CN109407113A - A kind of monitoring of woods window change in time and space and quantization method based on airborne laser radar - Google Patents

本发明提供一种基于机载激光雷达的林窗时空变化监测与量化方法，包括以下步骤：步骤1：基于多时相的激光雷达点云数据，识别林窗、绘制林窗边缘矢量图并保存林窗矢量数据集；步骤2：比较两个时期的林窗矢量数据集，构建林窗新生成、稳定、扩展、萎缩、位移或闭合状态的逻辑数学模型；步骤3：将两个时期的林窗矢量数据集叠合，进行林窗矢量多边形叠置分析，根据所述步骤2构建的逻辑数学模型，构建林窗矢量多边形变化状态的判别指标。本发明基于激光雷达三维点云数据进行林窗识别与空间分布的绘制，能快速、准确绘制多尺度林窗分布多边形，实现林窗空间分布的监测，解决林窗各种变化状态的科学表征问题，最终实现林窗时空变化监测与量化。

The present invention provides a method for monitoring and quantifying temporal and spatial changes of forest gaps based on airborne laser radar. Window vector data set; Step 2: Compare the gap vector data sets of the two periods, and construct a logical mathematical model of the newly generated, stable, expanded, shrinking, displaced or closed states of the gap; Step 3: Combine the gaps of the two periods The vector data sets are superimposed, and the superposition analysis of the gap vector polygon is performed, and the discriminant index of the change state of the gap vector polygon is constructed according to the logical mathematical model constructed in the step 2. Based on the three-dimensional point cloud data of the laser radar, the invention can identify the gaps and draw the spatial distribution, can quickly and accurately draw the multi-scale gap distribution polygons, realize the monitoring of the spatial distribution of the gaps, and solve the problem of scientific characterization of various changing states of the gaps. , and finally realize the monitoring and quantification of temporal and spatial changes of forest gaps.

A kind of monitoring of woods window change in time and space and quantization method based on airborne laser radar

Technical field

The invention belongs to forestry remote sensing technical fields, and in particular to a kind of woods window change in time and space based on airborne laser radar Monitoring and quantization method.

Background technique

Woods window is one of the hot issue of recent domestic forest management and administration concern." woods window (canopy gap) " one Word is initially proposed by British scholar Watt nineteen forty-seven, caused by indicating after one plant or more canopy arbor is dead in forest Chearance or small location, the space for being new individual invasion, occupying and updating.Forest cycle theory regards forest as spatially different Matter, " the flowing mosaics " changed on the time, the dynamic change of woods window spatial distribution (size, shape and position) is in certain journey The heterogeneity of Disturbed habit is enhanced on degree, affect the property of patch in group and its inlays situation, to influence group's knot Structure and dynamic process, the final structure and function for influencing landscape, therefore the quantization to woods window spatial distribution monitoring and changing condition, Be the key that comprehensive system recognize forest ecosystem long-term dynamics change procedure, be science carry out forest management and administration base Plinth.

In the different phase of woods window development, the heterogeneity of space structure and the heterogeneity of function are interweave variation or development 's.The change in time and space such as generation, stabilization (maintenance), extension, atrophy, displacement or the closure of woods window enhance woods window function heterogeneousization, Selection index system certainly will be played to plant intrusion, seed sprouting, seedling establishment, sapling up-growth.Limit by space data collection technology System, the heterogeneity that scholar analyzes the research of woods window woods window space structure with being concentrated mainly in single observation period " static state " are right The influence of species composition, form and the physiology of vegetation, update and growth etc., and " dynamic " monitor woods window change in time and space It studies less.The monitoring of woods window space structure relies primarily on or field investigation, including the direct method of measurement and photograph estimate method, Such method is time-consuming and laborious, high labor cost and is affected by extraneous factor；Again due to lack changing pattern quantizating index, Lead to that the characterization of woods window changes in spatial distribution state is objective, not system, affects the standardization processing of woods window dynamic monitoring.Cause This, it is necessary to a kind of method or technique is studied, realizes and the space-time dynamic of gap distribution is monitored, the specific woods window that solves generates, is steady The scientific characterization problems of the variable conditions such as fixed, extension, atrophy, displacement or closure, it is final to realize the monitoring of woods window change in time and space and amount Change.

Airborne laser radar (Airborne Laser Scanning, ALS) is a kind of emerging active remote sensing technology, The high-precision vegetation structure information of forest ecosystem, dimensional topography feature can be obtained on multiple space and time scales.ALS is to woods window High precision monitor has great potential in terms of recognizing forest disturbance regime, updating rule and inverting, But China still belongs to the starting stage in this respect.Therefore, reinforce the research in terms of the spatial distribution dynamic monitoring of woods window, especially with Advanced remote sensing technology quantifies woods window Dynamic mode of time and space, helps to improve the theoretical level of woods window research, further The mechanism that the dynamic law and bio-diversity for disclosing forest cycle maintain has raising forest management and administration level important Realistic meaning.

Summary of the invention:

It is an object of the present invention to provide a kind of, and quick, objective, accurate measurements woods window space based on airborne laser radar are divided The quantization method of cloth and its variable condition, in order to make up the technological gap of prior art middle forest window change in time and space monitoring and quantization.

In order to achieve the above object, the present invention adopts the following technical scheme:

A kind of monitoring of woods window change in time and space and quantization method based on airborne laser radar, comprising the following steps:

Step 1: the laser radar point cloud data based on multidate, identification woods window are drawn woods window edge polar plot and are saved Woods window vector data collection；

The laser radar point cloud data refers to by fixed wing aircraft or the acquisition of UAV flight's laser radar scanner Remotely-sensed data, the data of each laser point include X, Y, Z coordinate data and echo strength data；Utilize existing remote sensing images The correct throwing of processing software The Environment for Visualizing Images (ENVI, v5.3) setting point cloud data Shadow coordinate system, measurement unit are rice.

Step 2: comparing the woods window vector data collection in two periods, building woods window is newly-generated, stable, extension, atrophy, displacement Or the logic mathematical model of closed state；

Step 3: the woods window vector data collection in two periods is overlapped, woods window vector polygon Overlap Analysis is carried out, according to The logic mathematical model that the step 2 constructs constructs the discriminant criterion of woods window vector polygon variable condition.

Preferably, in the step 1, if woods window vector data integrates as G_nIf the value of some point (x, y) is in woods window G (x, y) identifies that the model of woods window is as follows:

In formula, G (x, y)=1 indicates that the corresponding position point (x, y) is woods window；G (x, y)=0 indicates that point (x, y) is corresponding Position be non-woods window；X, y are respectively the abscissa and ordinate of grid (x, y), and CHM (x, y) is the canopy on grid (x, y) Height value；A is the discrimination threshold of canopy edge average height.Discrimination threshold is determined by empirical value or on the spot sample investigation, is such as sentenced Other threshold value is 5 meters.

Preferably, in the step 1, CHM is the difference of Malabar Pied Hornbill model and earth's surface elevation model, it may be assumed that CHM (x, y) =DSM (x, y)-DEM (x, y)；

DSM is Malabar Pied Hornbill model in formula, and DEM is digital elevation model.

Firstly, original point cloud is divided into canopy point cloud and Ground Point using adaptive irregular triangle network filtering method Cloud.Canopy point cloud is interpolated to DSM using kriging analysis method, Ground Point cloud interpolation is instead generated into DEM apart from weight interpolation method. The characteristics of characterizing in view of the density and canopy of cloud, setting DSM and DEM have identical raster resolution.

Secondly, carry out difference operation to DSM and DEM using ENVI software, canopy height Raster Data Model CHM is obtained, and Classified on the basis of this according to the binary map G (x, y) that the height threshold a of woods window edge canopy carries out woods window and Fei Lin window.

Then, woods window recognition result is post-processed, to improve accuracy of identification, using ArcGIS software in G (x, y) two Morphologic filtering operation is carried out on the basis of value figure, it is final to determine that woods window edge simultaneously carries out vectoring operations to it, draw woods window side Edge polar plot simultaneously saves data set G_n。

Preferably, the specific steps of the step 2 include:

Step 21: setting first latter two period t_i,t_jThe woods window vector data collection of acquisition isI < j in formula；I, j > 0；

Step 22: comparing t_iThe woods window vector data collection and t in period_jThe woods window vector data collection in period, t_iThe woods window in period Data set and t_jPeriod, corresponding woods window data set was compared, in fact it could happen that three kinds of situations: the woods window polygon of two phases does not have Overlapping, it is completely overlapped with partly overlap.Not being overlapped indicates t_iWoods window exists, and t_jWoods window disappears (closure)；Or t_iWoods window is not In the presence of, and t_jWoods window is newly-generated.The completely overlapped woods window individual for indicating two periods all exists, and woods window maintains stable state not have There is generation geometry deformation, or upper extension or atrophy in the original location.Partly overlapping phenomenon occurs are due to woods window edge Vegetation renewal Or the heterogeneity of dead spatial distribution, lead to t_jWoods window may be subjected to displacement and (not deform), and atrophy is simultaneously displaced or extends simultaneously position The dynamic such as shifting.The logic mathematical model of newly-generated woods window, stable, extension, atrophy, displacement or closed state is defined respectively are as follows:

It is newly-generated:

Closure:

Stablize:

Extend non-displacement:

Atrophy non-displacement:

Extension has displacement:

Atrophy has displacement:

Preferably, the specific steps of the discriminant criterion of building woods window temporal and spatial orientation include: in the step 3

Step 31: respectively to t_i,t_jThe woods window vector polygon in two periods carries out topological structure coding, in formula subscript i or J indicates the data obtained in i-th or j period, and defines i < j；I, j > 0；

Step 32: by t_i,t_jThe woods window vector polygon in two periods carries out overlapping operation, the comprehensive t of output figure layer_i,t_jTwo The attribute of a period woods window vector data collection, reserve window is in t_i,t_jThe feature of all polygons in two periods；Figure layer overlapping Original polygon element is divided into new element in the process, new element combines the attribute of original two datasets, result It is usually exactly that a polygon is carried out intersection operation by the spatial distribution state of another polygon, to be divided into multiple more Side shape, while copying the attribute of input object in new object attribute list in attribute assigning process, the purpose for the arrangement is that The change information of geometrical characteristic (woods window edge) and attributive character (woods window, non-woods window) in woods window spatial distribution is fully retained；

Step 33: the shared segmental arc of the woods window vector polygon after overlapping being encoded, it is more then to calculate single woods window Side shape attributive character t_i,t_jIt is superimposed the ratio of area and original forest gap area；

Step 34: constructing woods using the step 32 and the lamination process in step 33, shared segmental arc and area ratio The discriminant criterion of window variation, monitor different times woods window is newly-generated, stable, extension, atrophy, displacement or closure variable condition And quantify its variation degree.

Preferably, respectively to t in the step 31_i,t_jThe woods window vector polygon in two periods carries out topological structure coding When, indicate within the scope of woods window or indicate that already existing woods window, O indicate that woods window range is outer or indicate non-woods window using E.

Preferably, in the step 33, to the shared segmental arc of the woods window vector polygon after overlapping be encoded to BB, BI, IB, BN, NB, wherein B indicates the boundary of polygon, and I indicates the inside of polygon, and N indicates non-boundary or inside.

Preferably, in the step 33, the specific formula for calculation of area ratio are as follows:

P(EE_ij)=Area (EE_ij)/Area(E_i),

P(EO_ij)=Area (EO_ij)/Area(E_i),

P(OE_ij)=Area (OE_ij)/Area(E_i)；

In formula, P indicates that the ratio of area, Area indicate woods window range or area,

E_i: t_iPeriod, the polygon range of single woods window,

EEi_j: assuming that some woods window t_iPeriod exists, t_jPeriod, there is also EE indicated t_i~t_jThe overlapping portion of woods window in period Point,

OE_ij: assuming that t_iPeriod, some woods window did not occurred, in t_jPeriod has occurred, and OE indicates t_i~t_jDuring woods window change Change range,

EO_ij, it is assumed that t_iSome woods window of period exists, in t_jPeriod has disappeared, and EO indicates t_i~t_jDuring woods window disappearance Range.

Preferably, in the step 31, topology is carried out to woods window vector polygon in different times using ArcGIS software Structured coding.

Technical solution of the present invention has the advantages that

The present invention realizes gap distribution (arrow using airborne laser radar technology and geographical spatial data topology analyzing method Measure polygon) it fast and accurately monitors and draws, to single woods window, variation characteristic is (such as newly-generated, steady within the different observation phases The states such as fixed, extension, atrophy, displacement or closure) progress is objective, systematically quantifies, it is precisely supervised to improve forest space structure The level of survey promotes the scientific characterization of woods window change in time and space state.Present invention enhances woods window spatial distribution dynamic monitoring sides The research in face quantifies woods window Dynamic mode of time and space especially with advanced remote sensing technology, helps to improve woods window and grind The standardization studied carefully, the mechanism that the dynamic law and bio-diversity for further disclosing forest cycle maintain, improves forest warp Seek management level.

(1) woods window spatial distribution state is the basis of woods window research, and the present invention utilizes airborne laser radar technology and geography Spatial data topology analyzing method realizes that woods window polygon is fast and accurately monitored and drawn, to single woods window in different observations Variation characteristic (states such as newly-generated, stable, extension, atrophy, displacement or closure) progress is objective in phase, systematically quantifies, To improve the level that forest space structure precisely monitors, promote the standardization level of woods window spatial variations situation characterization；It solves The dynamic monitoring problem of woods window spatial distribution and single woods window specifically change journey in landscape scale existing in the prior art The problems such as precise expression of degree.

(2) it is directed to the monitoring of woods window spatial distribution, the airborne laser radar technology that the present invention uses is a kind of active distant Sense technology obtains the three dimensional point cloud for accurately reflecting Forest Canopy space structure by laser radar penetration capacity, in this base The drafting of woods window identification and spatial distribution is carried out on plinth, this method can quickly, accurately draw multiple dimensioned gap distribution polygon, benefit It is compared with the laser point cloud data of multidate, can accurately reflect the changing condition of woods window spatial distribution；Solves the prior art Be concentrated mainly on present in on-site inspection method it is time-consuming and laborious, process is tedious, precision is not high, the constraint by natural environment is larger Etc. problems.

(3) the solution of the present invention can complicated geographic object and phenomenon is simplified and be abstracted into computer be indicated, Processing and analysis, quantify woods window changes in spatial distribution situation using geographical spatial data topology analyzing method, energy system, The variation characteristics such as precise expression woods window is newly-generated, stable, extension, atrophy, displacement or closure；It solves at present for woods window space Changes in distribution situation lacks the quantizating index of complete set, science, and actual mechanical process subjective judgement is more, and human factor is dry The technical problems such as the precise expression of changing condition are disturbed.

Detailed description of the invention

Fig. 1 is canopy laser point cloud and earth's surface laser point cloud schematic diagram；

Fig. 2 is polygon Overlaying analysis schematic diagram；

Fig. 3 is region Malabar Pied Hornbill MODEL C HM schematic diagram to be measured；

Fig. 4 is that CHM sectional view differentiates woods window by threshold value of 5 meters of height；

Fig. 5 is that region woods window to be measured identifies schematic diagram；

Fig. 6 is the woods window edge schematic diagram of " extension " variable condition, a), b) is respectively same woods window t₁And t₂Period data；

Fig. 7 is the woods window edge schematic diagram of " atrophy " variable condition, a), b) is respectively same woods window t₁And t₂Period data.

Specific embodiment

The preferred embodiment of the present invention presented below, to help the present invention is further understood.Those skilled in the art answer It solves, the explanation of the embodiment of the present invention is merely exemplary, the scheme being not meant to limit the present invention.

Embodiment 1:

Woods window change in time and space monitoring and quantization method of this programme based on airborne laser radar, specific implementation are divided into following 3 A step:

Step 1: the laser radar point cloud data based on multidate, identification woods window are drawn woods window edge polar plot and are saved Woods window vector data collection；

Laser radar point cloud data, which refers to, obtains remote sensing number by fixed wing aircraft or UAV flight's laser radar scanner According to the data of each laser point include X, Y, Z coordinate data and echo strength data；It is soft using existing remote sensing image processing The correct projection coordinate of part The Environment for Visualizing Images (ENVI, v5.3) setting point cloud data System, measurement unit are rice.

Wherein, if woods window vector data integrates as G_nIf the value of some point (x, y) is G (x, y) in woods window, woods is identified The model of window is as follows:

In formula, G (x, y)=1 indicates that the corresponding position point (x, y) is woods window；G (x, y)=0 indicates that point (x, y) is corresponding Position be non-woods window；X, y are respectively the abscissa and ordinate of grid (x, y), and CHM (x, y) is the canopy on grid (x, y) Height value；A is the discrimination threshold of canopy edge average height.Discrimination threshold is determined by empirical value or on the spot sample investigation, is such as sentenced Other threshold value is 5 meters.

CHM is the difference of Malabar Pied Hornbill model and earth's surface elevation model, it may be assumed that

CHM (x, y)=DSM (x, y)-DEM (x, y)；(2)

DSM is Malabar Pied Hornbill model in formula, and DEM is digital elevation model.

Firstly, original point cloud is divided into canopy point cloud and Ground Point using adaptive irregular triangle network filtering method Cloud, referring to figure 1.Canopy point cloud is interpolated to DSM using kriging analysis method, instead apart from weight interpolation method by Ground Point Cloud interpolation generates DEM.The characteristics of characterizing in view of the density and canopy of cloud, setting DSM and DEM is differentiated with identical grid Rate.

Secondly, carry out difference operation to DSM and DEM using ENVI software, canopy height Raster Data Model CHM is obtained, and Classified on the basis of this according to the binary map G (x, y) that the height threshold a of woods window edge canopy carries out woods window and Fei Lin window.

Then, woods window recognition result is post-processed, to improve accuracy of identification, using ArcGIS software in G (x, y) two Morphologic filtering operation is carried out on the basis of value figure, it is final to determine that woods window edge simultaneously carries out vectoring operations to it, draw woods window side Edge polar plot simultaneously saves data set G_n。

Step 2: comparing the woods window vector data collection in two periods, building woods window is newly-generated, stable, extension, atrophy, displacement Or the logic mathematical model of closed state；Specific steps include:

Step 21: setting first latter two period t_i,t_jThe woods window vector data collection of acquisition isI < j in formula；I, j > 0；

Step 22: comparing t_iThe woods window vector data collection and t in period_jThe woods window vector data collection in period, t_iThe woods window in period Data set and t_jPeriod, corresponding woods window data set was compared, in fact it could happen that three kinds of situations: the woods window polygon of two phases does not have Overlapping, it is completely overlapped with partly overlap.Not being overlapped indicates t_iWoods window exists, and t_jWoods window disappears (closure)；Or t_iWoods window is not In the presence of, and t_jWoods window is newly-generated.The completely overlapped woods window individual for indicating two periods all exists, and woods window maintains stable state not have There is generation geometry deformation, or upper extension or atrophy in the original location.Partly overlapping phenomenon occurs are due to woods window edge Vegetation renewal Or the heterogeneity of dead spatial distribution, lead to t_jWoods window may be subjected to displacement and (not deform), and atrophy is simultaneously displaced or extends simultaneously position The dynamic such as shifting.The logic mathematical model of newly-generated woods window, stable, extension, atrophy, displacement or closed state is defined respectively are as follows:

It is newly-generated:

Closure:

Stablize:

Extend non-displacement:

Atrophy non-displacement:

Extension has displacement:

Atrophy has displacement:

Step 3: the woods window vector data collection in two periods is overlapped, woods window vector polygon Overlap Analysis is carried out, according to The logic mathematical model that the step 2 constructs constructs the discriminant criterion of woods window vector polygon variable condition.

Wherein, the specific steps of the discriminant criterion of building woods window temporal and spatial orientation include:

Step 31: using ArcGIS software respectively to t_i,t_jThe woods window vector polygon in two periods carries out topological structure volume Yard, the data that subscript i or j expression i-th or j period obtain in formula, and define i < j；I, j > 0 indicates woods window range using E The interior or already existing woods window of expression, O indicate that woods window range is outer or indicate non-woods window.

Step 32: by t_i,t_jThe woods window vector polygon in two periods carries out overlapping operation, the comprehensive t of output figure layer_i,t_jTwo The attribute of a period woods window vector data collection, reserve window is in t_i,t_jThe feature of all polygons in two periods.Referring to attached drawing 2 It is shown, original polygon element is divided into new element in figure layer lamination process, new element combines original two datasets Attribute, result be usually exactly a polygon by another polygon spatial distribution state carry out intersection operation, from And multiple polygons are divided into, while copying the attribute of input object in new object attribute list in attribute assigning process, The purpose for the arrangement is that geometrical characteristic (woods window edge) and attributive character (woods window, non-woods window) in woods window spatial distribution is fully retained Change information；

Step 33: the shared segmental arc of the woods window vector polygon after overlapping being encoded, to the woods window vector after overlapping The shared segmental arc of polygon is encoded to BB, BI, IB, BN, NB, and wherein B indicates the boundary of polygon, and I indicates the inside of polygon, N indicates non-boundary or inside；Then single woods window polygon attribute feature t is calculated_i,t_jIt is superimposed area and original forest gap area Ratio, the specific formula for calculation of area ratio are as follows:

P(EE_ij)=Area (EE_ij)/Area(E_i),

P(EO_ij)=Area (EO_ij)/Area(E_i),

P(OE_ij)=Area (OE_ij)/Area(E_i)；

In formula, P indicates that the ratio of area, Area indicate woods window range or area,

E_i: t_iPeriod, the polygon range of single woods window,

EE_ij: assuming that some woods window t_iPeriod exists, t_jPeriod, there is also EE indicated t_i~t_jDuring woods window overlapping portion Point,

OE_ij: assuming that t_iPeriod, some woods window did not occurred, in t_jPeriod has occurred, and OE indicates t_i~t_jDuring woods window change Change range,

EO_ij, it is assumed that t_iSome woods window of period exists, in t_jPeriod has disappeared, and EO indicates t_i~t_jDuring woods window disappearance Range.

Step 34: constructing woods using the step 32 and the lamination process in step 33, shared segmental arc and area ratio The discriminant criterion of window variation, referring to shown in the following table 1, monitoring the newly-generated, stable of woods window, extension, atrophy, displacement or closure etc. become Change state and quantify its variation degree.

1 woods window dynamic change topological analysis table of table

Illustrate application of the invention below by way of specific example:

Study area's overview:

Research ground is located at Fu Shou mountain forest (28 ° 3 ' 00 " -28 ° 32 ' 30 " N, 113 ° 41 ' 15 "-of Northeast of Hunan 113 ° 45 ' 00 " E), it is located in Luoxiao mountain range and connects Yunshan Mountain offshoot, topography is high in the south and low in the north, more than 1200 rice of mean sea level, and mean inclination is 22-27 degree, are presented the landforms of hills and mountains overlapping.12.1 DEG C of average temperature of the whole year, Nian Zhao 1500 hours, frost-free period 217 days.Mainly Vegetation pattern is typical Mid-subtropical Evergreen Broadleaved Forests, upper layer trees tree species mainly have China fir, Pinus taiwanesis, Qinggang oak, bitter sweet oak, Sassafrases, Alnus Trabeculosa, hickory nut and Fagaceae.

Woods window investigation on the spot uses hemisphere face image method, determines gap size according to fish eye lens projection theory；Using angle Rule method or telescopic height finder measurement woods window edge wood height；Differential GPS or total station survey woods window center height above sea level and margin location It sets.

Field investigation method:

The sample prescription of 80 30m × 30m is set in permanent sample plot, the position of each sample prescription is determined with differential GPS, each Hemisphere face woods is obtained with the external fish eye lens of digital camera (wide-angle is 183 °, orthographic projection) at sample prescription center and diagonal line quartile It is preced with image, image direction is overlapped with magnetic north direction.With Gap Light Analyzer (GLA, V2.0, image processing software) to hat Layer photo is analyzed and draws out woods window edge.Investigation work selects fine shape respectively summer in 2016 and 2009 It is carried out under condition.40 are used as training sample data in 80 sample prescription data, remaining 40 as verifying sample data.

Airborne laser radar data:

t₁,t₂Laser radar data acquisition time is in June, 2009 and in September, 2016, t respectively₁Airborne lidar system For LMS-Q560, laser beam flying angle is 22.5 ° average, average spot size 50cm, and point cloud density is 2~6/m²。t₂It is airborne Laser scanning system is ALTM2050, and laser beam flying angle is 15 ° average, average spot size 25cm, and point cloud density is 2~10 A/m².Every Shu Jiguang includes the information such as the coordinate value, height value, intensity value of the first echo and last echo.Laser radar point Cloud data all use LAS format, and the geographical co-ordinate system using ENVI software set laser point cloud data is CGCS2000, reference Ellipsoid is WGS84, and projection code name is 38, and laser radar data is with a cloud storage format record (referring to shown in the following table 2).

2 laser point cloud storage format schematic table of table (each laser point corresponds to a line record)

Change in time and space monitoring and quantizing process are carried out to research area woods window based on the method for the present invention:

(1) the woods window monitoring based on laser radar point cloud data

Using adaptive irregular triangle network filtering method according to height value by t₁Airborne laser radar point cloud data divides For canopy point cloud and Ground Point cloud, and raster interpolation is carried out respectively and generates DSM and DEM, set the raster resolution of DSM and DEM It is all 2 meters.Grid difference operation is carried out to DSM and DEM using ENVI software, generates the canopy height model of 2 meters of resolution ratio CHM, referring to shown in attached drawing 3.

The differentiation of woods window is carried out on CHM using woods window identification model (formula 1), sets woods window edge canopy height threshold value a It is 5 meters, referring to shown in attached drawing 4, on CHM sectional view, vegetation height is woods window less than 5 meters, otherwise is crown canopy (non-woods window). Using the binary map of ENVI Software Create G (x, y) woods window and Fei Lin window range and morphologic filtering operation is carried out, using " corrosion " " expansion " algorithm eliminates the small―gap suture (within 2 meters) between forest or small " cavity " or grating image noise in crown canopy, really Fixed final woods window range carries out " grid conversion vector " operation on this basis, draws and save woods referring to shown in attached drawing 5 Window edge polygon vector data G₂₀₀₉。

t₂Airborne laser radar data operating process is such as above-mentioned t₁The edge polygon arrow of woods window is drawn and is saved in operation Measure data G₂₀₁₆。

(2) quantization of the woods window dynamic mode based on spatial analysis

Using the ArcCatalog module creation geographical data bank in ArcGIS software, polygon element collection, selection are created CGCS2000 geodetic coordinate system, respectively by G₂₀₀₉And G₂₀₁₆Woods window polygon vector data is imported into geographical data bank, creation The topological structure of vector data, setting XY tolerance are 0.1m.To G₂₀₀₉And G₂₀₁₆Polygon element carries out Overlaying analysis, specifically Operation is " joint ", according to the logic mathematical model of woods window variation namely formula (3)~(9) and the signal addition of 1 content of table Topology rule, output factor kind is by the institute of the polygon comprising the geometry union that represents all inputs and all input factor kinds There is field, topological relation verifying then is carried out to overlapping result, according to spatial analysis as a result, successively selecting to meet the variation of woods window The result of condition count and drawing result.

(3) interpretation of result

Expert data statistical analysis software Statistical Product and Service Solutions (SPSS, V19) to t₁And t₂The woods window edge and field investigation result (verifying sample) of laser radar data monitoring carry out linear fit precision Analysis, the difference for comparing context of methods and field investigation using paired-samples T-test (paired t-test), referring to shown in the following table 3.

The method of 3 this programme of table and the comparison sheet of field investigation result

X is field investigation value, and y is the method estimated value of this method.

Woods window edge position difference normal distribution conspicuousness Sig. >=0.05 illustrates there is statistics using paired-samples T-test method Learn meaning.The null hypothesis of paired-samples T-test is the t distribution that the distribution of position difference meets that average value is 0, t₁And t₂The bilateral of data Conspicuousness P is both greater than 0.05, and there is no significant differences for the woods window position and field investigation result for illustrating Monitoring by Lidar.From phase From the point of view of root-mean-square error, the monitoring result error of woods window position is smaller, and higher with the results relevance of field investigation.We The woods window position and distribution situation precision with higher that method measures, can be used for rapid survey different size, woods of different shapes Window.

t₁The woods window density that airborne laser radar data monitors is 11.67/hm^-2, forest gap area mean value is 81.02m², minimum 4.06m², maximum 727.43m², forest gap area integrated distribution is in interquartile-range IQR, the degree of bias of feature distribution It is 2.542, illustrates that it with positive deviation, that is, is mostly the woods window compared with small area, referring to shown in the following table 4.t₂Woods window also has similar Situation, density are 12.81/hm^-2, also based on small area woods window.

4 forest gap area (m of table²) essential characteristic statistical form

According to the logic mathematical model of variable condition, to t₁-t₂The variation of woods window is quantified, and ginseng is shown in Table 5, newly-generated Woods window is with 5~50m²The overwhelming majority is accounted for, conflicting mode mainly rolls over branch；Due to the lateral update and vertical update of woods window, small grade Other woods window is easier to be closed completely；Under " stabilization " state, it is greater than 300m²Great Lin window ratio it is maximum, " extension non-displacement " It is equally based on great Lin window under state, the woods window ratio of remaining rank is not much different, and main cause may be raw in great Lin window Border deteriorates, and part edge wood generates new confusion area, so that forest gap area becomes larger, referring to shown in attached drawing 6；" atrophy non-displacement " State 150m²Based on following woods window, reason may be that woods window laterally updates and accounts for leading role, referring to shown in attached drawing 7；" extension has Displacement " and " atrophy has displacement " two kinds of variable conditions, all ratio, main cause have been the updates of woods window edge wood to woods windows at different levels With interference and meanwhile act on, eventually led to the changes in spatial distribution of woods window edge.It is dry that above-mentioned change procedure not only demonstrates woods window The ecological process disturbed, has been completed at the same time the quantization work of various change state, provides accurate prison for forest ecology research Measured data provides information-based support and skill further to disclose the mechanism that the dynamic law of forest cycle is maintained with bio-diversity Art guarantee.

5 woods window dynamic change ration statistics table of table

Finally it should be noted that above embodiments are merely to illustrate the technical solution of the application rather than to its protection scope Limitation, although the application is described in detail referring to above-described embodiment, the those of ordinary skill in the field should Understand: those skilled in the art read the specific embodiment of application can still be carried out after the application various changes, modification or Equivalent replacement, but the above change, modification or equivalent replacement, in the application wait authorize or the claim of issued for approval protection model Within enclosing.