TWI224925B - 4-field motion adaptive deinterlacing - Google Patents
- ️Wed Dec 01 2004
TWI224925B - 4-field motion adaptive deinterlacing - Google Patents
4-field motion adaptive deinterlacing Download PDFInfo
-
Publication number
- TWI224925B TWI224925B TW92113387A TW92113387A TWI224925B TW I224925 B TWI224925 B TW I224925B TW 92113387 A TW92113387 A TW 92113387A TW 92113387 A TW92113387 A TW 92113387A TW I224925 B TWI224925 B TW I224925B Authority
- TW
- Taiwan Prior art keywords
- field
- pixel
- block
- difference
- picture Prior art date
- 2003-05-16
Links
Landscapes
- Television Systems (AREA)
Abstract
This invention presented a universal deinterlacing method and system by the motion detecting, the motion adaptive interpolation, the directional edge interpolation, the 3:2 pull-downed detecting and the 3:2 pull-downed recovery. Four-field motion detection can detect the motion area and the still area. By cooperating with the motion adaptive interpolation and the directional edge interpolation the picture possesses the very clear quality, high contrast, and high resolution. The 3:2 pull-downed detecting and the 3:2 pull-downed recovery can detect the 3: 2 pull-downed movie sequences and recover the interlaced fields into a progressive frames universally without too much overhead of computational power.
Description
1224925 五、發明說明（1) · 一、【發明所屬之技術領域】 本發明係有關於圖場解交錯法，特別是關於以動像偵 測、動像補償、邊緣方向内差法、3 : 2拉下偵測與3 : 2拉下 還原之圖場解交錯法。 一、【先前技術】 於轉換掃目苗頻率（scanning rate conversion)的解交 錯（d e i n t e r 1 a c i n g)技術在現今已經變得越來越重要。交 錯影像序列（inter lace(i seq[uence s)通常是以每秒60個圖 場（f 1 e 1 d )傳遞予觀眾，為了將這些交錯影像序列顯示於 循序影像裝置（pr〇gressive devices，如電腦、高晝質電 視HDTV) ’這些循序影像裝置需要重建圖場。如果解交錯 做得不好，觀眾就要忍受不舒服的視訊失真 (uncomfortable visual artifacts)’ 如錯齒邊（edge f 1 icker)、抖線（！ ine crawl ing)的這類缺失將會在輸出 圖忙（frame )上造成嚴重扭曲（dist〇rti〇n)。 一循序&像放置疋使用非交錯（non i n t er 1 aced )影像顯 示規七又稱為循序掃瞄（progressive scan)，也就是整 個，幅（frame)是以一條線一條線由上到下掃瞎，以一個 固疋更新頻率反覆重新掃猫整個圖幅，如75Hz(每秒75次） 验相a-像電視信號，採用的是交錯影像顯示規格，是 Ϊ 了 ΐ !以交錯方式用2或3個圖場來顯示。美國的NTSC 電視"ίέ Ϊ虎採用兩個P v ^ & · 可彳U圖％以各自不同的水平線來顯示整個圖 1224925 五、發明說明（2) 幅的影像資料 這樣觀幕就 如果是以每秒 面閃動，這樣 會變得不明顯 兩倍資訊傳輸 2拉下，在以（ 秒6 0個圖場時 代表從圖場偶 瞄線產生的圖 代表是由同一 樣，如歐洲的 下方式來產生 每秒反覆掃瞄3 〇次 圖場的混合影像。 眾將會覺得影像畫 較向的掃猫頻率才 低閃動而不需要以 。這種方式稱為3 : 幅細序列轉換為每 奇偶奇偶奇u (偶 代表從圖場奇數掃 偶奇偶”與π奇偶，， 式影像規格也是一 1 i n e )，是以2 : 2拉 ’並且依續將這兩個圖場 會覺的是每秒各6 0個獨立 3〇個循序掃瞄的圖幅，觀 的閃動需要在像60Hz這樣 ’因此交錯式規格能夠降 頻率來每秒更新β 〇個圖幅 丨：2拉下方式將每秒2 4個圖 ，是採π偶奇偶奇偶”或，1 數掃瞄線產生的圖場，奇 場。）來產生圖場，其中” 圖幅所產出。其他的交錯 PAL(phase alternation 每秒5 0個圖框的方式。 以3 : 2拉下方式或2 : 2拉下由一個圖幅來產生二或三個 的圖％’被稱為循序影像圖場（fields in progressive v i deo ) ’也就是同一個圖幅所產生的圖場的時間點是一樣 的。相反地，每一個圖場都是不同時間點的影像，這些圖 场就稱為父錯影像圖場（fields in interlaced video)。 兩者在顯示上最大的差別為在同一時間點的不同循序影像 圖場中’動像（raotion)會出現在相同位置，而在交錯影像 圖場中’每一圖場時間點都不同，動像在圖場中出現的位 置都會不同。1224925 V. Description of the invention (1) · 1. [Technical field to which the invention belongs] The present invention relates to the field de-interlacing method, in particular to moving image detection, moving image compensation, edge difference method, 3: The field de-interlacing method of 2 pull-down detection and 3: 2 pull-down reduction. 1. [Previous technology] De-interleaving (d e i n t e r 1 a c i n g) technology for converting scanning rate conversion has become more and more important today. Interlaced image sequences (inter lace (i seq [uence s) are usually delivered to the audience at 60 fields per second (f 1 e 1 d). In order to display these interlaced image sequences on progressive devices, (Eg, computer, HDTV) "Sequential video devices need to reconstruct the field. If de-interlacing is not done well, viewers will endure uncomfortable visual artifacts" such as edge f 1 Such as icker), ine crawling (! ine crawling) will cause severe distortion (distortion) on the output frame. A sequential & like placement 疋 using non int er 1 aced) image display rules are also called progressive scan, that is, the whole, the frame is scanned from top to bottom with a line and a line, and the cat is scanned again and again with a fixed update frequency Picture format, such as 75Hz (75 times per second) phase-checking a-like television signals, adopts the interlaced image display specification, which is ΐ! It is displayed in an interlaced manner with 2 or 3 fields. NTSC TV in the United States " ί Ϊ Tiger uses two P v ^ & · You can display the entire figure in different horizontal lines with different horizontal lines. 1224925 V. Description of the invention (2) image data. So if the screen flashes on a per second surface, it will become inconspicuous. Twice the information transmission 2 is pulled down. When 60 fields are displayed in 60 seconds, the image generated from the field line of the field is represented by the same, such as the European method to generate repeated scans 30 times per second. The mixed image of the field. The public will think that the frequency of the cat's sweep is low and it does not need to flicker. This method is called 3: the thin sequence is converted to each parity and odd parity u (even represents the odd number from the field) Sweeping parity "and π parity, the format image specification is also 1 ine), and the two fields are pulled at 2: 2 and one after the other will see 60 independent 30 sequential scans per second Aiming at the picture frame, the flicker of the view needs to be at 60Hz, so the interlaced specification can reduce the frequency to update β 0 frames per second. 丨: The 2 pull-down method will take 2 4 pictures per second, which is π even parity. Odd or even ", 1 field generated by the scanning line, odd field.) To generate a field, Medium "format. Other interlaced PAL (phase alternation 50 frames per second. 3: 2 pull down mode or 2: 2 pull down to generate two or three pictures % 'Is called "fields in progressive vi deo", that is, the time points of the fields generated by the same frame are the same. In contrast, each field is an image at a different point in time. These fields are called fields in interlaced video. The biggest difference between the two displays is that 'raotion' will appear at the same position in different sequential image fields at the same time point, while in the interlaced image field, 'each field time point is different. Images appear in different places in the field.
第8頁 1224925Page 8 1224925
έ 、一圖幅可由一奇數掃瞄線圖場與一偶數掃瞄線圖場所 、、且成’換5之’是以奇數掃瞄線與偶數掃瞄線交錯所組成 人因此以解交錯方法將一圖場產生一圖幅係依目的圖場配 a另一互補圖場，將彼此之掃瞄線交錯所組成。解交錯方 法粗略分為三大類一圖場内内差法（intra —field interpolation)、圖場間内差法（inter — interpolation)與動像調適性解交錯法（m〇ti〇n adaptive deinterlacing)。 圖場内内差法就是以一個原始圖場來建構一個完整的 圖框。如第一 A圖，複數個目前像素丨〇組成一目前圖場之 掃猫線，經圖場内内差法解交錯後，產生複數個内差像素 1 2，每一内差像1 2素係根據相鄰之像素丨〇所產生，以目前 像素1 0與内差像素1 2交錯組成一圖幅之掃瞄線。最簡單也 是最普遍的做法是雙線性内差法。雙線性内差法是以相鄰 四個像素的加權值平均後來決定新圖場的像素，所以雙線 性内差法可視為四個加權内差函式的總和，如下所示： F(X，，y，）= aF(x，y + 1) + bF(x，y —1) + cF(x+i，y) + dF(x—iy) F(x，y )表不在新圖場中位於（χ，，y,)位置的新像素，並 且a、b、c、d為相鄰像素的加權 . )。由此根據原始目前圖場就」\^/^(weightlng factors 兩圖場組合成一個新的圖幅，$彻再啊口两亚將 w 、仏雒μ α 足個方法也稱之為編織（A picture frame can be composed of an odd-numbered scan line field and an even-numbered scan line field, and it can be changed into 5. The odd-numbered line and the even-numbered line are interlaced. The generation of a picture field from a picture field is made up of a complementary picture field according to the purpose field, and the scanning lines of each other are staggered. De-interlacing methods are roughly divided into three major categories: intra-field interpolation, inter-field interpolation, and motion adaptive deinterlacing. The intra-field difference method is to construct a complete picture frame from an original picture field. As shown in the first picture A, a plurality of current pixels 丨 0 form a sweeping cat line of the current field. After de-interlacing in the field internal difference method, a plurality of internal difference pixels 12 are generated, each of which is a 12-element According to the neighboring pixels, the scanning lines of a picture frame are formed by interleaving the current pixel 10 and the inner difference pixel 12 alternately. The simplest and most common method is the bilinear internal difference method. The bilinear internal difference method is based on the weighted values of four adjacent pixels and then determines the pixels of the new field, so the bilinear internal difference method can be regarded as the sum of the four weighted internal difference functions, as shown below: F ( X ,, y,) = aF (x, y + 1) + bF (x, y — 1) + cF (x + i, y) + dF (x—iy) F (x, y) New pixels in the field at (χ ,, y,), and a, b, c, and d are the weights of neighboring pixels.). Therefore, according to the original current field, "\ ^ / ^ (weightlng factors, the two fields are combined into a new frame, and $ Che Zai Ai Liangya also refers to w and αμ α as a method of weaving (
Weave)。此雙線性内差法力；息从，入 a «以#咖^ & 在邊緣上會變得模糊或呈現鋸齒 狀’因為疋以相鄰的像辛炎來 1豕京來+均’不過卻是圖場内内差法Weave). This bilinear internal difference mana; interest from, into a «以 # 咖啡 ^ & on the edges will become blurred or jagged 'because I use adjacent images like Xin Yan to come 1 to Beijing + to both' Intrafield difference method
U24925 五、發明說明（4) 中最簡單的。 圖場 者來重建 尤其是在 像素1 4分 内差法解 係有目前 差法是擺 場組合或 相鄰圖場 動像向量 重建一個 合的新圖 通常以調 曰内差法是以在前一圖場或後一圖場找出最合適 二個圖幅，畫質上會比單一圖場内差法好一些， 靜像區域。如第一 8圖，複數個目前像素1〇斑前 別組成一目前圖場與前圖場之掃瞄線，經圖場間 ^錯後，產生一圖幅之掃瞄線，此圖幅之掃瞄線 像素1 〇與前像素丨4交錯組成。最簡單的圖場間内 動（B 〇 b )擺動就只是單純將偶數圖場以奇數圖 將奇數圖場以偶數圖場組合。在動像預估法是在 中找出與目前圖場中最相似的區塊並且計算出其 ，然後=相鄰圖場與動像向量用一動像預估程^ =的圖，，這方法可以產生一個與原圖場完美符 但是去要花費太多時間，所以圖場間内差法 適性動像解交錯法取代。 像調？象解交錯法最重要之處在於找出哪裡是動 像Iee域，如果動德/自、 就會儘可能地高。在二::f正確地完成，則輸出影像畫質 動像區域是以計算出^ 5周適性動像解交錯法中，圖場間 圖場與後圖場差代表動場與後圖場不同之處，此目前 。接下來可能以圖場内=域，而其他部份即為靜像區域 間内差法用於靜像區域。差法用於動像區域，並且以圖場 能由目前像素！ 〇、前德丄如第一 0圖，一圖幅之掃瞄線可 素1 4與内差像素丨2所組成，其中目 五 前 像 測 像 部 質 發明說明（5) ___ :素上0與前像素14分別為 後^差像素12係根據相鄰;斤產猫線上的 4，圖幅掃瞄線上動像相關|以 去在經動像偵 相關像素以前像素組成14。’===素12組成，靜 份動像區域摘測不到，以二象相：此種做法可能會有 失真。 Μ邛饧以錯誤像素内差造成畫 由上述可 雨個圖場組合 快速動像影像 因此擺動不適 普遍線性技術 新的圖場，再 恭感到不舒服 有一半，因為 出來的。而動 此得出一個綜 解交錯法是最合適的。 知，擺動是-種簡單的解交錯方法， ，來而已，使用擺動的解交錯方式在有 (fast-mot ion >時會產生嚴重的畫質夕 用於每秒6 0次的快速動像影像圖場。 為編織，編織是以圖場内内差法:種 與：前的圖場組合’這樣的畫面就不备讓： ，然而編織還是有缺點，在垂直 ㈢讓關 Ϊ出圖幅是由目前圖場以圖場内内土只 J預估法畫質好可是卻要花大量的展 二f方優點並能準確針測動像的調適:氣因 合適的。 项性動像 【發明内容】 本毛月之-主要目的為解決邊緣雜齒狀或模糊的現象 本&月之另一主要目的為以3: 2拉下 還原重建原始圖U24925 5. The simplest in the description of invention (4). The field player comes to rebuild, especially in the pixel 1 4 point internal difference method. The current difference method is the combination of the field or the moving image vector of the adjacent field to reconstruct a new picture. The internal difference method is usually based on the former. One picture field or the latter picture field finds the most suitable two picture frames, the picture quality will be better than the single picture field difference method, the still image area. As shown in the first figure 8, a plurality of current pixels 10 dots before each constitute a scanning line of the current field and the previous field. After the field is wrong, a scanning line of a frame is generated. Scan line pixels 10 and front pixels 4 are staggered. The simplest inter-field (B 0 b) swing is simply to combine even-numbered fields with odd-numbered fields and odd-numbered fields with even-numbered fields. The moving image estimation method is to find the most similar block in the current field and calculate it, and then use a moving image estimation process ^ = for the adjacent field and the moving image vector. This method It can produce a perfect match with the original field but it takes too much time, so the inter-field difference method and the adaptive moving image de-interlacing method are used instead. Like tone? The most important aspect of the image deinterlacing method is to find out where the dynamic image Iee domain is. If it is dynamic / self, it will be as high as possible. When 2 :: f is completed correctly, the output image quality moving image area is calculated by ^ 5 weeks. The adaptive moving image de-interlacing method indicates that the difference between the picture field and the back picture field represents that the moving field is different from the back picture field. Where this is currently. Next, the intra-field = domain may be used, and the other parts are used for the still image area. The difference method is used in the moving image area, and the field can be calculated from the current pixel! 〇 The former German figure is like the first picture 0, the scanning line of a picture frame can be composed of 14 and the inner difference pixel 丨 2, in which the description of the image of the front image of the fifth image (5) ___: prime 0 And the front pixel 14 is the rear ^ difference pixel 12 is based on the adjacent; 4 cat line on the cat line, the image scan line is related to the moving image | to make the pixel 14 before the moving image detection related pixels. ’=== Primary 12 composition, the static moving image area can not be measured, and the two images: this method may be distorted. Μ 邛 饧 was created with the wrong pixel internal difference. The combination of the above-mentioned rainy field is a fast moving image, so the swing is uncomfortable. The universal linear technology, the new field, is still half uncomfortable because it comes out. It is most appropriate to derive a comprehensive de-interlacing method. It is known that wobble is a simple de-interlacing method, and that ’s it. The de-interlacing method using wobble will produce severe picture quality when fast-mot ion is used. It is used for fast moving images 60 times per second. The image field. For weaving, weaving is based on the internal difference method of the field: the kind and the combination of the previous fields. This kind of picture is not ready: However, weaving still has disadvantages. From the current field, the image quality in the field is only good, but it takes a lot of advantages and can accurately measure the adjustment of the moving image: the gas factor is suitable. 】 Motoyuki-The main purpose is to solve the phenomenon of edge teeth or blurring. The other main purpose of Motoyuki is to restore the original picture with 3: 2 pull down.
第11 ·頁 1224925 五、發明說明（6) 本發明的再一主要目的為在偵測不到3 .· 2拉下模式時 在同時兼顧效率與晝質下將交錯圖場解交錯成循序式掃 瞄圖幅。 本發明提出一種四圖場動像的調適性解交錯方法，包 括：接收一圖場序列中之一四圖場、執行一動像偵測、產 生一同位圖場差序列、執行一 3 : 2拉下偵測、產生一圖場 解交錯結果。同時，本發明亦提出一種四圖場動像的調適 性解交錯系統，包括：一第一儲存裝置、一第二儲存裝置 一第三儲存裝置、一圖場資料流、一動像偵測裝置、一 動像補償裝置、一邊緣方向内差裝置、一像素切換裝置、 一圖場切換裝置、一 3: 2拉下還原裝置與一輸出切換裝置 另外，本發明提出一種四圖場動像的調適性解交錯方 法，包括：接收一圖場序列中之一四圖場、執行一動像偵 測與產生一圖幅。同時，本發明亦提出了一本發明亦提出 一種四圖場動像的調適性解交錯系統，包括：一第一儲存 裝置、一第二儲存裝置、一第三儲存裝置、一圖場資料流 、一動像偵測裝置、一動像補償裝置、一邊緣方向内差裝 置、一像素切換裝置與一圖場切換裝置。Page 11 · Page 1224925 V. Description of the invention (6) Another main purpose of the present invention is to de-interleave the interlaced field into a sequential pattern under the simultaneous consideration of efficiency and day quality when the pull-down mode of 3.2 is not detected. Scan the frame. The invention proposes a adaptive de-interlacing method for a four-picture field moving image, which includes: receiving a four-picture field in a field sequence, performing a moving image detection, generating a bit-field field difference sequence, and performing a 3: 2 pull Detect and generate a field deinterlacing result. At the same time, the invention also proposes an adaptive de-interlacing system for a four-picture field moving image, including: a first storage device, a second storage device, a third storage device, a field data stream, a moving image detection device, A moving image compensation device, an edge direction difference device, a pixel switching device, a field switching device, a 3: 2 pull-down reduction device, and an output switching device. In addition, the present invention proposes a four-field motion image adaptability. The de-interlacing method includes receiving a four-field in a field sequence, performing a moving image detection, and generating a frame. At the same time, the invention also proposes an adaptive de-interlacing system for a four-picture field moving image, which includes: a first storage device, a second storage device, a third storage device, and a field data stream. , A moving image detection device, a moving image compensation device, an edge direction difference device, a pixel switching device and a field switching device.
第12頁 1224925 五、發明說明（7) 為了解決在上述方法與系統中於畫面上之 能出現模糊或鋸齒邊的問題，本發明亦提出—邊緣方向可 差法來解決這樣的問題。此邊緣方向内差法^、'緣方向内 掃瞄線與一第二掃瞄線產生一介於第一掃瞄=根據—第一 線間之一邊緣方向内差掃瞄線之每一邊緣方、、。與第二掃瞄 各邊緣方向内差像素係複數對像素對中最相向内差像素， 平均值，這些像素對係以邊緣方向内差像^ $之像素對之 値不同方向於第一掃瞄線與第二掃瞄線各取二中心由複數 ” 一像素產生。 最後，本發明提出一種3: 2拉下解交鈣 接收-圖場序歹.】、產生每—第_圖場之—日位，包括： 艮一同位圖場差序列、執行3:2拉下 ^立圖％差、記 偵測結果、解交錯該圖場序列以 、—產^生—3 : 2拉下 圖^列。】日夺，本發明亦提出—種3 拉幅下序解列與輸公 ，包括··一第一儲存裝置、一 、曰系、冼 流、一 3 : 2拉下偵測裝置、一 3 •下置、一圖場資料 交錯裝置與一圖場輸出裝置。·遇原裝置、一圖場解 、動ί述邊ί發明中提出了一個具有四圖場動像偵測 動像補彳貝邊緣方向内差法、3 · 2拉下相：ρ,丨应3 . 2；te nr # 原的調適性動傻紐一 w t ύ ·罐下偵測興J · 2拉下退 能區分出動像盥靜^ rs Τ法與系統。四圖場動像偵測不僅 法所呈現的畫面ϊίί〗’配合動像補償與邊緣方向内差 。另外，3. ?知有m析的畫質、高度的對比與高解析度 .拉下偵測與3 : 2拉下還原能以少量的運算成本 1224925 五、發明說明（8) 將交錯圖場還原成循序式圖幅。 . 本發明的這些特徵與優點皆描述於下，並結合於請求 保護項中。 四、【實施方式】 本發明的一些實施例會詳細描述如下。然而，除了詳 細描述外，本發明還可以廣泛地在其他的實施例施行，且 本發明的範圍不受限定，其以之後的專利範圍為準。 再者，為提供更清楚的描述及更易理解本發明，圖示 内各部分並沒有依照其相對尺寸繪圖，某些尺寸與其他相 關尺度相比已經被誇張；不相關之細節部分也未完全繪出 ，以求圖示的簡潔。 由於本發明中所用到的圖場間解交錯法（如搖擺）與圖 場間解交錯法（如擺動），已於先前技術中詳細說明，故下 述說明中之描述不包含其完整流程。 本發明首先提出一種四圖場動像的調適性解交錯方法 0 。第二A圖為本發明第一較佳實施例之示意圖。首先，如 步驟2 0 0所示，接收一圖場序列中之四圖場，此四圖場係 一第一圖場、一第二圖場、一第三圖場與一第四圖場且依 · 序由先而後被接收。其中，此圖場序列之每一圖場與其同Page 12 1224925 V. Description of the invention (7) In order to solve the problem that blurry or jagged edges can appear on the screen in the above method and system, the present invention also proposes-edge direction difference method to solve such problems. This edge direction internal difference method ^, 'the scan line within the edge direction and a second scan line generate an interval between the first scan = according to-each of the edge lines ,,. The inner difference pixels in each edge direction of the second scan are the most opposite inner difference pixels in the plurality of pixel pairs. On average, these pixel pairs are in the first scan with different directions of the pixel difference in the edge direction ^ $. The line and the second scanning line each take two centers and are generated by a complex number of one pixel. Finally, the present invention proposes a 3: 2 pull-down disjoint calcium reception-field sequence sequence. Day position, including: Gen map bit field difference sequence, perform 3: 2 pull down ^ erection% difference, record detection results, deinterleave the field sequence to,-produce ^ 3-2 pull down picture ^ Column.] The day of the win, the present invention also proposes-a kind of 3 stencils in order to delist and lose, including: a first storage device, a system, a stream, a 3: 2 pull down detection device, 1 3 • Lower, a field data interleaving device and a field output device. · The original device, a field solution, and moving edges are described in the invention. A four-field motion image detection motion image supplement is proposed.彳 Shell edge direction internal difference method, 3 · 2 pull down the phase: ρ, 丨 Ying 3.2; te nr # The original adaptability to move a stupid one wt ω · detection under the jar · 2 pull down can distinguish the moving image ^ rs Τ method and system. The four-picture field moving image detection method is not only the picture presented by the method 配合〗 '' combined with the moving image compensation and edge difference. In addition, 3. There are image quality, height contrast and high resolution analysis. Pull-down detection and 3: 2 pull-down reduction can be performed with a small amount of calculation cost 1224925 V. Description of the invention (8) The interlaced field is restored into a sequential frame .. These features and advantages of the present invention are described below, and are incorporated in the claims. [Embodiment] Some embodiments of the present invention will be described in detail below. However, in addition to the detailed description, the present invention can also be widely The ground is implemented in other embodiments, and the scope of the present invention is not limited, which is subject to the scope of the subsequent patents. Furthermore, in order to provide a clearer description and easier understanding of the present invention, the parts in the diagram have not been in accordance with it Relative size drawings, some dimensions have been exaggerated compared to other related scales; irrelevant details have not been completely drawn in order to simplify the illustration. Because the inter-field de-interlacing method used in the present invention ( The method of de-interlacing between field and field (such as swing) has been described in detail in the prior art, so the description in the following description does not include its complete flow. The present invention first proposes an adaptive de-interlacing of four-field field image Method 0. The second picture A is a schematic diagram of the first preferred embodiment of the present invention. First, as shown in step 200, four picture fields in a picture sequence are received, and the four picture fields are a first picture field. , A second field, a third field, and a fourth field are received in order, one after the other. Among them, each field in this field sequence is the same as
第14頁 1224925 五、發明說明（9) 位圖場係在以不同時間下，以相同複數個像素位置之像素 . 所構成，並且，這些像素位置構成複數條掃目苗線。換句話 說，每個圖場有數條掃瞄線，每一條掃瞄線有數個像素， 每一個像素位於一個像素位置上，並且，在不同時間下， 同一像素位置可能為不同圖場的某一像素。然後，如步驟 ， 2 1 0所示，執行一動像偵測，用以產生一動像偵測結果。 接下來，如步驟2 2 0所示，產生一同位圖場差序列，此同 位圖場差序列係由動像偵測結果所產生。再接下來，如步 驟2 3 0所示，執行一 3 : 2拉下偵測，此3 : 2拉下偵測係依同 位圖場差序列產生一 3 : 2拉下偵測結果。最後，如步驟2 4 0 所示，產生一圖場解交錯結果，此圖場解交錯結果係依3 : 2拉下結果以一圖場解交錯法產生。 另外，上述同位圖場差序列之產生方法，包含產生第 三圖場之一同位圖場差與記錄一同位圖場差序列。如第二 D圖所示，首先，如步驟2 2 1所示，產生該第三圖場之一同 位圖場差，第三圖場之一同位圖場差係根據該動像動像偵 測產生。此第三圖場之同位差可用第一圖場與第三圖場間 差異像數總和表示，或用是否相似表示（如只記錄是、否 來分別表示相似與不相似）。然後，如步驟2 2 2所示，記錄 0 一同位圖場差序列，此同位圖場差序列係於第三圖場之同 位圖場差產生後記錄，此同位圖場差序列係由包含第三圖 場之同位圖場差及向前推算之複數個同位圖場差所組成。Page 14 1224925 V. Description of the invention (9) The bitmap field is composed of pixels at the same plurality of pixel positions at different times, and these pixel positions constitute a plurality of scanning lines. In other words, each field has several scanning lines, each scanning line has several pixels, each pixel is located at a pixel position, and at different times, the same pixel position may be a certain field in different fields. Pixels. Then, as shown in step 2 10, a moving image detection is performed to generate a moving image detection result. Next, as shown in step 220, a bitmap field difference sequence is generated, and the bitmap field difference sequence is generated from the motion image detection result. Next, as shown in step 230, a 3: 2 pull-down detection is performed, and the 3: 2 pull-down detection generates a 3: 2 pull-down detection result according to the isotopic field difference sequence. Finally, as shown in step 2 40, a field de-interlacing result is generated. The field de-interlacing result is generated by a 3: 2 pull-down result using a field de-interlacing method. In addition, the generating method of the above-mentioned bitmap field difference sequence includes generating a bitmap field difference sequence of one of the third field and recording the bitmap field difference. As shown in the second diagram D, first, as shown in step 2 21, an isotopic field difference of the third field is generated. The isotopic field difference of the third field is detected based on the moving image. produce. This third field's isoparallel can be expressed by the sum of the difference image number between the first field and the third field, or whether it is similar (such as recording only yes and no to indicate similarity and dissimilarity respectively). Then, as shown in step 2 2 2, record 0 with the bitmap field difference sequence. This isotope field difference sequence is recorded after the third bitfield is generated. This isotope field difference sequence is composed of The three-field field is composed of the same-field image field difference and a plurality of the same-field field fields calculated forward.
第15頁 1224925 五、發明說明（ίο) 上述3 : 2拉下偵測係於該同位圖場差序列中搜尋複數 _ 個起始圖場差，該些起始圖場差皆小於一第二門檻限值且 每隔五次出現一次。此同位圖場差可以為兩同位圖場中各 像素位置之兩相關像素之差值的總和。此兩相關像素分別 屬於此兩同位圖場，並且此差值可以為亮度差、彩度差、 . 三原色（RGB)差等等。另外，同位圖場差也可用是否相似 的方式來儲存，「是」（true)表示小於第一門檻限值，反 之，「否」（f a 1 s e )表是不小於第一門檻限值。此外，上 述3 : 2拉下偵測結果可分別為已偵測出3 : 2拉下與尚未偵測 出3 : 2拉下來表示。例如，當起始圖場差之總數超過一第 .. 三門檻限值，則3 : 2拉下偵測結果為已偵測出3 : 2拉下，否 則，3 : 2拉下偵測結果為尚未偵測出3 : 2拉下。依不同的3 : 2拉下結果以不同的圖場解交錯法來產生圖場解交錯結果 。例如，當3 : 2拉下偵測結果為已偵測出3 : 2拉下時，以一 3 : 2拉下還原法來產生圖場解交錯結果，或當3 .· 2拉下偵測 結果為尚未偵測出3 : 2拉下時，則以圖場内内差法、圖場 間内差法或調適性解交錯法來產生圖場解交錯結果。其中 ，調適性解交錯法係以圖場内内差法與圖場間内差法分別 解交錯該動像區域與該靜像區域。上述動像區域係由動像 偵測結果所產生，而在一圖場中動像區域以外之區域為靜 0 像區域。例如，當3 : 2偵測結果為尚未偵測出3 : 2拉下時， 則圖場解交錯結果為一圖幅，此圖幅以由第三圖場與一内 差圖場之各掃瞄線交錯所產生，此内差圖場具有與第二圖 . 場相同之各像素位置，並且以各像素位置之内差像素所組Page 15 1224925 V. Description of the invention (ίο) The above 3: 2 pull-down detection is to search for a plurality of initial picture field differences in the isotopic field difference sequence, and these initial picture field differences are less than a second Thresholds and occur every five times. This isomap field difference can be the sum of the differences between two related pixels at each pixel position in the two comap fields. The two related pixels belong to the two co-located image fields, respectively, and the difference can be a difference in brightness, a difference in chroma, a difference in three primary colors (RGB), and so on. In addition, the field map of the same bitmap can also be stored in a similar way. "True" indicates that it is less than the first threshold value; otherwise, "no" (f a 1 s e) table is not less than the first threshold value. In addition, the detection results of the 3: 2 pull-down described above can be respectively indicated as 3: 2 pull-down detected and 3: 2 pull-down not detected. For example, when the total number of field differences in the initial picture exceeds the first threshold of three thresholds, a 3: 2 pulldown detection result is detected as a 3: 2 pulldown, otherwise, a 3: 2 pulldown detection result Pull down for 3: 2 not detected yet. According to different 3: 2 pull down results, different field de-interlacing methods are used to generate field de-interlacing results. For example, when 3: 2 pull-down detection result is that 3: 2 pull-down has been detected, a 3: 2 pull-down reduction method is used to generate a field de-interlacing result, or when 3. · 2 pull-down detection As a result, when the 3: 2 pull-down has not been detected, the intra-field intra-field difference method, inter-field intra-field difference method, or adaptive de-interlacing method is used to generate the field de-interlacing result. Among them, the adaptive de-interlacing method de-interleaves the moving image area and the still image area by the intra-field difference method and the inter-field difference method, respectively. The above moving image area is generated by the result of the moving image detection, and the area other than the moving image area in a field is a static 0 image area. For example, when the 3: 2 detection result is that the 3: 2 pull-down has not been detected, the field de-interlacing result is a frame, and the frame is scanned by the third field and an internal difference field. The interstitial field has the same pixel position as that in the second field.
第16頁 1224925 五、發明說明（11) 成而所有内差像素係分別以圖場内内差法與圖場間内 法解交錯動像區域與靜像區域所產生。此外，上述之 内=法與圖場間内差法可分別為邊緣方向内 動： 補償法。 7 1豕 當3: 結果係以 圖場之 圖場組合 生第一圖 圖幅與一 場與第五 第四圖幅 圖場之下 、第四圖 2偵測…果為已偵測出3 : 2拉下時，則圖場解交錯 了 3 : 2拉下還原法產生，此3: 2拉下還原法係當第 同位圖場差為起始圖場差時，以第一圖場與第二 ，交錯或以第二圖場與第三圖場組合解交錯以^ 場二第二圖場與第三圖場之一第一圖幅、一第二 第二圖幅，再將第四圖場與一第五圖場以第四圖 圖j組合解交錯以產生第四圖場與第五圖場之一 2 一 =五圖幅，此第五圖場係於圖場序列中第四 一圖場。其中，第一圖幅、第二圖幅、第三圖幅 幅與第五圖幅係連續的五個圖幅。 2U所第-二^為上述之動像偵測之示意圖。首先，如步驟 由複數不個像/戶Λ第三圖場成複數個原區塊’各原區塊係 去第二ί 1步驟212所示，偵測各原區塊之第一區塊， :個相Γ二中原區塊之相對位置以一搜尋範圍内存在至少 相;Γΐ”；則第一區塊存在，並且為最接近原區塊ΐ ，、置之相似區塊，否則，此第一區塊不存在。1 ^相似區塊係與該原區塊之一第一 /、，Page 16 1224925 V. Description of the invention (11) All internal difference pixels are generated by de-interlacing the moving image area and the still image area by the intra-field difference method and the inter-field method. In addition, the above-mentioned internal = method and the inter-field difference method can be the edge-direction internal motion: compensation method. 7 1Dang 3: The result is a combination of the first field and the fifth field and the fourth and fourth fields, and the fourth and second fields are detected based on the combination of the fields of the field. When 2 is pulled down, the field is de-interlaced by the 3: 2 pull-down reduction method. This 3: 2 pull-down reduction system uses the first field and the first field difference when the first isotopic field difference is the initial field difference. Second, interleave or deinterleave with the combination of the second field and the third field to ^ field two one of the second field and the third field of the first frame, a second second frame, and then the fourth frame The field and a fifth field are de-interleaved by combining the fourth field j with the fourth field j to generate one of the fourth field and the fifth field. 2 = 5 frames. This fifth field is the fourth field in the field sequence. Picture field. Among them, the first, second, third, and fifth frames are five consecutive frames. 2U place-2 ^ is a schematic diagram of the above-mentioned moving image detection. First, as shown in step 212, the first block of each original block is detected from a plurality of original blocks in the third field into a plurality of original blocks. : Phase Γ The relative position of the original block in the second phase exists in at least one phase within a search range; Γΐ "; then the first block exists and is the closest block to the original block ，, otherwise, the first block A block does not exist. 1 ^ A similar block is one of the original block.
^塊差小於一門檻限值^ Block difference is less than a threshold value
IEH 1224925 五、發明說明（12) 接下來，如步騾21 3所示，古十笪夂居厂么 '當第一區塊存在時，此第 ^异各原Q塊之第一向量， ;至原區塊之相對位置：：向：；從區塊之相對位置 丨來，如步驟21 4所示，偵測各 向里不存在。再接下 丨區塊位於第二圖場中第二 ^ *之第二區塊，此第二 |位置中央，當一;，t相對位置與原區塊之相對 則，該第二區塊以塊；；'，：該第二區塊存在，否 丨距離第二區塊一第一向第四圖場t，並且為 區塊之-第二區塊差小於當第三區塊與第二 ’否則’此第三區塊不存在。值，則此第三區塊存在 |异各原區塊之移動向量，者二 2步驟21 6所不，計 量係從第二區塊至原區田=£鬼存在時’則此移動向 !，此原區塊的i動是：:Ί原區塊不存在。另外 I位置的移動向量。 〃區塊所對應到之所有像素 |對的ί;第；= 塊之所有相關像* :區塊。同樣地，第二分區塊與該* 有相關像素對的總和，1由，—弟—&塊與第三區塊中所 別屬於第二區塊與第三區塊。：中：m:之各像素分 差、彩度差等等。 宁所述之差值可以是亮度 此外，上述動像補償法係以第 圖場上之一第— 像素 1麵 I麵IEH 1224925 V. Description of the invention (12) Next, as shown in step 骡 21 3, is the ancient ten dwelling house 'when the first block exists, the first vector of this original Q block is different ,; To the relative position of the original block :: To :; From the relative position of the block, as shown in step 21 4, detect the absence of each direction. The next block is located in the second block of the second field in the second field. This second | position is in the center. When one ;, the relative position of t is relative to the original block, the second block starts with Block;;,: the second block exists, no 丨 the first to fourth field t from the second block, and the difference between the second block and the second block is less than when the third block and the second block 'Otherwise' this third block does not exist. Value, then this third block exists | different from the original block's movement vector, or the second step 2 21 6 is not, the measurement is from the second block to the original area == when the ghost exists' then this move to! The i-movement of this original block is: The Ίyuan block does not exist. In addition, the motion vector at I position.所有 All pixels corresponding to the block | of the pair; the first; = all relevant images of the block *: block. Similarly, the sum of the pixel pairs associated with the second sub-block and the *, 1 by, the --- brother and the third block belong to the second block and the third block. : Medium: m: Pixel difference, chroma difference, and so on. In addition, the difference may be brightness. In addition, the above-mentioned moving image compensation method is based on
第18頁 U24925 五、發明說明（13) _______ I來產生一内差像素，從此第一 f像素所在之像素位置為此 :所在之像素位置至内差 |動向量。 此内差像素所對應之像素位置的移 第二C圖為上述動像僧一立 圖場2 6、第三圖場2 7虚楚、不思圖’第一圖場2 5、第 偵測第三圖場27之原區^ 圖,％ 28為連續之四圖場，為 第一區塊250，此第一區* ’頁先於第一圖場25找尋一 一區塊250有找到的話，計管^原〃區塊極為相似。如果第 原區塊2 7 0相對位置之第二才曰彳之第一區塊2 5 0相對位置至 位於第一區塊2 5 0相對位番。接下來，於第二圖場2 6 二區塊261與在第四圖場 ，、區塊270相對位置中央之第 一向量驗第三區；= 三區塊281極為相似， 1乂如果第二區塊261與第 位置至原區塊270相對位番$向量為從第二區塊251相對 區塊270時，則用動像補J法=圖場間内差法解交錯原 產生圖幅29之一圖幅區塊“I 一區塊261與原區塊27〇 此外，所有移動 動像 於些 靜像 動向 屬於 動之 塊屬 中， 之移 量的 像素 ，可視為動像區域里：子在之原區塊屬於快速移 微移動之動像或去 之，具有移動向量之原區 區域中各個像靜像，可視為靜像區域。其 量。換句話說，f之移動向量亦為所屬原區塊 動像區域，即動俊：：像素位置中，$有移動向 動像&域為所有移動向量不存在之Page 18 U24925 V. Description of the invention (13) _______ I to generate an internal difference pixel, from which the pixel position of the first f pixel is: the pixel position to the internal difference | motion vector. The second C picture of the shift of the pixel position corresponding to this inner difference pixel is the above-mentioned moving image monk-upright field 2 6 and the third field 2 7 The original area of the third field 27 ^ Figure,% 28 is the fourth field in a row, which is the first block 250. This first area * 'pages look for the first field 25 before the first block 250 is found. The accounting ^ original block is very similar. If the second block relative position of the original block 270 is the second relative position of the first block 250, it is located at the relative position of the first block 250. Next, in the second field, the second block 261 and the fourth field, the first vector in the center of the relative position of the block 270, check the third region; = three blocks 281 are very similar, 1 if the second The relative position between block 261 and the original block 270 is $. When the vector is from block 251 to block 270, the moving image supplement J method = inter-field difference method is used to deinterleave the original image frame 29 One of the picture block "I-block 261 and the original block 27. In addition, all the moving images in the still image movement belong to the moving block genus, and the pixels of the moving amount can be regarded as the moving image area: The original block in the original block belongs to the fast moving micro-moving or moving image. The still image of each image in the original region area with the motion vector can be regarded as the still image area. Its amount. In other words, the motion vector of f also belongs to The motion image area of the original block, that is, motion :: In the pixel position, $ has a motion to the motion image & field, where all motion vectors do not exist
第19頁 1224925 五、發明說明（14) 位置的聯集。反之，靜 — 位置的聯集。 4所有移動向量存在之像素 上述邊緣方向内差法係 =，來，生一介於第一掃•線〜掃瞒線與一第二掃 方向内差柃瞄線之每一邊緣方仓^弟〜掃瞄線間之一邊 ί Ϊ素係複數對像素對中最相：之J像素，各邊緣方向内 該掃猫線與該第線：：心：複數個不同方向； =個不同方向係可用接“i:;;素產生。其中上述 下方向、左左上至右右下方丄至 丨步驟匕緣方向内差法，其 於第三圖場之一第圖場之掃晦線係； ㈣像素;間；產生各動 具來產生，此最相似像=似 t素對，係以各動像像辛動”素相關之複數個 卜線與第二掃晦線各象；=產由；數個不同方向於第一 因此，本發明夕| 透緣方向内差法Page 19 1224925 V. Description of the invention (14) Union of positions. Conversely, static — a union of locations. 4 All pixels with motion vectors exist within the above-mentioned edge direction difference method =, then, generate one edge between the first scan line ~ the concealment line and a second scan direction inside difference. One edge between the scanning lines is a complex pixel pair with the most phase pair: J pixels, the cat line and the second line in each edge direction :: heart: a plurality of different directions; = different directions are available for connection "I: ;; prime production. The above-mentioned downward direction, left-upper-left, right-lower-right, and 丨 -step edge difference internal difference method, which is one of the third field of the first field of the scan field line; ㈣ pixels; The most similar image is the t-prime pair, which consists of the multiple lines and the second scan line that are related to each other by the moving image element; = production reason; several Different directions are first. Therefore, the present invention | Peripheral difference method
月之第二較佳實施例為— 第20頁 1224925 五、發明說明（15) ，第四圖為邊緣方& .μ女插# μ 向内差法之示意圖。第k-1線與第k+1線 為一上方#猫線4 2愈 a 俱右下方掃瞄線4 6，在此二掃瞄線間以 认# ^又争各 斤、線4 4作為第赚° X ( η，k )表示此新線4 4 的第η個像素，而μ ^ ,0 , ,or fc 上方原知瞄線42中之像素421、422、423 、4 2 4、 42 5-御 ^ττ ^ μ a 二万原掃瞒線中之像素461、462、463、464 、， /、之相鄰像素，此像素係以下列步驟產生。 百先，如下列計算出五個方向的像素差（a、b、c、d、e) a=[X(n-2, k〜1)一χ(η + 2, k+1 )] b=[X(n-l，k+1)] c=[X(n，k-1)〜χ(η，k+1)] d=[X(n+1，k〜l)-X(n-1，k+1)] e=[X(n + 2, k〜l) 一x(n — 2, k+1)] 法產生此新像素443。 、左上到右下方向、右 向與右右上至左左下方 對，以内差法產生此新 此新像素相鄰之兩原掃 内差法能偵測並重建出 k成之鋸齒邊。以五個 本务明其中一個例子， 本叙明中並不受限制。 然後沿這差值最小方向以内差 亦即以此新像素為中心之垂直方向 上至左下方向、左左匕5六六下士 向等五個方向找出差異最小之像素 像素，這些像素對係以此五方向於 猫線各取一像素產生。用邊緣方向 邊緣來避免可能出現之突出像素所 方向兩個像素的邊緣方向偵測僅為 以幾個方向來做邊緣方向内差法於 據 此，本發明之一第三較佳實 施例為一種四圖場動像The second preferred embodiment of the month is-page 20 1224925 V. Description of the invention (15). The fourth figure is a schematic diagram of the edge square & The k-1th line and the k + 1th line are one above. # 猫 线 4 2yua Ju the lower right scanning line 4 6, between the two scanning lines to recognize # ^ and fight each catty, line 4 4 as The first earned ° X (η, k) represents the n-th pixel of this new line 4 4, and the pixels 421, 422, 423, 4 2 4, 42 in the original sight line 42 above μ ^, 0,, or fc 5- 御 ^ ττ ^ μ a Pixels 461, 462, 463, 464, and /, adjacent pixels in the 20,000 original sweep line, this pixel is generated by the following steps. Baixian, calculate the pixel difference in five directions (a, b, c, d, e) as follows: a = [X (n-2, k ~ 1) -χ (η + 2, k + 1)] b = [X (nl, k + 1)] c = [X (n, k-1) ~ χ (η, k + 1)] d = [X (n + 1, k ~ l) -X (n- 1, k + 1)] e = [X (n + 2, k ~ l)-x (n — 2, k + 1)] method to generate this new pixel 443. , Upper left to lower right direction, right direction and right upper right to left lower left pair, this new is generated by the inner difference method. Two original scans adjacent to this new pixel. The inner difference method can detect and reconstruct k-tooth jagged edges. Take one of the five principals as an example. There are no restrictions in this statement. Then find the smallest difference in the direction of the smallest difference and the inner difference, that is, the vertical direction from the new pixel to the lower left direction, and the left and right direction. These five directions are generated by taking one pixel from each cat line. The edge direction edge is used to avoid possible protruding pixels. The edge direction detection of two pixels is only based on the edge direction internal difference method. Based on this, one of the third preferred embodiments of the present invention is a Four-picture field dynamic image
1224925 五、發明說明（16) 的 調適性解交錯方法，包括接收一圖場序列中之一 執行一動像偵測與產生一圖幅。第三圖所示 1四圖場 動像的調適性解交錯方法之示意圖。首先，如:$四圖場 示，接收一圖場序列中之一四圖場，此四圖場^〇所 一第二圖場、一第三圖場與一第四圖場且依=第一圖 後被接收，其中，圖場序列之每一圖場與其同^序由先而 不同時間下，以相同複數個像素位置之二‘;所$場係在 且，這些像素位置構成複數條掃瞄線。然後，如步= 所示，執行前述之動像偵測，產生該第三圖場之二動像1 域，此動像區域外之區域為一靜像區域。最後，如步驟 3 3 0所示，產生一圖幅，此圖幅圖幅係由該第三圖場與一 内差圖場之各掃瞄線交錯所產生，此内差圖場具有與第二 圖％相同之各像素位置’並且以各該像素位置之一内差像 素所組成，而所有内差像素係分別以圖場内内差法與前述 之動態補償法解交錯動像區域與靜像區域所產生。^中， 圖％内内差法可為邊緣方向内差法。 —本發明進一步提出一 3: 2拉下解交錯方法，如第五圖 ’第五圖為本發明之第四較佳實施例之示意圖。首先，如 步驟510所示，#收一圖場序列，此圖場序列係由 貢料流所接收；然後，如步驟520所示，產生每一第一圖 :之圖場差：此同位圖場差係於每次此圖-場序-列之 =收完成時產生’ &同位圖場差係根據第一圖場 ” 一第一圖場所產生’此第二圖場係為第一圖場之前同位1224925 5. The adaptive de-interlacing method of invention description (16) includes receiving one of a field sequence, performing a moving image detection and generating a frame. Figure 3 shows a schematic diagram of the adaptive de-interlacing method of the four-picture field. First of all, as shown in the $ 4 field, one of the four fields in a sequence of fields is received. The four fields are a second field, a third field, and a fourth field. A picture is received later, where each field of the field sequence and its sequence are different from each other at the same time at the same number of pixel positions'; the fields are in the same, and these pixel positions constitute a plurality of Scan line. Then, as shown in step =, the aforementioned moving image detection is performed to generate two moving image 1 fields of the third field. The area outside the moving image area is a still image area. Finally, as shown in step 3 3 0, a picture frame is generated. The picture frame is generated by the interlaced scanning lines of the third field and an internal difference field. The internal difference field has The two image locations are the same at each pixel position 'and are composed of one inner difference pixel of each pixel position, and all the inner difference pixels are deinterlaced between the moving image area and the still image using the field internal difference method and the aforementioned dynamic compensation method, respectively. Area. In the figure, the inner difference method of the graph% may be the inner difference method of the edge direction. -The present invention further proposes a 3: 2 pull-down de-interlacing method, such as the fifth figure. The fifth figure is a schematic diagram of the fourth preferred embodiment of the present invention. First, as shown in step 510, #receive a field sequence, this field sequence is received by the material stream; then, as shown in step 520, each first picture: the field difference of the picture: this co-location map The field difference is generated every time when this picture-field sequence-column = received is completed. &Amp; isotopic field difference is generated according to the first field "a first field" This second field is the first field Pre-court parity
第22頁Page 22
1224925 五、發明說明（17) 圖場，亦即為第一圖場之前一圖場之前一圖場；接下來， 如步驟5 3 0所示，記錄一同位圖場差序列，此同位圖場差 序列係於第一圖場之同位圖場差產生後記錄，並且同位圖 場差序列係由包含第一圖場之同位圖場差及向前推算之複 數個同位圖場差所組成；更進一步地，如步驟5 4 0所示， 執行上述之3 : 2拉下偵測以產生一 3 : 2拉下偵測結果，此3 : 2拉下偵測結果係依同位圖場差序列產生；於執行3 : 2拉下 债測之後，如步驟5 5 0所示，解交錯圖場序列以產生一圖 幅序列，此圖幅序列係依3 : 2拉下偵測結果以一圖場解交 錯法產生；最後，如步驟5 6 0所示，輸出此圖幅序列，此 圖幅序列係由一圖幅資料流輸出。 上述之同位圖場差可為第一圖場與其前同位圖場之差 異像素數量。當此同位圖場差小於一第一門檻限值時，可 視為此第一圖場與其同位圖場相同。另外，同為圖場差也 可用是否相似的方式來儲存，「是」（true)表示小於第一 門檻限值，反之，「否」（f a 1 s e )表示不小於第一門檻限 值。 據此，當3 : 2拉下偵測結果為尚未偵測出3 : 2拉下時， 則以一般圖場解交錯法，例如，圖場内内差法、圖場間内 差法與調適性解交錯法。當3 : 2拉下偵測結果為已偵測出 3 : 2拉下時，則以3 : 2拉下還原法產生圖幅序列。例如，每 一同位圖場差為根據此第一圖場與一第二圖場所產生，此1224925 V. Description of the invention (17) The field, that is, the field before the first field; then, as shown in step 5 3 0, record the sequence of the bitmap field difference, this comap field The difference sequence is recorded after the isotopic field difference of the first field is generated, and the isotopic field difference sequence is composed of the isotopic field difference including the first field and a plurality of isotopic field differences calculated forward; more Further, as shown in step 540, the above 3: 2 pull-down detection is performed to generate a 3: 2 pull-down detection result, and the 3: 2 pull-down detection result is generated according to the field map sequence of the same bitmap ; After performing the 3: 2 pull-down debt test, as shown in step 5 50, deinterleave the field sequence to generate a frame sequence. This frame sequence is based on the 3: 2 pull-down detection result to a field. The de-interlacing method is generated; finally, as shown in step 560, the frame sequence is output, and the frame sequence is output by a frame data stream. The above-mentioned co-located field difference may be the difference between the first field and the previous co-located field. When the field difference of this co-located image is less than a first threshold value, it can be considered that this first field is the same as its co-located field. In addition, the same field difference can also be stored in a similar manner. "Yes" (true) means less than the first threshold value. Conversely, "No" (f a 1 s e) means not less than the first threshold value. According to this, when the 3: 2 pull-down detection result is that the 3: 2 pull-down has not been detected, the general field de-interlacing method is used, for example, the field internal difference method, the field internal difference method, and adaptability. De-interlacing. When the 3: 2 pull-down detection result is that the 3: 2 pull-down is detected, a 3: 2 pull-down reduction method is used to generate a map sequence. For example, each bitmap field difference is generated according to this first field and a second field.
第23頁 1224925 五、發明說明（18) 第二圖場為第一圖場之前同位圖場。此3 : 2拉下還原法係 將產生每一該起始圖場差之第一圖場和第二圖場及此兩圖 場間之一第三圖場以第一圖場與第三圖場或以第二圖場與 第三圖場以交錯掃瞄線的方式來產生一第一圖幅、一第二 圖幅與一第三圖幅，再將一第四圖場與一第五圖場以此第 四圖場與此第五圖場以交錯掃瞄線的方式來產生一第四圖 幅與一第五圖幅，此第四圖場與第五圖場分別為第一圖場 之下一圖場與下同位圖場。 參考第六圖，圖幅6 1、6 2、6 3、6 4為一圖幅資料流， 圖場61卜612、61 3(由圖幅61而來）、圖場621、6 22 (由圖 幅6 2而來）、圖場6 3 1、6 3 2、6 3 3 (由圖幅6 3而來）與圖場 6 4 1、6 4 2 (由圖幅6 4而來）係以3 : 2拉下產生。在這情形下 ，圖場6 11與圖場6 1 3的圖場差和圖場6 3 1與圖場6 3 3的圖場 差必然為零，可知同位圖場差每差五次就會為零，換句話 說，如果同位圖場差每差五次就為零的話，每一個圖場就 能夠被還原，也就是同位圖場差為零的兩個圖場與在兩者 間的圖場係由相同的圖幅產生，而接下來的兩個圖場係由 另一個圖幅產生。所以圖場6 11、6 1 2、6 1 3可以將圖場6 1 2 與6 1 3組合來產生三個相同的圖幅，且圖場6 2 1與6 2 2可以 組合圖場6 2 1與6 2 2來產生兩個相同的圖幅，其餘圖場依此 類推。 本發明之第五較佳實施例係一 3 ·· 2拉下解交錯系統。Page 23 1224925 V. Description of the invention (18) The second field is the same position field before the first field. The 3: 2 pull-down reduction method will generate a first field and a second field for each of the initial field differences and a third field between the first field and the third field Field or the second field and the third field to generate a first frame, a second frame and a third frame by staggered scanning lines, and then a fourth field and a fifth field The picture field uses this fourth picture field and the fifth picture field to generate a fourth picture frame and a fifth picture frame in a staggered scanning line manner. The fourth picture field and the fifth picture field are the first picture, respectively. The field below the field is the same as the field below. Referring to the sixth picture, picture frames 6 1, 6, 2, 6, 3, and 6 are a picture frame data stream, picture fields 61, 612, 61 3 (from picture frame 61), picture fields 621, 6 22 (from Frame 6 2), field 6 3 1, 6, 3 2, 6 3 3 (from frame 63) and field 6 4 1, 6 4 2 (from frame 64) Pull down at 3: 2 to produce. In this case, the field difference between the field 6 11 and the field 6 1 3 and the field difference between the field 6 3 1 and the field 6 3 3 must be zero. Is zero, in other words, if the co-located field difference is zero every five times, each field can be restored, that is, the two fields with the co-located field field zero and the map between them The field is generated from the same frame, while the next two fields are generated from the other frame. So field 6 11, 6, 1 2, 6 1 3 can combine fields 6 1 2 and 6 1 3 to produce three identical frames, and fields 6 2 1 and 6 2 2 can be combined to field 6 2 1 and 6 2 2 to generate two identical picture frames, and the rest of the picture fields can be deduced by analogy. The fifth preferred embodiment of the present invention is a 3 ·· 2 pull-down de-interlacing system.
第24頁 1224925 五、發明說明（19) 如第七圖所示，此3 : 2拉下解交錯系統包含一第一儲存裝 置7 3、一第二儲存裝置7 2、一圖場資料流7卜一 3 .· 2拉下 偵測裝置74、一 3 : 2拉下還原裝置75、一圖場解交錯裝置 7 6與一圖場輸出裝置7 7。第一儲存裝置7 3係用以儲存並輸 出一第一圖場7 3 0，此第一圖場7 3 0係由一第二儲存裝置7 2 輸入，並且經由一段時間後輸出。另外，第二儲存裝置7 2 係用以儲存並輸出一第二圖場7 2 0，此第二圖場7 2 0係由一 圖場資料流7 1，並且經由該段時間後輸出。此圖場資料流 7 1係用以輸入一第三圖場7 1 0，其中，第一圖場7 3 0、第二 圖場7 2 0與第三圖場7 3 0係一圖場序列中依序由先而後被接 收之三圖場。此外，圖場序列之每一圖場與其同位圖場係 在不同時間下，以相同複數個像素位置之一像素所構成， 並且，這些像素位置構成複數條掃瞄線。3 : 2拉下偵測裝 置7 4係根據第一圖場與第三圖場產生上述之3 : 2同位圖場 差序列，並且以上述之3 : 2拉下偵測產生上述之3 : 2拉下偵 測結果742與該第三圖場之一起始訊號744。接下來，3 : 2 拉下還原裝置依第三圖場之起始訊號與第一圖場、第二圖 場與第三圖場以上述之3: 2拉下還原產生一 3: 2拉下圖幅序 列。另外，圖場解交錯裝置係依第一圖場、第二圖場與第 三圖場以一圖場解交錯法產生一圖幅序列。最後，由圖場 輸出裝置，依3 : 2拉下結果於3 : 2拉下圖幅序列該圖幅序列 擇一輸出。其中，該第三圖場之起始信號係於3 ·· 2拉下偵 測發現第三圖場之同位圖場差屬於上述之起始圖場差時送 予3 : 2拉下還原裝置，使得3 : 2拉下還原裝置能夠判別第三Page 24 1224925 V. Description of the invention (19) As shown in the seventh figure, this 3: 2 pull-down de-interlacing system includes a first storage device 7 3, a second storage device 7 2, a field data stream 7 [1] 3. 2 pull down the detection device 74, 3: 2 pull down the reduction device 75, a field de-interlacing device 76 and a field output device 7 7. The first storage device 73 is used to store and output a first field 730. The first field 730 is input from a second storage device 72 and output after a period of time. In addition, the second storage device 7 2 is used to store and output a second field 7 2 0. The second field 7 2 0 is a field data stream 7 1 and is output after a period of time. This field data stream 7 1 is used to input a third field 7 1 0, wherein the first field 7 3 0, the second field 7 2 0 and the third field 7 3 0 are a field sequence The three fields that have been received in turn have been received in order. In addition, each field of the field sequence and its co-located field are composed of one pixel at the same plurality of pixel positions at different times, and these pixel positions constitute a plurality of scanning lines. The 3: 2 pull-down detection device 7 4 generates the above-mentioned 3: 2 parity field difference sequence according to the first field and the third field, and the above-mentioned 3: 2 pull-down detection generates the above-mentioned 3: 2 The detection result 742 and the start signal 744 of one of the third fields are pulled down. Next, the 3: 2 pull-down reduction device generates a 3: 2 pull-down according to the initial signal of the third field and the first field, the second field, and the third field. View frame sequence. In addition, the field de-interlacing device generates a frame sequence by a field de-interlacing method according to the first field, the second field, and the third field. Finally, the picture field output device pulls the result according to 3: 2 and draws the picture sequence at 3: 2. The picture sequence is selected and output. Wherein, the start signal of the third picture field is sent to the 3: 2 pull-down reduction device when it is detected that the field difference of the third image field is equal to the above-mentioned initial picture field difference. Enables 3: 2 pull-down reduction device to identify third
第25頁 1224925 五、發明說明（20) 圖場於圖場序列中的位置。 據此，本發明首先提出一種四圖 · 錯系統，第八圖為本發明之第六較 =像的調適性解交 括-第-储存裝置84、4!儲==例之示意圖，包 置82、—圖場資料⑽、_動像補償裝儲存裝 内差裝置87、一像素切換裝置88與一圖86认一邊緣方向 第一儲存裝置84係用以儲存並輸出一—刀、裝置89。此 一圖場840係由一第二儲存裝置83輸入，—圖場840,此第 間後輸出。另夕卜，第二健存裝置83係”且赵由一段時 圖場830,此第二圖場83〇係由一 j儲存並輸出第二 且經由一段時間後輪出。再者，=f存裝置82輸入，並 存並輸出第三圖場820，此第圖彡曰―存裝置82係甩以儲 係用以輸入-，四圖場81。，第輸一出圖：二 後被接收之四圖場，其列中依序由先而 連續之四個圖場，動像偵測裝^ 條知瞄線。。根據此 產生一移動向量序列852與一^\^係以/述之動像谓測來 根據此移動向量序列852與第H ^序列854。接下來， 上述動像補償方法產生一第一一傻 由動像補償裝置8 6以 序列862係包含各像素位置之—像第素序/ 8 62’此第—像素 第一像素所組成。另外，Page 25 1224925 V. Description of the invention (20) The position of the field in the field sequence. Based on this, the present invention first proposes a four-picture error system, and the eighth picture is the sixth embodiment of the present invention. 82.—field data ⑽, _motion compensation device, storage device, internal difference device 87, a pixel switching device 88, and a picture 86 to identify an edge direction. The first storage device 84 is used to store and output a knife, device 89. . This field 840 is input by a second storage device 83—field 840, and then output after this period. In addition, the second health storage device 83 series "and Zhao Youshi's picture field 830, this second picture field 83o is stored by a j and output second and rotated out after a period of time. Furthermore, = f The storage device 82 inputs, coexists, and outputs the third field 820. This figure 彡 means that the “storage device 82 is dumped by the storage system for input-, the four fields 81.”, the first one is output: the second one is received. A four-picture field, in which four fields are successively arranged in order, and the moving image detection device is equipped with a line of sight. Based on this, a motion vector sequence 852 and a ^ \ ^ are described by Image predicate measurement is based on this motion vector sequence 852 and the H ^^ sequence 854. Next, the above-mentioned motion image compensation method generates a first-in-one motion image compensation device 86 which contains each pixel position in a sequence 862-like the first Prime order / 8 62'This first-pixel first pixel. In addition,
1224925 五、發明說明（21) 邊緣方向内差裝置8 7根據第三、^ ^ ^ ^ 法產生一第二像素序列872，努以上述之邊緣方向内差 像素位置之一第二像青 第一像素序列872係由各該 據像素決策序列854於第一\ °然後由像素切換裝置88依 8 72間切換以產生一内;：序列86 2與第二像素序列 882係*複數個内差圖場所組成。5來此内Λ圖場序列 置89輸出成-圖幅序列m， ^^ ’由圖場切換裝 幅所產生，各該圖幅係由第三圖俨田序列8 9 2係由複數個圖 單位交錯構成。 一 ^每與内差圖場以掃瞒線為 圖場所組成，各該 ，各内差掃瞄線係 成。依上述之動像 =動向量，動像偵 策序列，例如，上 移動向量所組成， 之—像素決策所級 像素切換裝置依此 ，此像素切換穿番 r，上述動ΐϊ 動向量之像】 n-像素不存在 二、、且成，各該内 内差掃瞄線係由 内差場序列•由複數個内差 由複數個該像素位置之一 =、·且成 偵測可產生第三圖場# ϊ所組 測裝置據此產生移c位置之 述移動向量序列俜ά:里序列與像素決 並且上述像m各該像素位置之該 成，當此各像素位置 像素決策以第一像夸=動向量存在時， 依此像素決策以第’否則 償褒置係於第二圖c像素。 為第-像素，當移動向距上 。上述内差圖場序列2::存在時’則 差圖場係由複數條内二數個内差圖 差知目㊣線所組成，1224925 V. Description of the invention (21) Edge direction internal difference device 8 7 generates a second pixel sequence 872 according to the third and ^ ^ ^ ^ methods, and uses one of the above-mentioned edge direction internal difference pixel positions as the second image green first The pixel sequence 872 is generated by the pixel decision sequence 854 at the first \ ° and then switched by the pixel switching device 88 in accordance with 8 72 ;: sequence 86 2 and the second pixel sequence 882 are a plurality of internal difference maps Place composition. In this case, the Λ field sequence is set to 89 and output as a -frame sequence m, ^^ 'is generated by the field switching format, each of the frame is from the third field Putian sequence 8 9 2 series from a plurality of images Units are staggered. Each internal difference map field is composed of sweeping lines as the map location, and each internal difference scanning line is formed. According to the above motion image = motion vector, the motion image detection sequence is composed of, for example, the upward motion vector, which is the pixel switching device at the pixel decision level, and this pixel switching passes through r, the image of the above motion motion vector] The n-pixel does not have two, and each, and each of the internal difference scanning lines is composed of a sequence of internal differences. • From a plurality of internal differences, from one of the plurality of pixel positions. Picture field # ϊThe set of measurement devices generates a sequence of motion vectors according to the position of c. The sequence is determined by the pixel and the above image m is determined by the pixel position. When the pixel position of each pixel position is determined by the first image, When the exaggeration = motion vector exists, the pixel decision is based on the pixel of the second picture c based on the 'otherwise' decision. Is the-pixel when moving to the distance. The above internal difference map field sequence 2 :: When it exists', the difference map field is composed of two or more internal difference maps,
第27頁 五、發明說明（22) 複數個該像素位置之一内本 一 測所產生第三® y tb — 冢素所組成。依上Η, + < 像：測裝置據此產生移動：ί: 向量可用以ϊ: :上述移動向量序列係由各該=立、：素决策序列，例如 =，並且上述像素決策序列係由各像=移動向量所組 所，，且成，當此像素位置之移動^ ±置之一像素決策 依此像素決策以第一像素為=】纷’像素切換裝置 裝置依此像素決策以第：：素：：傻：則’此像素切換 像素為第一像辛^二士像素位置距上述移動向量之 存在，第料“多動向I不存在時，則該第—像素不 第九 實施例係 適性解交 裝置92。 圖場差序 起始訊號 9 1係於每 、第二圖 法產生一 係依輪出 輸出。 圖為本發明之第七較佳實際例之示意 第五較佳實際例之延伸，係—種四圖場動2 = 錯系統，更包括3: 2拉下還原裝置91與輸出切換 上述之動像偵測裝置85更包括上述之產生一同位 列與執行上述之3 ·· 2拉下偵測來產生一第三圖支， 9 5 8與一輸出決策9 5 6。另外，3 : 2拉下還原裝#置 次收到第三圖場之起始訊號時9 58，依第一'圖"場 %、弟二圖場、第四圖場以上述之3: 2拉下還原 3: 2拉下圖幅序列912。最後，由輸出切換裝$置'^ 決策9 5 6以圖幅序列892與3: 2拉下圖幅序列"912擇Page 27 V. Description of the invention (22) The third ® y tb — made up of the first and second test results in one of the pixel positions. According to the above, + < Image: The measuring device generates movement based on this: ί: The vector can be used for ϊ :: The above motion vector sequence is determined by the sequence: prime decision sequence, such as =, and the above pixel decision sequence is determined by Each image = a group of motion vectors, and when the pixel position is moved ^ ± one pixel is determined according to this pixel decision. The first pixel is set to []. The pixel switching device is set according to this pixel decision: : Prime :: Silly: Then 'this pixel switching pixel is the existence of the first image Xin ^ 2 pixel position from the above-mentioned movement vector, the material "if the multi-motion direction I does not exist, the first pixel is not the ninth embodiment. Appropriate de-interleaving device 92. The field difference sequence start signal 9 1 is generated by each and the second method, and a series is output by rotation. The figure is a schematic illustration of the seventh preferred practical example of the present invention and the fifth preferred practical example. The extension is a kind of four-picture field motion 2 = wrong system, including 3: 2 pull-down reduction device 91 and output switching. The above-mentioned moving image detection device 85 further includes the above-mentioned generation, and performs the above-mentioned 3. · 2 pull down detection to generate a third figure, 9 5 8 and a lose Make a decision 9 5 6. In addition, 3: 2 pull down the reduction device # Zhici received the start signal of the third field 9 58, according to the first 'field', the second field, the fourth field The field is reduced by 3: 2 as described above, and the 3: 2 sequence is pulled down. 912. Finally, the output switching device is set to ^. Decision 9 5 6 is used to sequence 892 and 3: 2 to pull down the sequence of pictures " 912 select
1224925 五、發明說明（23) 其中，第三圖場之起始信號係於3 : 2拉下偵測發現第 三圖場之同位圖場差屬於上述之起始圖場差時送予3: 2拉 下還原裝置，使得3 : 2拉下還原裝置能夠判別第三圖場於 圖場序列中的位置。 3 : 2拉下還原裝置係於每次收到一第三圖場之起始訊 號時，依上述連續之四圖場以上述之3 : 2拉下還原法產生 一 3 ·. 2拉下圖幅序列。然後輸出切換裝置依一輸出決策於 圖幅序列與該3 : 2拉下圖幅序列擇一輸出。 當第三圖場之同位圖場差為起始圖場差之一時，動像 偵測裝置產生此第三圖場起始訊號。上述3 : 2拉下還原裝 置係於收到該第三圖場起始訊號時，以3 : 2拉下還原法產 生3 : 2拉下圖幅序列。 另外，輸出切換裝置依該輸出決策輸出。例如，當起 始圖場差之總數超過一第三門檻限值時，輸出切換裝置依 輸出決策選擇以3 : 2拉下圖幅序列輸出，否則輸出切換裝 置依輸出決策選擇以圖幅序列輸出。 以上所述僅為本發明之較佳實施例而已，並非用以限 定本發明之申請專利範圍；凡其他為脫離本發明所揭示之 精神下所完成之等效改變或修飾，均應包含在下述之申請 專利範圍。1224925 V. Description of the invention (23) Among them, the start signal of the third field is detected at 3: 2, and the co-located field difference of the third field is found to be 3: The 2 pull-down reduction device enables the 3: 2 pull-down reduction device to determine the position of the third field in the field sequence. The 3: 2 pull-down reduction device generates a 3 ·· 2 pull-down picture each time when a start signal of a third field is received, according to the above-mentioned four consecutive fields using the above 3: 2 pull-down reduction method. Frame sequence. Then, the output switching device selects one output from the frame sequence and the 3: 2 pull-down frame sequence according to an output decision. When the co-located field difference of the third field is one of the initial field fields, the moving image detection device generates the third field start signal. The above 3: 2 pull-down reduction device generates a 3: 2 pull-down image sequence using the 3: 2 pull-down reduction method when receiving the start signal of the third field. In addition, the output switching device outputs according to the output decision. For example, when the total number of initial picture field differences exceeds a third threshold value, the output switching device pulls down the picture frame sequence by 3: 2 according to the output decision choice, otherwise the output switching device chooses to output the picture frame sequence according to the output decision . The above is only a preferred embodiment of the present invention, and is not intended to limit the scope of patent application of the present invention; all other equivalent changes or modifications made without departing from the spirit disclosed by the present invention shall be included in the following The scope of patent application.
第29頁 1224925 圖式簡單說明 五、【圖式簡單說明】 本發明相對於先前技藝之優點與好處在於參考下列圖 示與具體實施例比較後將更容易顯現，其中： 第一 A圖至第一 C圖為圖場解交錯法之示意圖； 第二A圖至第二D圖為本發明之第一較佳實施例之示意 圖， 第三圖為本發明之第二較佳實施例之示意圖； 第四圖為本發明之第三較佳實施例之示意圖； 第五圖與第六圖為本發明之第四較佳實施例之示意圖 第七圖為本發明之第五較佳實施例之示意圖； 第八圖為本發明之第六較佳實施例之示意圖； 第九圖為本發明之第七較佳實施例之示意圖。 主要部分之代表符號： . 25第一圖場1224925 Brief description of the drawings on page 29. [Simplified description of the drawings] The advantages and benefits of the present invention over the prior art are that it will be easier to see after comparison with the specific embodiments with reference to the following drawings, where: Figure C is a schematic diagram of the field de-interlacing method; Figures A through D are schematic diagrams of the first preferred embodiment of the present invention, and Figure 3 is a schematic diagram of the second preferred embodiment of the present invention; The fourth diagram is a schematic diagram of the third preferred embodiment of the present invention; the fifth and sixth diagrams are schematic diagrams of the fourth preferred embodiment of the present invention; the seventh diagram is the schematic diagram of the fifth preferred embodiment of the present invention; Figure 8 is a schematic diagram of a sixth preferred embodiment of the present invention; Figure 9 is a schematic diagram of a seventh preferred embodiment of the present invention. The main part of the symbol:. 25 first field
第30頁 1224925 圖式簡單說明 2 5 0第一區塊 2 5 2移動向量 2 6第二圖場 2 6 1第二區塊 27第三圖場 270第三區塊 28第四圖場 2 8 1第四區塊 29圖幅 Φ Φ 2 9 1 0圖幅區塊 4 2上方原掃瞄線 4 2 1上方原掃瞄線之像素 4 2 2上方原掃瞄線之像素 4 2 3上方原掃瞄線之像素 4 2 4上方原掃瞄線之像素 425上方原掃瞄線之像素 4 4新線 443新像素 4 6下方原掃瞄線 4 6 1下方原掃瞄線之像素 4 6 2下方原掃瞄線之像素 4 6 3下方原掃瞄線之像素 4 6 4下方原掃瞄線之像素 4 6 5下方原掃瞄線之像素Page 24 1224925 Schematic description 2 5 0 First block 2 5 2 Motion vector 2 6 Second field 2 6 1 Second block 27 Third field 270 Third block 28 Fourth field 2 8 1Fourth block 29 picture frame Φ Φ 2 9 1 0 picture block block 4 2 above the original scan line 4 2 1 pixels above the original scan line 4 2 2 pixels above the original scan line 4 2 3 above the original Pixels of the scanning line 4 2 4 Pixels of the original scanning line above 425 Pixels of the original scanning line 4 4 New line 443 New pixels 4 6 Original scanning line below 4 6 1 Pixels of the original scanning line 4 6 2 Pixels under the original scan line 4 6 3 Pixels under the original scan line 4 6 4 Pixels under the original scan line 4 6 5 Pixels under the original scan line
第-3l·、頁 1224925 圖式簡單說明 61圖幅 6 1 1圖場 6 1 2圖場 6 1 3圖場 62圖幅 6 2 1圖場 62 2圖場 63圖幅 631圖場 6 3 2圖場 633圖場 64圖幅 641圖場 642圖場 7 1圖場資料流 72第二儲存裝置 73第一儲存裝置 74 3 : 2拉下偵測裝置 7 5 3 : 2拉下還原裝置 76圖場解交錯裝置 7 7圖場輸出裝置 8 1圖場資料流 8 1 0第四圖場 8 2第三儲存裝置Section-3l ·, Page 1224925 Schematic Illustration 61 Frame 6 1 1 Frame 6 1 2 Frame 6 1 3 Frame 62 Frame 6 2 1 Frame 62 2 Frame 63 Frame 631 Frame 6 3 2 Field 633 Field 64 Field 641 Field 642 Field 7 1 Field data stream 72 Second storage device 73 First storage device 74 3: 2 pull down detection device 7 5 3: 2 pull down reduction device 76 Figure Field de-interlacing device 7 7 Field output device 8 1 Field data stream 8 1 0 Fourth field 8 2 Third storage device
第32頁 1224925 圖式簡單說明 〇 82 0第三圖場 83第二儲存裝置 830第二圖場 84第一儲存裝置 840第一圖場 8 5動像偵測器 85 2移動向量序列 8 5 4決策序列 8 6動像補償裝置 862第一像素序列 87邊緣方向内差裝置 8 7 2第二像素序列 8 8圖場切換裝置 8 8 2内差圖場序列 8 9圖幅切換裝置 8 9 2圖幅序列 9 1 3 ·. 2拉下還原裝置 9 1 2 3 : 2拉下圖幅 9 5 6輸出決策 9 5 8起始訊號1224925 Illustration on page 32 Brief description 082 0 Third field 83 Second storage device 830 Second field 84 First storage device 840 First field 8 5 Motion detector 85 2 Motion vector sequence 8 5 4 Decision sequence 8 6 Motion image compensation device 862 First pixel sequence 87 Edge-difference device 8 7 2 Second pixel sequence 8 8 Field switching device 8 8 2 Internal-difference field sequence 8 9 Frame switching device 8 9 2 Frame sequence 9 1 3 ·. 2 Pull down the reduction device 9 1 2 3: 2 Pull down the picture frame 9 5 6 Output decision 9 5 8 Start signal
第33頁Page 33
Claims (1)
1224925 六、申請專利範圍 1. 一種四圖場動像的調適性解交錯方法，包括： 接收一圖場序列中之一四圖場，該四圖場係一第一圖 場、一第二圖場、一第三圖場與一第四圖場且依序由先而 後被接收，其中，該圖場序列之每一圖場與其同位圖場係 在以不同時間下，以相同複數個像素位置之一像素所構成 並且，該些像素位置構成複數條掃瞄線； 執行一動像偵測，係用以產生一動像偵測結果； 產生一同位圖場差序列，該同位圖場差序列係由該動 像偵測結果所產生； 執行一 3 : 2拉下偵測，該3 : 2拉下偵測係依該同位圖場 差序列產生一 3 : 2拉下偵測結果；及 產生一圖場解交錯結果，該圖場解交錯結果係依該3 : 2拉下結果以一圖場解交錯法產生。 2. 如申請專利範圍第1項所述四圖場動像的調適性解交錯 方法，其中上述動像偵測，包括： 劃分該第三圖場成複數個原區塊，各該原區塊係由複 數個該像素所組成，各該像素係各別對應到該些像素位置 之一； 偵測各該原區塊之一第一區塊，當該第一圖場中該原 區塊之相對位置以一搜尋範圍内存在至少一個相似區塊， 則該第一區塊存在且為最接近該原區塊之相對位置之該相 似區塊，否則，該第一區塊不存在，其中，該相似區塊係 與該原區塊之一第一區塊差小於一第一門檻限值；1224925 6. Scope of patent application 1. A method for adaptive de-interlacing of a four-picture field including: receiving one of the four-picture fields in a sequence of one field, the four-picture field being a first picture field and a second picture The field, a third field, and a fourth field are sequentially received, where each field in the field sequence and its co-located field are at different times and at the same plurality of pixel positions. A pixel is formed and the pixel positions constitute a plurality of scanning lines; a moving image detection is performed to generate a moving image detection result; a bitmap field difference sequence is generated, and the bitmap field difference sequence is composed of The moving image detection result is generated; a 3: 2 pull-down detection is performed, and the 3: 2 pull-down detection generates a 3: 2 pull-down detection result according to the isotopic field difference sequence; and a picture is generated Field de-interlacing results. The field de-interlacing results of the picture are generated by a field de-interlacing method according to the 3: 2 pull-down result. 2. The adaptive de-interlacing method for a four-picture field moving image as described in item 1 of the scope of patent application, wherein the moving image detection includes: dividing the third picture field into a plurality of original blocks, each of the original blocks Is composed of a plurality of the pixels, and each of the pixels corresponds to one of the pixel positions respectively; a first block of each of the original blocks is detected, and when the original block in the first field is The relative position has at least one similar block within a search range, then the first block exists and is the similar block closest to the relative position of the original block; otherwise, the first block does not exist, where, The difference between the similar block and a first block of the original block is less than a first threshold value; 第34頁 六、申請專利範圍 計 ’則該 相對位 偵 二圖場 央，當 第二區 偵 四圖場 區塊與 該第三 計 該移 塊不存 所對應 算各該原區 第一向量係 置，否則， 測各該原區 中該第一區 一第三區塊 塊不存在； 測各該原區 中距該第二 δ亥第二區塊 區塊存在， 鼻各該原區 動向量係從 在，同時， 到之各該像 塊之一第一向量，者篦 _ 從兮楚 广t 田4弟一區塊存在時 攸※第一區塊之相 1 哕笛 A曰 相對位置至該原區塊之 ”亥第一向量不存在； $之一第二區塊，該第二區於 存. ，…亥原£塊之相對位置中 存在時，則該第二區塊存在，否則，^ 塊之該第三區堍，兮赞一 區揷j g &兒该弟二區塊位於該第 Γ :弟1量之相對位置，當該第： 之-弟二區塊差小於一第一門播限：- 否則，該第三區塊不存在；及 ， 塊之該移動向量’當該第二區塊存在 δ亥第二區塊至”亥原區塊，否則，兮原 各該原區塊之該移動向量亦為該原^ ^ 素位置的該移動向量。 μ塊 •、如申請專利範圍第2項所述四圖場動像的調適性解交 一法’其中上述第一區塊差係該原區塊與該第一區塊$， 上相關第一像素對之差值的總和，該相關第一像素對之: e亥像素分別屬於該原區塊與該第一區塊，並且該第二區土 差係該第二區塊與該第一區塊中每一相關第二像素對=鬼 值的總和，該相關第二像素對之各該像素分別屬於該差 區塊與該第三區塊。 w —Page 34 VI. The scope of the patent application is' the second detection field center of the relative detection. When the fourth detection field block of the second area and the third measurement block are not stored, the first vector of each original area is calculated. System, otherwise, the first block and the third block in the original region are not detected; the existence of the second block from the second delta block in the original region is detected; The quantity is from the same time, at the same time, to the first vector of each of the image blocks, which is from the beginning of the first block. Phase 1 of the first block. Until the first block of the original block, the first vector does not exist; if one of the second blocks exists, the second block exists.... The second block exists when the relative position of the original block exists. Otherwise, ^ block of the third block 兮, Xizan one block 揷 jg & Er Gedi block 2 is located in the relative position of the Γ: yi 1 amount, when the ：--xi block difference is less than one First threshold:-Otherwise, the third block does not exist; and, the motion vector of the block 'when the second block exists, the second block is δ. "Hai original block, otherwise the original motion vectors Xi of each of the original blocks of the original ^ ^ also for the motion vectors prime location. μ block •, as described in item 2 of the scope of the patent application, a method of adaptive dissection of the four-field motion image 'where the first block is the difference between the original block and the first block $, The sum of the difference between the pixel pairs, the related first pixel pair: e Hai pixels belong to the original block and the first block, respectively, and the second region soil difference is the second block and the first region Each related second pixel pair in the block = the sum of ghost values, and each of the pixels of the related second pixel pair belongs to the difference block and the third block, respectively. w — 1224925 六、申請專利範圍 4. 如申請專利範圍第3項所述四圖場動像的調適性解交錯 方法，其中上述像素差係亮度差。 5. 如申請專利範圍第1項所述四圖場動像的調適性解交錯 方法，其中上述同位圖場差序列之產生方法，包含： 產生該第三圖場之一同位圖場差，該同位圖場差係根 據該動像偵測產生；及 記錄該同位圖場差序列，該同位圖場差序列係於該第 三圖場之該同位圖場差產生後記錄，該同位圖場差序列係 由包含該第三圖場之該同位圖場差及向前推算之複數個該 同位圖場差所組成。 6. 如申請專利範圍第1項所述四圖場動像的調適性.解交錯 方法，其中上述3 : 2拉下偵測係於該同位圖場差序列中搜 尋複數個起始圖場差，該些起始圖場差皆小於一第二門檻 限值且每隔五次出現一次，當該起始圖場差之總數超過一 第三門檻限值，則該3: 2拉下偵測結果為已偵測出3 : 2拉下 ，否則，該3 : 2拉下偵測結果為尚未偵測出3 ·. 2拉下。 7. 如申請專利範圍第6項所述四圖場動像的調適性解交錯 方法，其中上述動像偵測結果係一動像區域，該動像區域 係所有該移動向量不存在之該原區塊所對應到的所有該像 素位置之聯集，並且該動像區域外之所有其他該像素位置 之聯集為一靜像區域。1224925 6. Scope of patent application 4. As described in item 3 of the scope of patent application, the adaptive de-interlacing method of the four-picture field dynamic image is used, wherein the pixel difference is the brightness difference. 5. The adaptive de-interlacing method for a four-picture field moving image as described in item 1 of the scope of the patent application, wherein the method for generating the isotopic field difference sequence includes: generating one isotopic field difference of the third field, and The isotopic field difference is generated according to the moving image detection; and the isotopic field difference sequence is recorded, and the isotopic field difference sequence is recorded after the isotopic field difference of the third field is generated, and the isotopic field difference is recorded. The sequence is composed of the isotopic field difference including the third field and a plurality of the isotopic field differences calculated forward. 6. The adaptability of the four-picture field motion image as described in item 1 of the scope of the patent application. The de-interlacing method, in which the above 3: 2 pull-down detection is based on searching the plurality of initial picture field differences in the same-image field difference sequence The initial picture field differences are less than a second threshold value and occur every five times. When the total of the initial picture field differences exceeds a third threshold value, the 3: 2 pull-down detection The result is that the 3: 2 pulldown has been detected, otherwise, the result of the 3: 2 pulldown detection is that the 3 · .2 pulldown has not been detected. 7. The adaptive de-interlacing method for a four-picture field moving image as described in item 6 of the scope of the patent application, wherein the above-mentioned moving image detection result is a moving image area, and the moving image area is all the original areas where the motion vector does not exist. The associated set of all the pixel positions corresponding to the block, and the associated set of all other pixel positions outside the moving image area is a still image area. 第36頁 、申請專利範圍 方法，當3和範圍第7項所述四圖場動像的調適性解交錯 |解交錯結果发偵'則結果為尚未偵測出3 ··2拉下時，則該圖場 |所產生之一二由該第三圖場與一内差圖場之各掃瞄線交錯 該像素位置幅，該内差圖場具有與該第二圖場相同之各 而所有該内矣並且以各該像素位置之一内差像素所組成， I解交錯該動推像素係分別以圖場内内差法與圖場間内差法 I 像區域與該靜像區域所產生。 I方法，其利範圍第8項所述四圖場動像的調適性解交錯 像素係“ 一動=】：2差法係將該靜像區域中各該内差 場上之一第一你主貝/產生，該動像補償法係以該第二圖 至該内差像夸m ^產生，從該第一像素所在之該像素位置 位置的該像素位置為該内差像素所對應之像素On page 36, the method of applying for the patent scope, when the adaptability of the four-picture field dynamic image described in 3 and 7 of the scope is de-interlaced | the de-interlaced result is detected, and the result is that 3 ·· 2 has not been detected. Then the field | field produced by the scanning field is interlaced by the scanning lines of the third field and an internal difference field, and the pixel position width is the same as that of the second field. The internal error is composed of an internal difference pixel at each of the pixel positions, and the de-interlacing of the dynamic push pixels is generated by the image field internal difference method and the field internal difference method I image area and the still image area, respectively. Method I, the adaptability of the four-picture field dynamic image described in item 8 of the de-interlacing pixel system "One move =]: 2 difference method. One of the internal difference fields in the static image area. The motion image compensation method is generated from the second image to the internal difference image, and the pixel position from the pixel position of the first pixel is the pixel corresponding to the internal difference pixel. 產 一動像Produce a moving image 之Of 生各 一最 專利範圍第8項所述四圖場 中沭闰M舒動像的調適性解交 中上达圖％間内I法係^ —邊緣方向内差法， 内ί = 該掃I線中屬於該動像區域之每 ί第亥圖：中’該内差圖場之該掃晦 線與一第二掃 ；“於 該動像像素，該動像像素m 相似像素斜之平均值來：：於：，數個像素對 座生，该攻相似像素對The adaptive resolution of the 沭 闰 M-shuttle image in the four-picture field described in item 8 of each patent, and the up-and-down map of the interstitial I method ^ —the edge direction internal difference method, where = = the scan I Each line in the line that belongs to the moving image area: Middle 'The scan line of the internal difference field and a second scan; "for the moving image pixel, the moving image pixel m is the average of the similar pixel slopes Here :: Yu :, several pixel pairs are born, this attack is similar to pixel pairs 第37頁 1224925 六、申請專利範圍 為該些像素對中具有最小差值之像素對；及 取得各該動像像素相關之複數個像素對，係以該動像 像素為中心由複數個不同方向於該第一掃瞄線與該第二掃 瞒線各取一像素產生。 11.如申請專利範圍第1 0項所述四圖場動像的調適性解交 錯方法，其中上述複數個像素對係最接近該動像像素之五 對像素對，該些像素對之方向分別為垂直方向、左上到右 下方向、右上至左下方向、左左上至右右下方向與右右上 至左左下方向。 1 2.如申請專利範圍第6項所述四圖場動像的調適性解交 錯方法，當3 : 2偵測結果為已偵測出3 .· 2拉下時，則該圖場 解交錯結果係以一 3 : 2拉下還原法產生，該3 : 2拉下還原法 係當該第三圖場之同位圖場差為該起始圖場差時，以該第 一圖場與該第二圖場以交錯掃瞄線的方式來產生該第一圖 場、該第二圖場與該第三圖場之一第一圖幅、一第二圖幅 與一第三圖幅，再將該第四圖場與一第五圖場以該第四圖 場與該第五圖場以交錯掃瞄線的方式來產生該第四圖場與 該第五圖場之一第四圖幅與一第五圖幅，該第五圖場係於 該圖場序列中該第四圖場之下一圖場。 1 3. —種四圖場動像的調適性解交錯系統，包括： 一第一儲存裝置，用以儲存並輸出一第一圖場，該第Page 37 1224925 6. The scope of the patent application is the pixel pair with the smallest difference among the pixel pairs; and obtaining a plurality of pixel pairs related to each of the moving image pixels is based on the moving image pixel as a center from a plurality of different directions One pixel is taken from each of the first scanning line and the second scanning line. 11. The adaptive de-interlacing method for a four-picture field moving image as described in item 10 of the scope of the patent application, wherein the plurality of pixel pairs are five pixel pairs closest to the moving image pixel, and the directions of the pixel pairs are respectively It is vertical, top left to bottom right, top right to bottom left, left top to right bottom right, and top right to bottom left left. 1 2. According to the adaptive de-interlacing method of the four-picture field moving image as described in item 6 of the scope of the patent application, when the 3: 2 detection result is that 3. · 2 has been detected, the field is de-interlaced The result is generated by a 3: 2 pull-down reduction method, and when the difference between the isotopic field of the third field is the initial field difference, the first field and the The second picture field generates the first picture field, one of the second picture field and one of the third picture field, a second picture frame and a third picture frame by means of staggered scanning lines, and then The fourth field and a fifth field are generated by the fourth field and the fifth field in a staggered scanning line manner to generate a fourth frame of the fourth field and the fifth field. With a fifth frame, the fifth field is a field below the fourth field in the field sequence. 1 3. An adaptive de-interlacing system for a four-image field including: a first storage device for storing and outputting a first image field; 第38頁 1224925 六、申請專利範圍 一圖場係由一第二儲存裝置輸入，並且經由一段時間後輸 出； 該第二儲存裝置，用以儲存並輸出一第二圖場，該第 二圖場係由一第三儲存裝置輸入，並且經由該段時間後輸 出； 該第三儲存裝置，用以儲存並輸出一第三圖場，該第 三圖場係由一圖場資料流輸入，並且經由該段時間後輸出 一圖場貧料流，用以輸入'一^弟四圖場，該弟^一圖場、 該第二圖場、該第三圖場與該第四圖場係一圖場序列中依 序由先而後被接收之四圖場，其中，該圖場序列之每一圖 場與其同位圖場係在不同時間下，以相同複數個像素位置 之一像素所構成，並且，該些像素位置構成複數條掃猫線 y 一動像憤測裝置，用以產生一移動向量序列、一像素 決策序列、一第三圖場起始訊號與一輸出決策，該移動向 量序列、該像素決策序列、該第三圖場起始訊號與該輸出 決策係根據該四圖場產生； 一動像補償裝置，係根據該移動向量序列與該第二圖 場產生一第一像素序列，該第一像素序列係由各該像素位 置之一第一像素所組成； 一邊緣方向内差裝置，係根據該第三圖場產生一第二 像素序列，該第二像素序列係由各該像素位置之一第二像 素所組成；Page 24 1224925 VI. Patent application scope A picture field is input by a second storage device and output after a period of time; the second storage device is used to store and output a second picture field, the second picture field It is input by a third storage device and output after a period of time; the third storage device is used to store and output a third field, which is input from a field data stream, and After this period of time, a picture field lean stream is output for inputting a picture field, the picture field, the second picture field, the third picture field, and the fourth picture field. The field sequence consists of four fields that have been received one after another. Among them, each field of the field sequence and its co-located field are composed of one pixel at the same plurality of pixel positions at different times, and, These pixel positions constitute a plurality of cat-line scan lines and a moving image detection device for generating a motion vector sequence, a pixel decision sequence, a third field start signal and an output decision. The motion vector sequence, the pixel Decision sequence, the first The field start signal and the output decision are generated based on the four fields; a moving image compensation device generates a first pixel sequence according to the motion vector sequence and the second field, and the first pixel sequence is generated by each of the One of the pixel positions is composed of a first pixel; an edge direction internal difference device generates a second pixel sequence according to the third field; the second pixel sequence is composed of one second pixel at each of the pixel positions; 第39頁 1224925 六、申請專利範圍 一像素切換裝置，用以依該像素決策序列於該第一像 素序列與該第二像素序列間切換以產生一内差圖場序列， 該内差圖場序列係由複數個内差圖場所組成； 一圖場切換裝置，用以輸出成一圖幅序列，該圖幅序 列係由複數個圖幅所產生，各該圖幅係由該第三圖場與該 内差圖場以掃瞄線為單位交錯構成； 一 3 : 2拉下還原裝置，該3 : 2拉下還原裝置係於每次收 到該第三圖場之起始訊號時，依該四圖場以3 : 2拉下還原 法產生一 3 : 2拉下圖幅序列；及 一輸出切換裝置’該輸出切換裝置係依該輸出決策以 該圖幅序列與該3 : 2拉下圖幅序列擇一輸出。 1 4.如申請專利範圍第1 3項所述四圖場動像的調適性解交 錯系統，其中上述動像偵測裝置係於該第一圖場接收完成 時執行一動像摘測、產生一同位圖場差序列與執行一 3 : 2 拉下偵測。 1 5.如申請專利範圍第1 4項所述四圖場動像的調適性解交 錯系統，其中上述動像偵測，包括： 劃分該第三圖場成複數個原區塊，各該原區塊係由複 數個該像素所組成，各該像素係各別對應到該些像素位置 之一； 偵測各該原區塊之一第一區塊，當該第一圖場中該原 區塊之相對位置以一搜尋範圍内存在至少一個相似區塊，Page 1224925 6. Patent application scope A pixel switching device for switching between the first pixel sequence and the second pixel sequence according to the pixel decision sequence to generate an internal difference field sequence, the internal difference field sequence Is composed of a plurality of internal difference map places; a field switching device is used to output a sequence of picture frames, the picture sequence is generated by a plurality of picture frames, and each of the picture frames is composed of the third picture field and the The internal difference field is staggered with the scanning line as a unit; a 3: 2 pull-down reduction device, the 3: 2 pull-down reduction device is based on the four times each time it receives the start signal of the third field The picture field uses a 3: 2 pull-down reduction method to generate a 3: 2 pull-down picture sequence; and an output switching device 'the output switching device is based on the output decision to use the picture sequence and the 3: 2 pull-down picture frame Sequence alternative output. 1 4. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 13 of the scope of the patent application, the above-mentioned moving image detection device performs a moving image extraction test and generates together when the first field is received. Bitmap field difference sequence and perform a 3: 2 pull-down detection. 1 5. The adaptive de-interlacing system for a four-picture field moving image as described in item 14 of the scope of the patent application, wherein the moving image detection includes: dividing the third picture field into a plurality of original blocks, each of which The block is composed of a plurality of the pixels, and each of the pixels corresponds to one of the pixel positions respectively; a first block of each of the original blocks is detected, and when the original region in the first field is detected The relative position of the blocks has at least one similar block within a search range, 第40頁 ^^5 …申請專利範圍 f該第-區塊 =區塊，否則 /、該原區塊之 計算各該 則該第一向 f十位置，否 偵測各該 圖場中該第 > pj> § 一第三 二區塊不存 偵測各該 =圖場中距該 區塊與該第二 該第一广 禾二區塊存 計算各該 議移動向量 塊不存在，同 所對應到之各 存在且為最接近該原區塊之相對位置之該相 該第一區堍不存在，其中，該相似區塊係 相 央 一第一 原區塊 量係從 則，該 原區塊 一區塊 區塊存 在； 原區塊 第二區 區塊之在，否 原區塊 係從該 時，各 該像素 一區塊不存在’ 區塊差小於一第 之一第一向量， 該第一區塊之相 第一向量不存在 之一第二區塊’ 之相對位置與該 在時’則該第一 門松限值； 當該第一區塊存在時 對位置至該原區塊之 該第二區塊位於該第 原區塊之相對位置中 區塊存在，否則，該 該第三區塊位於該第 相對位置，當該第三 之該第三區塊’ 塊該第一向量之 一第二區塊差小於一第一門檻限值， ·如申請專利範圍 ，系統’其中上述第 母—相關第一像素對 各該像素分別屬於該 塊差係該第二區塊與 則，該第三區塊 之該移動向量， 第二區塊至該原 該原區塊之該移 位置的該移動向 第1 5項所述四圖 一區塊差係該原 之差值的總和， 原區塊與該第一 該第一區塊中每 不存在，及 當該第二區塊存在時 區塊’否則，該原區 動向量亦為該原區塊 量。 場動像的調適性解交 區塊與該第一區塊中 該相關第一像素對之 區塊，並且該第二區 一相關第二像素對之Page 40 ^^ 5… the scope of the patent application f the-block = block, otherwise /, the calculation of the original block should be the first f f ten positions, whether to detect the first in each field > pj > § Blocks 1, 32 and 2 are not detected. Each of these = blocks in the field and the second and first Guanghe 2 blocks do not exist. Corresponding to the existence of the phase and the first area of the phase closest to the relative position of the original block, the first area 堍 does not exist, wherein the similar block is the phase of the central area, the first original block amount is the rule, the original area Block 1 block exists; block 2 exists in the original block, whether the original block is from then on, each block of the pixel does not exist 'block difference is less than a first first vector, the first The relative position of a block where the first vector does not exist, the relative position of a second block, and the presence, then the first threshold value; when the first block exists, the position to the original block of the The second block exists in the relative position of the original block, otherwise, the third block is located in the relative position When the third block of the third block 'block one of the first vector the second block difference is less than a first threshold value, such as the scope of a patent application, the system' wherein the first mother-related first pixel For each of the pixels belonging to the block difference is the second block and the rule, the movement vector of the third block, and the movement of the second block to the original block's moved position to the first 5 The difference of the four figures and one block mentioned in the item is the sum of the difference between the original block, each of the original block and the first and the first block does not exist, and the block exists when the second block exists; otherwise, the original block The zone motion vector is also the original block amount. The adaptive disjoint block of the field image is the block of the relevant first pixel pair in the first block, and the relevant second pixel pair of the second area is 第41頁 1224925 六、申請專利範圍 差值的總和，該相關第二像素對之各該像素分別屬於該第 二區塊與該第三區塊。 1 7.如申請專利範圍第1 6項所述四圖場動像的調適性解交 錯系統，其中上述像素差係亮度差。 1 8.如申請專利範圍第1 4項所述四圖場動像的調適性解交 錯系統，其中上述移動向量序列係由各該像素位置之該移 動向量所組成。 1 9.如申請專利範圍第1 3項所述四圖場動像的調適性解交 錯系統，其中上述產生同位圖場差之方法，包括： 產生各該第三圖場之一同位圖場差，該同位圖場差係 該些原區塊中該移動向量不存在之總數；及 記錄該同位圖場差序列，該同位圖場差序列係於該第 三圖場之該同位圖場差產生後記錄，該同位圖場差序列係 由包含該第三圖場之該同位圖場差及向前推算之複數個該 同位圖場差所組成。 2 0 .如申請專利範圍第1 9項所述四圖場動像的調適性解交 錯系統，其中上述3 : 2拉下偵測係於該同位圖場差序列中 搜尋複數個起始圖場差，該些起始圖場差皆小於一第二門 檻限值且每隔五次出現一次。Page 41 1224925 VI. Patent Application Sum of the difference, each pixel of the relevant second pixel pair belongs to the second block and the third block respectively. 1 7. The adaptive deinterleaving system for a four-picture field moving image as described in item 16 of the scope of the patent application, wherein the above-mentioned pixel difference is a difference in brightness. 1 8. The adaptive de-interleaving system of the four-picture field moving image as described in item 14 of the scope of the patent application, wherein the above-mentioned motion vector sequence is composed of the motion vector of each pixel position. 19. The adaptive de-interlacing system for a four-picture field moving image as described in item 13 of the scope of the patent application, wherein the method for generating an isotopic field difference includes: generating an isotopic field difference for each of the third fields , The co-located field difference is the total number of the motion vectors that do not exist in the original blocks; and the co-located field difference sequence is recorded, which is generated from the co-located field difference of the third field After the recording, the sequence of the isotopic field difference is composed of the field of the isotopic field including the third field and a plurality of the field of the isotopic field calculated forward. 2 0. The adaptive de-interlacing system of the four-picture field moving image as described in item 19 of the scope of the patent application, wherein the above 3: 2 pull-down detection is to search for a plurality of starting picture fields in the sequence of the same image field difference Poor, these initial picture field differences are less than a second threshold value and occur every five times. 第-42、頁 1224925 六、申請專利範圍 2 1.如申請專利範圍第1 4項所述四圖場動像的調適性解交 . 錯系統，其中上述内差圖場序列係由複數個内差圖場所組 成，各該内差圖場係由複數條内差掃瞄線所組成，各内差 掃瞄線係由複數個該像素位置之一内差像素所組成。 2 2 .如申請專利範圍第1 5項所述四圖場動像的調適性解交 錯系統，其中上述像素決策序列係由各該像素位置之一像 素決策所組成，當該像素位置之移動向量存在時，該像素 切換裝置依該像素決策以該第一像素為該内差像素，否則 ，該像素切換裝置依該像素決策以該第二像素為該内差像 $ 素。 2 3 .如申請專利範圍第2 2項所述四圖場動像的調適性解交 錯系統，其中上述動像補償裝置係於該第二圖場以各該像 素位置距該移動向量之像素為第一像素，當該移動向量不 存在時，則該第一像素不存在。 2 4 .如申請專利範圍第1 4項所述四圖場動像的調適性解交 錯系統，其中上述邊緣方向内差裝置係依該第三圖場以邊 緣方向内差法產生一邊緣方向内差圖場，該邊緣方向内差 β 圖場係由複數條邊緣方向内差掃瞄線所組成，各該内差掃 瞄線係由複數個該第二像素所組成。 2 5 .如申請專利範圍第2 4項所述四圖場動像的調適性解交Page-42, page 1224925 6. Application for Patent Scope 2 1. The adaptive disassembling of the four-picture field dynamic image as described in item 14 of the scope of patent application. Wrong system, in which the above-mentioned internal difference field sequence is composed of a plurality of internal The internal difference image field is composed of multiple internal difference scanning lines, and each internal difference scanning line is composed of multiple internal difference pixels at one of the pixel positions. 2 2. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 15 of the scope of the patent application, wherein the above-mentioned pixel decision sequence is composed of one pixel decision of each pixel position, and the movement vector of the pixel position When present, the pixel switching device uses the first pixel as the internal difference pixel according to the pixel decision, otherwise, the pixel switching device uses the second pixel as the internal difference pixel according to the pixel decision. 2 3. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 22 of the scope of the patent application, the above-mentioned moving image compensation device is based on the second picture field, and the pixels of each pixel position away from the motion vector are First pixel, when the motion vector does not exist, the first pixel does not exist. 24. The adaptive de-interlacing system for a four-picture field moving image as described in item 14 of the scope of the patent application, wherein the edge direction internal difference device generates an edge direction internal difference according to the third field using the edge direction internal difference method. The difference image field, the edge direction internal difference β field is composed of a plurality of edge direction internal difference scanning lines, and each of the internal difference scanning lines is composed of a plurality of the second pixels. 2 5. Adaptation of four-field motion image as described in item 24 of the scope of patent application 第43頁 1224925 六、申請專利範圍 錯系統，其中上述邊緣方向内差法，包含： 產生該邊緣方向内差圖場之各該邊緣方向内差掃瞄線 ，該邊緣方向内差掃瞄線係以產生於介於該第三圖場之一 第一掃瞄線與一第二掃瞄線之該第二像素所構成； 產生各該第二像素，該第二像素係於一複數個像素對 中，以一最相似像素對之平均值來產生，該最相似像素對 為該些像素對中具有最小差值之像素對；及 取得各該第二像素相關之複數個像素對，係以該動像 像素為中心由複數個不同方向於該第一掃瞄線與該第二掃 瞄線各取一像素產生。 2 6 .如申請專利範圍第2 5項所述四圖場動像的調適性解交 錯系統，其中上述複數個像素對係最接近該動像像素之五 對像素對，該些像素對之方向分別為垂直方向、左上到右 下方向、右上至左下方向、左左上至右右下方向與右右上 至左左下方向。 2 7.如申請專利範圍第1 9項所述四圖場動像的調適性解交 錯系統，當該第三圖場之同位圖場差為該起始圖場差時， 該動像偵測裝置產生該第三圖場起始訊號。 2 8 .如申請專利範圍第1 9項所述四圖場動像的調適性解交 錯系統，當該起始圖場差之總數超過一第三門檻限值，該 輸出切換裝置依該輸出決策選擇以該3 : 2拉下圖幅序列輸Page 24 1224925 6. The patent application range error system, wherein the above-mentioned edge direction internal difference method includes: generating each edge direction internal difference scanning line of the edge direction internal difference field, and the edge direction internal difference scanning line system Is composed of the second pixel generated between a first scan line and a second scan line in the third field; each second pixel is generated, and the second pixel is in a plurality of pixel pairs In the method, an average value of a most similar pixel pair is generated, and the most similar pixel pair is a pixel pair having a smallest difference among the pixel pairs; and a plurality of pixel pairs related to each of the second pixels are obtained by using the The moving image pixel is generated by taking a pixel in each of the first scanning line and the second scanning line in different directions. 2 6. The adaptive de-interlacing system of the four-picture field moving image as described in item 25 of the scope of patent application, wherein the plurality of pixel pairs are the five pixel pairs closest to the moving image pixel, and the directions of the pixel pairs They are vertical direction, upper left to lower right direction, upper right to lower left direction, left upper left to right lower right direction, and right upper right to lower left lower direction. 2 7. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 19 of the scope of the patent application, when the co-located field difference of the third field is the initial field difference, the moving image detection The device generates the third field start signal. 2 8. According to the adaptive de-interlacing system of the four-picture field moving image described in item 19 of the scope of patent application, when the total number of field differences of the initial picture exceeds a third threshold value, the output switching device makes a decision according to the output Choose to use this 3: 2 to pull down the picture sequence and input 第44頁 1224925 六、申請專利範圍 出，否則該輸出切換裝置依該輸出決策選擇以該圖幅序列 ， 輸出。 2 9 .如申請專利範圍第1 9項所述四圖場動像的調適性解交 錯系統，其中上述3 : 2拉下還原裝置係於收到該第三圖場 起始訊號時，以3 : 2拉下還原法產生該3 : 2拉下圖幅序列。 3 0 .如申請專利範圍第2 9項所述四圖場動像的調適性解交 錯系統，其中上述3 .· 2拉下還原法係將產生每一該起始圖 場差之該第一圖場和該第三圖場與該第一圖場與該第三圖 | 場間的該第二圖場以該第一圖場與該第二圖場以交錯掃瞄 線的方式來產生一第一圖幅、一第二圖幅與一第三圖幅， 再將該第四圖場與一第五圖場以該第四圖場與該第五圖場 以交錯掃瞄線的方式來以產生一第四圖幅與一第五圖幅， 該第五圖場係於該圖場序列中該第一圖場之下一圖場，該 第一圖幅、該第二圖幅、該第三圖幅、該第四圖幅與該第 五圖幅係該3 : 2拉下圖幅中連續五個圖幅。 3 1. —種四圖場動像的調適性解交錯方法，包括： 接收一圖場序列中之一四圖場’該四圖場係一第一圖 看 場、一第二圖場、一第三圖場與一第四圖場且依序由先而 後被接收，其中，該圖場序列之每一圖場與其同位圖場係 在不同時間下，以相同複數個像素位置之一像素所構成， 並且，該些像素位置構成複數條掃瞄線；Page 44 1224925 6. The scope of patent application is out, otherwise, the output switching device chooses to output in the sequence of the picture frame according to the output decision. 2 9. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 19 of the scope of the patent application, wherein the above 3: 2 pull-down reduction device is 3 when receiving the start signal of the third picture field. The 2: 2 pull-down reduction method produces the 3: 2 pull-down sequence of pictures. 30. According to the adaptive de-interlacing system of the four-picture field moving image as described in item 29 of the scope of the patent application, wherein the above-mentioned 3 .2 pull-down reduction system will generate the first of each of the initial picture field differences. The second field between the field and the third field and the first field and the third field | The field between the first field and the second field are generated by interlaced scanning lines. The first frame, a second frame, and a third frame, and then the fourth and fifth fields are interlaced by the fourth and fifth fields. To generate a fourth frame and a fifth frame, the fifth field is a field below the first field in the field sequence, the first frame, the second frame, the The third frame, the fourth frame, and the fifth frame are five consecutive frames in the 3: 2 pull-down frame. 3 1. A adaptive de-interlacing method for a four-picture field including: receiving one of the four-picture fields in a sequence of one-picture fields; the four-picture field is a first picture viewing field, a second picture field, a The third field and a fourth field are received sequentially and sequentially, where each field in the field sequence and its co-located field are at different times, using one pixel at the same plurality of pixel positions. Structure, and the pixel positions constitute a plurality of scanning lines; -、第45頁 1224925 六、申請專利範圍 執行一動 動像區域外之 產生一圖 各掃瞄線交錯 之各該像素位 成，而所有該 法解交錯該動 像偵測，產生該第三圖場之一動像區域，嗜 區域為一靜像區域；及 ^ 幅，該圖幅係由該第三圖場與一内差圖場之 所產生，該内差圖場具有與該第二圖場相同 置，並且以各該像素位置之一内差像素所組 2差像素係分別以圖場内内差法與動態補 像區域與該靜像區域所產生。 32·如 錯方法 劃 數個該 之一； 偵 區塊之 則該第 似區塊 與該原 計 ，該第 對位置 偵 二圖場 央，當 申請專 ，其中 分該第 像素所 測各該 相對位 一區塊 ’否則 區塊之 算各該 一向量 ，否則 測各該 中該第 一第三 利範圍 上述動 三圖場 組成， 第3 1項所述四圖場動像的調適性解交 像偵測，包括： 成複數個原區塊，各該原區塊係由複 各該像素係各別對應到該些像素位置 原區塊之一第一區 置以一搜尋範圍内 存在且為最接近該 ，該第一區塊不存 一第一區塊差小於 原區塊 係從該 ，該第 之一第一向 第一區塊之 向量不存 原區塊之一第二區 一區塊之相對位置 區塊存在時，則該 塊，當該第一圖 存在至少一個相 原區塊之相對位 在，其中，該相 一第一門檻限值 量，當該第一區 相對位置至該原 在； 塊，該第二區塊 與該原 第二區 區塊之相 塊存在， 場中該原 似區塊， 置之該相 似區塊係 塊存在時 區塊之相 位於該第 對位置中 否則，該-, Page 45, 1224925 6. The patent application scope is executed to generate a picture outside the moving image area, each scanning line is interlaced, and all the pixels are deinterlaced, and all the methods are used to deinterleave the moving image detection to generate the third image. One of the fields is a moving image area, and the addiction area is a still image area; and ^ frames are generated by the third field and an internal difference field, the internal difference field having the same as the second field The two difference pixels that are the same and are set with one inner difference pixel at each pixel position are generated by the inner difference method of the field, the dynamic complementary region, and the still image region, respectively. 32. If the wrong method is used, draw one of the several; if the block is detected, the first similar block and the original plan, the second pair of position detection field map center, when applying for the special, which divided the second pixel measured each Relatively one block 'Otherwise the calculation of each block is a vector, otherwise the first and third profit ranges are composed of the above three moving fields, and the adaptive solution of the moving image of the four fields described in item 31 Cross-image detection includes: forming a plurality of original blocks, each of the original blocks being a corresponding one of the original blocks corresponding to the pixel positions in the first block; To get the closest, the first block does not store a first block. The difference between the first block and the original block is smaller than the original block. The vector from the first to the first block does not store one of the original blocks. Relative position of a block When a block exists, the block exists when the relative position of at least one phase original block exists in the first graph, where the phase is a first threshold value, and when the relative position of the first region reaches The original; block, the second block and the original second block block , Field like the original block, the set of blocks similar to blocks of coefficients of the blocks there is positioned relative to the first position or in the 1224925 六、申請專利範圍 第二區塊不存在； 偵測各該原區塊之該第三區塊，該第三區塊位於該第 四圖場中距該第二區塊該第一向量之相對位置，當該第三 區塊與該第二區塊之一第二區塊差小於一第一門檻限值， 該第三區塊存在，否則，該第三區塊不存在；及 計鼻各該原區塊之該移動向Έ ’當該第二區塊存在時 ，該移動向量係從該第二區塊至該原區塊，否則，該原區 塊不存在，同時，各該原區塊之該移動向量亦為該原區塊 所對應到之内各該像素位置的該移動向量。 3 3 .如申請專利範圍第3 2項所述四圖場動像的調適性解交 錯方法，其中上述第一區塊差係該原區塊與該第一區塊中 每一相關第一像素對之差值的總和，該相關第一像素對之 各該像素分別屬於該原區塊與該第一區塊，並且該第二區 塊差係該第二區塊與該第一區塊中每一相關第二像素對之 差值的總和，該相關第二像素對之各該像素分別屬於該第 二區塊與該第三區塊。 3 4.如申請專利範圍第3 3項所述四圖場動像的調適性解交 錯方法，其中上述像素差係亮度差。 3 5 .如申請專利範圍第3 2項所述四圖場動像的調適性解交 錯方法，其中上述動像區域係所有該移動向量不存在之該 原區塊所對應到的所有該像素位置之聯集，並且該動像區1224925 6. The second block of the patent application scope does not exist; the third block of each of the original blocks is detected, and the third block is located in the fourth field from the first vector of the second block Relative position, when the difference between the third block and one of the second blocks is less than a first threshold value, the third block exists; otherwise, the third block does not exist; and When the second block exists, the movement vector is from the second block to the original block; otherwise, the original block does not exist, and each of the original blocks does not exist. The motion vector of the block is also the motion vector of each of the pixel positions corresponding to the original block. 3 3. The adaptive de-interlacing method for a four-picture field moving image as described in item 32 of the scope of the patent application, wherein the first block difference is the original first block and each related first pixel in the first block The sum of the difference between the pair, each pixel of the related first pixel pair belongs to the original block and the first block, and the second block is the difference between the second block and the first block The sum of the differences of each related second pixel pair, and each pixel of the related second pixel pair belongs to the second block and the third block, respectively. 3 4. The adaptive de-interleaving method for a four-picture field moving image as described in item 33 of the scope of the patent application, wherein the above-mentioned pixel difference is a difference in brightness. 35. The adaptive de-interlacing method for a four-picture field moving image as described in item 32 of the scope of the patent application, wherein the moving image area is all the pixel positions corresponding to the original block where the motion vector does not exist Connected set, and the moving image area 第47頁 二、申請專利範圍 ，之所有其他該像素位置之聯集為一靜像區域 3β· ‘ 錯方法，Ϊ t ΐ J ί第32項所述四圖場動像的調適性解交 像素產生：ΐΐ動像補償法係以該第二圖場上之-第： 所在…素位置至該内差像素 向量。 Λ内差像素所對應之像素位置的該移動 钟如申請專利範圍第 ^方法，其中上述圖場 產生該内差圖場之 兮Ϊ像像素，在該圖幅 第三圖場之一第一掃 中產生各該動像像素 ，以一最相似像素對 為該些像素對中具有最 取得各該動像像素 —素為中心由複數個不 吗線各取一像素產生。 31項所述四圖場動像的調適性解交 間内差法係一邊緣方向内差法，包 各該掃猫線中屬於該動像區域之每 中，該内差圖場之該掃瞄線係介於 瞄線與一第二掃瞄線間； 、 ，該動像像素係於-複數個像素對 之平均值來產生，言亥最相似像素對 小差值之像素對；及 I对 二目關之複數個像素對，係、以該動像 同方向㈣第-掃猫線與該第二掃 .如申请專利範圍第3 了货祕阳固 錯方法，其中上述複i像二場動像的調適性解 象素對，該些像㈣之方向分別 像像素之五 且方向、左上到右 1224925 六、申請專利範圍 方向、右上至左下方向、左左上至右右下方向與右右上至 左左下方向。 像置儲 動裝二 場存第 圖儲一 四一由 種第係 一 一場 . 圖 的 存 第輸 該後 ，間 :場時 括圖段 包一 1 ，第由 統一經 系出且 錯輸並 交並 ， 解存入 性儲、輸 適以置 調用裝 出 第輸 該後 ，間 場時 圖段 二該 第由 一 經 出且 輸並 並’ 存入 儲輸 以置 用裝 ，存 置儲 裝三 存第 儲一 二由 第係 該場 圖 第出 該輸 ，後 場間 圖時 三段 第該 一由 出經 輸且 並並 存， 儲入 以輸 用流 ，料 置資 裝場 存圖 儲一 三由 第係 該場 • ’IS1I 出 三 一圖場資料流，用以輸入一第四圖場，該第一圖場、 該第二圖場、該第三圖場與該第四圖場係一圖場序列中依 序由先而後被接收之四圖場，其中，該圖場序列之每一圖 場與其同位圖場係在不同時間下，以相同複數個像素位置 之一像素所構成，並且，該些像素位置構成複數條掃瞄線 一動像偵測裝置，用以產生一移動向量序列與一像素 決策序列，該移動向量序列與該像素決策序列係根據該第 一圖場、該第二圖場、該第三場與該第四圖場產生； 一動像補償裝置，係根據該移動向量序列與該第三圖 場產生一第一像素序列，該第一像素序列係由各該像素位 IPage 47 II. The scope of the patent application, all other associated sets of this pixel position are a static image area 3β · 'Wrong method, Ϊ t ΐ J ί The adaptability of the four-picture field dynamic image described in item 32 to disjoin the pixels Generate: The method of compensating the moving image is based on the -th: on the second field, where the prime position is to the internal difference pixel vector. The moving clock at the pixel position corresponding to the Λ internal difference pixel is as described in the patent application method ^, wherein the above-mentioned field generates an image pixel of the internal difference field, which is scanned in one of the third fields of the frame first. Each of the moving image pixels is generated by taking a most similar pixel pair as the pixel pair having the most obtained each of the moving image pixels-prime as a center and taking one pixel from each of a plurality of lines. The adaptive inter-difference internal difference method of the four-image field described in item 31 is an edge-direction internal difference method, including each of the cat line belonging to the dynamic image area, and the internal image of the internal image field. The aiming line is between the aiming line and a second scanning line;,, the moving image pixel is generated by an average of a plurality of pixel pairs, and the pixel pair with the smallest difference between the most similar pixel pairs; and I For a plurality of pixel pairs of the second eye, the first scan line and the second scan are in the same direction as the moving image. For example, the third scope of the patent application is the method of fixing the goods, and the above complex i image is two. The adaptive resolution of field motion image pairs of pixels, the directions of these images are like the direction of five pixels and the direction, top left to right 1224925 VI, patent application direction, top right to bottom left, left top to right bottom right and top right to top Left to left direction. The storage of the image is stored in two fields, and the map is stored in the map. The map is stored in the map. The map is stored in the map. The map is stored in the map. Parallel and merge, after deducting into the storage, the transfer is appropriate to load and transfer, and then the second segment of the interim time is to be exported and merged, and then stored into the storage and transfer to use the equipment and the storage equipment. The three deposits, the first deposit, the first deposit, the second deposit, the first deposit and the export, and the three paragraphs in the inter-field diagram, the first and the first, the lose and the coexistence. The third field is the field • 'IS1I outputs the Trinity field data stream for inputting a fourth field, the first field, the second field, the third field, and the fourth field A field sequence consists of four fields that have been received one after another. Among them, each field of the field sequence and its co-located field are composed of one pixel at the same plurality of pixel positions at different times. In addition, the pixel positions constitute a plurality of scanning lines and a moving image detection device for generating a A motion vector sequence and a pixel decision sequence are generated according to the first field, the second field, the third field, and the fourth field; a motion image compensation device, A first pixel sequence is generated according to the motion vector sequence and the third field, and the first pixel sequence is determined by each of the pixel bits I 第49頁 一一 第一像 ~邊緣方 ^素序列，該 素所組成； 一像素切 素序列與該第 該内差圖場Ϊ 一圖場切 列係由複數個 内差圖場以掃 4 0 ·如 錯系統 時執行 劃 數個該 之一； 偵 區塊之 則該第 似區塊 與該原 計 1則該 中請專 ’其中 一動像 分該第 像素所 素所組成； 向内差裝置，係根據該第二圖場產生一第二 第二像素序列係由各該像素位置之一第二像 換裝置，用以依該像素決策序列於該第一像 一像素序列間切換以產生一内差圖場序列， 列係由複數個內差圖場所組成；及 換4置’用以輸出成一圖幅序列，該圖幅序 圖巾田所產生，各該圖幅係由該第三圖場與該 略線為單位交錯構成。 利範圍第3 9項所述四圖場動像的調適性解交 上述動像偵測裝置係於該第一圖場接收完成 4貞測方法，包括： 三圖場成複數個原區塊，各該原區塊係由複 組成’各該像素係各別對應到該些像素位置 <1 測各該原區塊之一第一區塊，當該第一圖場中該原 置以一搜，範圍内存在至少一個相似區塊， 存在f為最接近該原區塊之相對位置之該相 ΐ不存在’其中’該相似區塊係 °° Α差小於一第一門檻限值；區塊第—向量，當該第-區塊存在時 f係從該苐一區塊之相對位置至該原區塊之 相對位 一區塊 ，否則 區塊之 算各該 第一向Page 49: The first image ~ the edge square element sequence, which is composed of a pixel; a pixel slice sequence and the first internal difference field 图 The field sequence is composed of a plurality of internal difference fields. 0 · If the system is faulty, execute one of several ones; the block that is detected is the first block and the original number is 1 and the special one is composed of the first pixel; the inward difference A device for generating a second second pixel sequence according to the second field is a second image changing device for each of the pixel positions, and is used for switching between the first image-pixel sequence according to the pixel decision sequence to generate An internal difference map field sequence, the column system is composed of a plurality of internal difference map places; and 4 sets of 'for outputting a sequence of map frames, the map sequence sequence map generated by the field, each of the map series is from the third map The field is interlaced with the outline. The adaptability of the four-picture field moving image described in item 39 of the scope of interest is that the above-mentioned moving-image detection device completes the four test methods after receiving the first picture field, including: the three-picture field is composed of a plurality of original blocks, Each of the original blocks is composed of a plurality of 'each of the pixel systems respectively corresponding to the pixel positions < 1 to measure a first block of each of the original blocks, and when the original field is Search, there is at least one similar block in the range, and the existence of f is the relative position closest to the original block's relative position does not exist 'where' the similar block is °° Δ difference is less than a first threshold value; Block number-vector, when the first block exists, f is a block from the relative position of the first block to the relative position of the original block, otherwise the block count is the first direction 第50頁 1224925 六、申請專利範圍 相對位置，否則，該第一向量不存在； · 偵測各該原區塊之一第二區塊，該第二區塊位於該第 二圖場中該第一區塊之相對位置與該原區塊之相對位置中 央，當一第三區塊存在時，則該第二區塊存在，否則，該 第二區塊不存在， · 債測各該原區塊之該第三區塊，該第三區塊位於該第 四圖場中距該第二區塊該第一向量之相對位置，當該第三 區塊與該第二區塊之一第二區塊差小於一第一門檻限值， 該第三區塊存在，否則，該第三區塊不存在；及 計算各該原區塊之該移動向量，當該第二區塊存在時 $ ，該移動向量係從該第二區塊至該原區塊，否則，該原區 塊不存在，同時，各該原區塊之該移動向量亦為該原區塊 所對應到之内各該像素位置的該移動向量。 4 1.如申請專利範圍第4 0項所述四圖場動像的調適性解交 錯系統，其中上述第一區塊差係該原區塊與該第一區塊中 每一相關第一像素對之差值的總和，該相關第一像素對之 各該像素分別屬於該原區塊與該第一區塊，並且該第二區 塊差係該第二區塊與該第一區塊中每一相關第二像素對之 差值的總和，該相關第二像素對之各該像素分別屬於該第 · 二區塊與該第三區塊。 4 2 .如申請專利範圍第4 1項所述四圖場動像的調適性解交 · 錯系統，其中上述像素差係亮度差。Page 50 1224925 6. Relative position of patent application scope, otherwise, the first vector does not exist; · Detect one of the original blocks, the second block, and the second block is located in the second field in the second field. The relative position of a block is in the center of the relative position of the original block. When a third block exists, the second block exists; otherwise, the second block does not exist. The third block of the block, the third block is located in the fourth field relative to the first vector of the second block, when the third block and one of the second block are second The block difference is less than a first threshold value, the third block exists, otherwise, the third block does not exist; and the movement vector of each original block is calculated, when the second block exists, $, The motion vector is from the second block to the original block, otherwise, the original block does not exist, and the motion vector of each original block is also the pixel corresponding to the original block. The movement vector of the position. 4 1. The adaptive de-interlacing system for a four-field motion image as described in item 40 of the scope of the patent application, wherein the first block difference is the first pixel and each related first pixel in the first block. The sum of the difference between the pair, each pixel of the related first pixel pair belongs to the original block and the first block, and the second block is the difference between the second block and the first block The sum of the differences of each related second pixel pair, and each pixel of the related second pixel pair belongs to the second block and the third block, respectively. 4 2. The adaptive de-interleaving and error-correcting system for a four-picture field moving image as described in item 41 of the scope of patent application, wherein the above-mentioned pixel difference is a difference in brightness. 第51 -頁 1224925 六、申請專利範圍 4 3 .如申請專利範圍第4 0項所述四圖場動像的調適性解交 錯系統，其中上述移動向量序列係由各該像素位置之該移 動向量所組成。 4 4.如申請專利範圍第4 0項所述四圖場動像的調適性解交 錯系統，其中上述内差圖場序列係由複數個内差圖場所組 成，各該内差圖場係由複數條内差掃瞄線所組成，各内差 掃瞄線係由複數個該像素位置之一内差像素所組成。 4 5 .如申請專利範圍第4 4項所述四圖場動像的調適性解交 錯系統，其中上述像素決策序列係由各該像素位置之一像 素決策所組成，當該像素位置之移動向量存在時，該像素 切換裝置依該像素決策以該第一像素為該内差像素，否 則，該像素切換裝置依該像素決策以該第二像素為該内差 像素。 4 6 .如申請專利範圍第4 0項所述四圖場動像的調適性解交 錯系統，其中上述動像補償裝置係於該第二圖場以各該像 素位置距該移動向量之像素為第一像素，當該移動向量不 存在時，則該第一像素不存在。 4 7 .如申請專利範圍第4 0項所述四圖場動像的調適性解交 錯系統，其中上述邊緣方向内差裝置係依該第三圖場以邊Page 51-Page 1224925 VI. Patent Application Range 4 3. The adaptive de-interlacing system of the four-picture field moving image as described in item 40 of the patent application range, wherein the above motion vector sequence is the motion vector of each pixel position Composed of. 4 4. The adaptive de-interlacing system of the four-picture field moving image as described in item 40 of the scope of patent application, wherein the above-mentioned internal difference map field sequence is composed of a plurality of internal difference map sites, and each of the internal difference map fields is composed of A plurality of internal difference scanning lines are formed, and each internal difference scanning line is composed of a plurality of internal difference pixels at one of the pixel positions. 4 5. According to the adaptive de-interlacing system of the four-picture field moving image described in item 4 of the scope of the patent application, the above pixel decision sequence is composed of one pixel decision of each pixel position, and when the pixel vector's motion vector When present, the pixel switching device uses the first pixel as the internal difference pixel according to the pixel decision, otherwise, the pixel switching device uses the second pixel as the internal difference pixel according to the pixel decision. 46. The adaptive de-interlacing system of the four-picture field moving image as described in item 40 of the scope of the patent application, wherein the moving image compensation device is located in the second picture field, and the pixels of each pixel position away from the motion vector are First pixel, when the motion vector does not exist, the first pixel does not exist. 47. The adaptive de-interleaving system of the four-picture field moving image as described in item 40 of the scope of patent application, wherein the above-mentioned edge direction internal difference device is based on the third picture field. 第52頁 丄 ^-s-- 申清專利範圍 -------______ 緣方向内差法產生—邊緣 ^ 圖場係由複數停邊緣6 11差圖場，該邊緣方 略'線伟由通^ 向内差掃猫線所％成，内差 由複數個該第二像素所組成。成’各該内差掃 ^心如申請專利範圍第4 卜糸2“其中上述邊緣方向内差：圖2像的調適性解交 ，該邊緣緣方向内差圖場之各該方6 邊緣方向内差掃晦線 遺緣方向内差择 丨弟-掃晦線與—第二掃r線：以產生於介於該第三圖%::線 中產生各該第二像："：；該=像素所構成 丨為該像素對之平均：ίίί於；？數個像素對 4::第中具::;差值之像素ί;:攻相似像素斟 像素為中，、、Α —像素相關之複數個像青料. 瞄線各取：像2個不同方向於該第—掃瞄唛：以該動像 像素產生。 Τ 線與該第二掃 I錯系統，：f 7範圍第48項所述四圖場動像的 對像素對/、I上述複數個像素對係最接# 2調適性解交 下方向对*該些像素對之方向分別為；ί:動像像素之五 卜左左下至左下方向、左左上至右J;;、左上到* 方向。 右下方向與右右上 5〇 產；邊；方向内差法，包含： 猫線間之一邊麵 ji於一第一掃瞄線與一第 第53頁 1224925 六、申請專利範圍 一 I方向内差掃瞄、線， 丨係根據該第一掃瞄線盥:亥：向：差掃瞄線之每一内差像素 產生每1邊緣“内；C ; |之一最相似像素對取其平 ，丄係猎由複數個像素對 為該複數個像素對中具有一差佶=，該最相似像素對係 取得每一該邊緣方向肉M值取小之像素對；及 Μ該邊緣方向内差像；：=相：之複數個像素對， …亥第二掃瞄線各取—像素：：不问方向於該第 51. 如申請專利範圍第 述複數個像素對係該2f邊緣方向内 像素對之方向分別為垂像像素之五對i辛V: 左左上至右右下方向與右方向、右上至 52. -種3:2拉下解交錯方 i左左下方向。 接收一圖場序列匕括： 圖場資料流所接 I收； ％序列係由一 產生每一第_圖場之— 於母次該圖場序列之該第—圖上圖場差，該 圖場差係根據該第一圖場與3接收完成時產：因；2 記錚—nf%之刖同位圖場；所產生，該第二圖 丨己錄冋位圖場差序列，咕 圖場之該同位R "亥同位圖i碁至+ I由包含嗜第」Ϊ场差產生後記錄，：差序列係於該第 ° 圖場之該同位圖場差及;;冋位圖場差序列係 及向前推算之複數個同Page 52 丄 ^ -s-- Application for Patent Scope -------______ Generated by the internal difference method of the edge direction—Edge ^ The picture field is composed of a complex stop edge 6 11 difference picture field, and the edge strategy 'line Weiyou The internal difference is formed by scanning the cat line, and the internal difference is composed of a plurality of the second pixels. If the difference is within the scope of the patent application No. 4 bu 2 "wherein the above-mentioned internal difference in the edge direction: the adaptability of the image shown in Figure 2 disjoint, each edge of the edge difference in the field of the image 6 edge direction Internal Difference Scanning Lines In the direction of the margin, internal difference 丨 brother-scanning line and-the second scanning r line: to produce each of the second images between the third% :: line: ":; This = pixel constitutes the average of the pixel pair: ίίί; several pixel pairs 4 :: 第 中 中 ::; the difference of the pixel ί :: attack similar pixels, consider the pixels as medium, and Α — pixels A plurality of related image materials. Each line of sight is taken: two different directions of the image are taken at the first-scanning scan: generated by the moving image pixels. The T-line and the second scan I error system, f 48 range 48th The pair of pixel pairs of the four-picture field moving image described in the item above, the above-mentioned plurality of pixel pairs are most closely connected to each other. # 2 Adaptability The direction of the intersecting downward direction. The directions of the pixel pairs are: Left-left to bottom-left direction, left-left to top-right J ;;, top-left to * direction. Bottom-right direction and right-top-right 50 production; side; direction difference method, including: between cat lines One of the sides is a first scanning line and a page 53 of 1224925. 6. The scope of patent application-a differential scanning, line in the I direction, according to the first scanning line: Hai: Direction: Differential scanning Each intra-difference pixel of the sight line produces every 1 edge "inside; C; | one of the most similar pixel pairs is flattened, that is, a plurality of pixel pairs has a difference in the plurality of pixel pairs. The most similar pixel pair is to obtain each pixel pair with a small M value in the edge direction; and M in the edge direction difference image;: = phase: a plurality of pixel pairs, ... each pixel in the second scanning line ：: Regardless of the direction of the 51st. As mentioned in the scope of the patent application, the plurality of pixel pairs is the direction of the pixel pairs in the 2f edge direction, which are five pairs of vertical image pixels i Sin. V: Left top left to right right bottom direction and Right direction, upper right to 52.-Kind of 3: 2 pull down the de-interlaced square i left left lower direction. Receiving a picture field sequence: Received by the picture field data stream; The% sequence is generated by one of each of the first picture field—the difference between the picture field on the first and the second picture field sequence. The field difference is produced according to the first picture field and 3 when the reception is completed: cause; 2 records—nf% of the co-located picture field; generated, the second picture has recorded the field difference sequence of the bit map, Gou field The parity R " Hai parity maps i 碁 to + I are recorded after the field difference including the tropism "is generated: the difference sequence is based on the parity map field difference of the °° field; and Sequences and multiple forward projections 第54頁 1224925 六、申請專利範圍 位圖場差所組成； 執行3 : 2拉下偵測以產生一 3 : 2拉下偵測結果，該3 : 2 拉下偵測結果係依該同位圖場差序列產生； 解交錯該圖場序列以產生一圖幅序列，該圖幅序列係 依該3 : 2拉下偵測結果以一圖場解交錯法產生；及 輸出該圖幅序列，.該圖幅序列係由一圖幅資料流輸出 5 3 .如申請專利範圍第5 2項所述3 : 2拉下解交錯方法，其 中上述同位圖場差係該第一圖場與該第一圖場之該前同位 $ 圖場之差異像素數量。 5 4.如申請專利範圍第5 2項所述3 : 2拉下解交錯方法，其 中上述3 : 2拉下偵測係於該同位圖場差序列中搜尋複數個 起始圖場差，該些起始圖場差皆小於一第一門檻限值且每 隔五次出現一次，當該起始圖場差之總數超過一第二門檻 限值，則該3 : 2拉下偵測結果為已偵測出3 : 2拉下，否則， 該3 : 2拉下偵測結果為尚未偵測出3 : 2拉下。 5 5 .如申請專利範圍第5 4項所述3 : 2拉下解交錯方法，係 ί 當3 : 2拉下偵測結果為尚未偵測出3 : 2拉下時，該圖場解交 錯法包括：圖場内内差法，圖場間内差法與調適性解交錯 法0Page 54 1224925 6. The patent application range consists of bitmap field difference; perform 3: 2 pull-down detection to generate a 3: 2 pull-down detection result, and the 3: 2 pull-down detection result is based on the same bitmap Field difference sequence generation; de-interlacing the field sequence to generate a frame sequence, the frame sequence is generated by a field de-interlacing method according to the 3: 2 pull-down detection result; and outputting the frame sequence. The frame sequence is output from a frame data stream 5 3. As described in item 52 of the patent application scope, the 3: 2 pull-down de-interlacing method is used, wherein the difference between the above-mentioned isotopic field is the first field and the first field. The number of pixels in the field of the previous parity $ field. 5 4. The 3: 2 pull-down de-interlacing method as described in item 52 of the scope of the patent application, wherein the above 3: 2 pull-down detection is searching for a plurality of initial picture field differences in the same-image field difference sequence. These initial picture field differences are less than a first threshold value and occur every five times. When the total of the initial picture field differences exceeds a second threshold value, the detection result of the 3: 2 pull-down is A 3: 2 pulldown has been detected, otherwise, the 3: 2 pulldown detection result is that a 3: 2 pulldown has not been detected. 5 5. The method of 3: 2 pull-down de-interlacing as described in item 5 4 of the scope of patent application is that when the detection result of 3: 2 pull-down is not detected yet 3: 2 pull-down, the field is de-interlaced. Methods include: intra-field intra-field difference method, inter-field intra-field difference method and adaptive de-interlacing method0 第55頁 1224925 六、申請專利範圍 5 6 .如申請專利範圍第5 4項所述3 .· 2拉下解交錯方法，其 ， 中上述圖場解交錯法係當3 : 2拉下偵測結果為已偵測出3 : 2 拉下時以3 : 2拉下還原法產生，該3 : 2拉下還原法係將產生 每一該起始圖場差之該第一圖場與一第二圖場及該一圖場 與該第二圖場間之一第三圖場以該第二圖場與該第三圖場 以交錯掃瞄線的方式來產生一第一圖幅、一第二圖幅與一 第三圖幅，再將一第四圖場與一第五圖場以該第四圖場與 該第五圖場以交錯掃瞄線的方式來產生一第四圖幅與一第 五圖幅，該第二圖場、第四圖場與該第五圖場分別為該第 一圖場之前同位圖場、下一圖場與下同位圖場。 Φ 5 7 . —種3 : 2拉下解交錯系統，包括： 一第一儲存裝置，用以儲存並輸出一第一圖場，該第 一圖場係由一第二儲存裝置輸入，並且經由一段時間後輸 出； 該第二儲存裝置，用以儲存並輸出一第二圖場，該第 二圖場係由一，圖場資料流輸入，並且經由該段時間後輸出 該圖場資料流，用以輸入一第三圖場，該第一圖場、 該第二圖場與該第三圖場係一圖場序列中依序由先而後被 < 接收之三圖場，其中，該圖場序列之每一圖場與其同位圖 場係在不同時間下，以相同複數個像素位置之一像素所構 成，並且，該些像素位置構成複數條掃瞄線； 一 3 : 2拉下偵測裝置，係根據該第一圖場與該第三圖Page 55 1224925 VI. Application scope 5 6. As described in item 54 of the scope of application patent 3. 2 pull-down de-interlacing method, where the above field de-interlacing method is when 3: 2 pull-down detection The result is that the 3: 2 pull-down reduction method has been detected when 3: 2 pull-down is detected, and the 3: 2 pull-down reduction method will generate the first field and a first The second field and a third field between the first field and the second field generate a first frame, a first field by using the second field and the third field in a staggered scanning line. A second frame and a third frame, and then a fourth field and a fifth field are generated by the fourth field and the fifth field in an interlaced scanning line to generate a fourth frame and A fifth picture frame, the second picture field, the fourth picture field, and the fifth picture field are respectively the co-located picture field before the first picture field, the next picture field and the lower co-located picture field. Φ 5 7. — Kind of 3: 2 pull-down de-interlacing system, including: a first storage device for storing and outputting a first field, the first field is input by a second storage device, and Output after a period of time; the second storage device is used for storing and outputting a second field, which is input by a field data stream, and outputs the field data stream after the period of time, For inputting a third field, the first field, the second field, and the third field are three fields sequentially received in sequence by a field sequence, where the image Each field of the field sequence and its co-located field are composed of one pixel at the same plurality of pixel positions at different times, and the pixel positions constitute a plurality of scanning lines; a 3: 2 pull-down detection Device according to the first field and the third field 第56頁 1224925 六、申請專利範圍 場產生一 3 : 2同位圖場差序列，並且以一 3 : 2拉下偵測產生 一 3 : 2拉下偵測結果與該第三圖場之一起始訊號； 一 3 : 2拉下還原裝置，係依該第三圖場之該起始訊號 與該第一圖場、該第二圖場與該第三圖場以一 3: 2拉下還 原產生一 3 : 2拉下圖幅序列； 一圖場解交錯裝置，係依該第一圖場、該第二圖場與 該第三圖場以一圖場解交錯法產生一圖幅序列；及 一圖場輸出裝置，係依該3 .· 2拉下結果於該3 : 2拉下圖 幅序列與該圖幅序列擇一輸出。 5 8 .如申請專利範圍第5 7項所述3 : 2拉下解交錯系統，其 中上述產生同位圖場差之方法，包括： 產生各該第三圖場之一同位圖場差；及 記錄一同位圖場差序列，該同位圖場差序列係於該第 三圖場之該同位圖場差產生後記錄，該同位圖場差序列係 由包含該第三圖場之該同位圖場差及向前推算之複數個該 同位圖場差所組成。 5 9 .如申請專利範圍第5 8項所述3 : 2拉下解交錯系統，其 中上述同位圖場差係該第一圖場與該第三圖場之差異像素 數量。 6 0 .如申請專利範圍第5 7項所述3 : 2拉下解交錯系統，其 中上述3 : 2拉下偵測係於該同位圖場差序列中搜尋複數個Page 56 1224925 VI. The patent application field generates a 3: 2 parity map field difference sequence and starts with a 3: 2 pull-down detection to generate a 3: 2 pull-down detection result starting with one of the third field Signal: 3: 2 pull-down reduction device is generated according to the initial signal of the third field and the first field, the second field and the third field are restored by a 3: 2 A 3: 2 pull-down sequence; a field de-interlacing device generates a frame sequence according to the first field, the second field and the third field by a field de-interlacing method; and A picture field output device selects and outputs one of the picture sequence and the picture sequence according to the 3... 2 pull down result. 58. The 3: 2 pull-down de-interlacing system as described in item 57 of the scope of patent application, wherein the method for generating an isotopic field difference includes: generating an isotopic field difference for each of the third fields; and recording Co-location field difference sequence, the co-location field difference sequence is recorded after the co-location field difference of the third field is generated, and the co-location field difference sequence is composed of the co-location field difference including the third field And the forward-computed multiple isotopic map field differences. 59. The 3: 2 pull-down de-interlacing system as described in item 58 of the scope of patent application, wherein the difference between the above-mentioned isotopic field is the number of pixels of the difference between the first field and the third field. 60. The 3: 2 pull-down de-interlacing system as described in item 57 of the scope of patent application, wherein the above 3: 2 pull-down detection is based on searching a plurality of the same bitmap field difference sequences. 第57頁 1224925 六、申請專利範圍 起始圖場差，該些起始圖場差皆小於一第一門檻限值且每 隔五次出現一次，當該起始圖場差之總數超過一第二門檻 限值，則該3 : 2拉下偵測結果為已偵測出3 : 2拉下，否則， 該3 : 2拉下偵測結果為尚未偵測出3 : 2拉下。 6 1.如申請專利範圍第5 7項所述3 : 2拉下解交錯系統，其 中上述圖場解交錯法包括：圖場内内差法，圖場間内差法 與調適性解交錯法。 6 2 .如申請專利範圍第5 8項所述3 : 2拉下解交錯系統，其 中上述3 : 2拉下偵測裝置係於該第三圖場之同位圖場差為 該起始圖場時發出一第三圖場之起始訊號。 6 3 .如申請專利範圍第6 2項所述3 : 2拉下解交錯系統，其 中上述3 : 2拉下還原裝置係於收到該第三圖場之該起始訊 號時以3 : 2拉下還原法產生該3 : 2拉下圖幅序列，該3 : 2拉 下還原法係將產生每一該起始圖場差之該第一圖場與一第 三圖場及該一圖場與該第三圖場間之該第二圖場以該第二 圖場與該第三圖場以交錯掃瞄線的方式來產生一第一圖幅 、一第二圖幅與一第三圖幅，再將一第四圖場與一第五圖 場以該第四圖場與該第五圖場以交錯掃瞄線的方式來產生 一第四圖幅與一第五圖幅，該第四圖場與該第五圖場分別 為該第三圖場之下一圖場與下同位圖場。Page 57 1224925 6. The initial picture field differences in the scope of patent application. These initial picture field differences are less than a first threshold value and appear every five times. When the total of the initial picture field differences exceeds one If the threshold value is two, the 3: 2 pull-down detection result is that a 3: 2 pull-down has been detected; otherwise, the 3: 2 pull-down detection result is that a 3: 2 pull-down has not been detected. 6 1. The 3: 2 pull-down de-interlacing system as described in item 57 of the scope of patent application, wherein the above-mentioned field de-interlacing methods include: within-field internal difference method, between-field internal difference method and adaptive de-interlacing method. 62. The 3: 2 pull-down de-interlacing system as described in item 58 of the scope of patent application, wherein the above-mentioned 3: 2 pull-down detection device is in the same field of the third field as the starting field. A start signal of the third field is issued at any time. 63. The 3: 2 pull-down de-interlacing system as described in item 62 of the scope of patent application, wherein the above 3: 2 pull-down reduction device is 3: 2 when it receives the start signal of the third field. The pull-down reduction method generates the 3: 2 pull-down image sequence, and the 3: 2 pull-down reduction system will generate the first field, a third field, and the one field for each of the initial field differences. The second field between the field and the third field generates a first frame, a second frame, and a third field by using the second field and the third field in a staggered scanning line. A picture frame, and then a fourth picture field and a fifth picture field are used to generate a fourth picture frame and a fifth picture frame in a staggered scanning line manner using the fourth picture field and the fifth picture field. The fourth field and the fifth field are the next field and the co-located field below the third field, respectively. 第58頁Page 58
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW92113387A TWI224925B (en) | 2003-05-16 | 2003-05-16 | 4-field motion adaptive deinterlacing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW92113387A TWI224925B (en) | 2003-05-16 | 2003-05-16 | 4-field motion adaptive deinterlacing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TW200427316A TW200427316A (en) | 2004-12-01 |
| TWI224925B true TWI224925B (en) | 2004-12-01 |
Family
ID=34568359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW92113387A TWI224925B (en) | 2003-05-16 | 2003-05-16 | 4-field motion adaptive deinterlacing |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI224925B (en) |
-
2003
- 2003-05-16 TW TW92113387A patent/TWI224925B/en not_active IP Right Cessation
Also Published As
| Publication number | Publication date |
|---|---|
| TW200427316A (en) | 2004-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4159606B2 (en) | 2008-10-01 | Motion estimation |
| US7652721B1 (en) | 2010-01-26 | Video interlacing using object motion estimation |
| KR100996216B1 (en) | 2010-11-24 | High quality deinterlacing and frame multiplication circuits and methods |
| EP0018856B1 (en) | 1990-08-16 | Television display system |
| KR100303728B1 (en) | 2001-09-29 | Deinterlacing method of interlaced scanning video |
| KR100403364B1 (en) | 2003-10-30 | Apparatus and method for deinterlace of video signal |
| CA1193712A (en) | 1985-09-17 | Double-scanning non-interlace television receiver |
| US6975359B2 (en) | 2005-12-13 | Method and system for motion and edge-adaptive signal frame rate up-conversion |
| US20060256238A1 (en) | 2006-11-16 | Frame interpolation and apparatus using frame interpolation |
| JP3855761B2 (en) | 2006-12-13 | Image signal processing apparatus and method |
| JPWO2004068852A1 (en) | 2006-05-25 | Imaging device |
| JP4772562B2 (en) | 2011-09-14 | Pull-down signal detection device, pull-down signal detection method, progressive scan conversion device, and progressive scan conversion method |
| US20080136963A1 (en) | 2008-06-12 | Method and apparatus for reconstructing image |
| TW200400755A (en) | 2004-01-01 | Method and system for edge-adaptive interpolation for interlace-to-progressive conversion |
| CN100433791C (en) | 2008-11-12 | Film mode correction in still areas |
| KR20050037293A (en) | 2005-04-21 | Artifacts suppress apparatus of video signal and a method thereof |
| US9100531B1 (en) | 2015-08-04 | Method and apparatus for performing robust cadence detection in a video deinterlacer |
| US7129989B2 (en) | 2006-10-31 | Four-field motion adaptive de-interlacing |
| TWI224925B (en) | 2004-12-01 | 4-field motion adaptive deinterlacing |
| JP5241632B2 (en) | 2013-07-17 | Image processing circuit and image processing method |
| CN103119931B (en) | 2016-04-13 | For the method and apparatus of de interlacing video data |
| WO2006082426A2 (en) | 2006-08-10 | De-interlacing of video data |
| US7423692B2 (en) | 2008-09-09 | De-interlace method and method for generating de-interlace algorithm |
| CN101888473B (en) | 2012-05-23 | Text protection device and dynamic adaptive deinterlacing device |
| TW200534699A (en) | 2005-10-16 | De-interlacing module and related method capable of adaptively de-interlacing a video field |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 2013-09-01 | MM4A | Annulment or lapse of patent due to non-payment of fees |