Video monitoring vision field analysis method based on improved PDERL algorithm

Abstract

The invention discloses a PDERL-based video monitoring vision field analysis method, and relates to the field of GIS space analysis. The method comprises the following steps: acquiring DSM topographic data and video monitoring parameter information; on the basis of information such as a video monitoring field angle and a sight distance, a partition strategy is utilized to realize that long-distance grid points on the same sight line can utilize short-distance grid point information in any field angle range; different strategies are adopted to optimize the initial reference line according to whether the observation point is located at the grid point or not; for each analysis sub-region, determining starting and ending coordinates of transverse / longitudinal grid lines of the analysis region according to a segmentation line formed by a field angle, and judging a relationship between a starting segmentation line and a boundary region value, so as to realize optimization of grid point search; and if a plurality of areas exist, merging visual results of the areas, generating a corresponding grid visibility file, and visualizing the results in the GIS. According to the method, the effective balance between the speed and the precision is realized in the field of vision field analysis, and the method has an important reference value for a vision field analysis scene which needs quick response and has a relatively high requirement on the precision.

Description

Technical Field

[0001] The present invention relates to the field of GIS spatial analysis, and in particular to a video surveillance view analysis method based on an improved PDERL algorithm, which is mainly applicable to view analysis of video surveillance and realizes visualization. Background Art

[0002] With the rapid development of society and economy, public safety issues have received more and more attention. As an extension of human visual range, video surveillance is an important technical means to ensure urban safety and is widely used in crowded public places and national security areas. However, due to the complex and changeable terrain, it is impossible to fully evaluate the visual range based only on the camera's viewing distance, field of view and other parameters. It is necessary to comprehensively consider factors such as actual terrain, building entities, etc. for analysis. View analysis systematically examines the scope of vision to achieve a comprehensive assessment of a specific area. It has a wide range of applications in various fields: such as visual quality assessment, territorial planning, site selection, spatial layout, coverage assessment, archaeological site protection, etc.

[0003] In recent years, as the spatial resolution and elevation values ​​of DEM have become increasingly refined, traditional viewshed analysis algorithms cannot fully utilize DEM data information, and the errors of the algorithms themselves have become increasingly prominent. In most scenarios, it is necessary to ensure the accuracy of the algorithm while improving the computational efficiency, which is of great significance for the widespread application of viewshed algorithms and the improvement of their actual performance. The PDERL algorithm is a fast approximate viewshed algorithm that proposes a PDE spatial reference system. It constructs and updates reference lines to judge the visibility of the survey line, achieving a good balance between accuracy and speed. The XPDERL algorithm changes the partitioning strategy so that the observation points can make the most of the grid line information in each region. Experiments have shown that the XPDERL algorithm can improve its efficiency to a level close to that of the approximate algorithm at the expense of some accuracy.

[0004] The above inventions all regard the visual range as a circular visual range, but video surveillance is a fan-shaped visual range analysis. In addition to the video surveillance latitude and longitude coordinates, viewing distance, and elevation, the field of view angle is also required to determine the analysis range. The existing algorithms are not suitable for video surveillance visual range analysis. To solve the above problems, the present invention proposes a video surveillance fast approximate visual range analysis SPDERL algorithm based on the PDERL algorithm and the XPDERL partitioning strategy idea to ensure its practicality and effectiveness in video surveillance visual range analysis, which has important reference value for visual range analysis scenarios that require fast response and have high requirements for accuracy. Summary of the invention

[0005] The purpose of the present invention is to provide a video surveillance view analysis method based on an improved PDERL algorithm to achieve fast approximate view analysis of video surveillance.

[0006] To achieve the above object, the present invention provides the following steps:

[0007] (1) Obtain DSM data and video surveillance parameters under three types of terrain, and establish a two-dimensional array of visible range based on the parameters;

[0008] (2) Determine the partitioning strategy based on the field of view and visible area;

[0009] (3) Optimize the initial reference line by selecting different strategies based on the location of the observation points;

[0010] (4) Determine the start and end boundary coordinates within the visible range based on the segmentation line equation and the grid coordinates of the video monitoring points;

[0011] (5) Optimization of grid point search within the visible area;

[0012] (6) Merge the two-dimensional visual arrays of each sub-region and visualize the results.

[0013] Furthermore, step (1) specifically includes the following steps:

[0014] Read the DSM (Digital Surface Model) elevation value and remove the abnormal value, and store it in the two-dimensional array H. Then, according to the DSM data geographic transformation parameters (upper left corner x coordinate, horizontal resolution, rotation parameter, upper left corner y coordinate, rotation parameter, vertical resolution), the DSM elevation array is uniformly stored with the lower left corner as the origin. The starting longitude and latitude of the terrain file is (lon_s, lat_s), and the horizontal and vertical grid point spacing is d x ,d y ;

[0015] Read the parameters in the monitoring attribute data table, including the line of sight r, the latitude and longitude of the monitoring point (lon c ,lat c ), scanning start angle α, scanning end angle β and elevation value H a , convert the video surveillance latitude and longitude into the grid point coordinate value P in combination with the geographic conversion parameters of the DSM file c (x c ,y c );

[0016] Based on the monitoring attribute parameters, a sector area with the monitoring point as the center and the viewing distance r as the radius is first determined. The angular range of this area is determined by the scanning start angle α and the scanning end angle β. The two boundary dividing lines L1 and L2 correspond to the monitoring point P respectively. c And the straight lines with slopes of tanα and tanβ are shown in equation (1) and equation (2):

[0017] L1:y1=tanα*(xxc )+y c (1)

[0018] L2:y2=tanβ*(xx c )+y c (2)

[0019] Furthermore, step (2) specifically includes the following steps:

[0020] Specific zoning strategies are used for different area ranges. The minimum angle is 0 degrees, indicating the north direction, and the angle increases clockwise to a maximum of 360 degrees. The angle range is calculated based on the input starting and ending angles and is divided into four categories: [0,90], [90,180], [180,270], [270,360], and [0,360]. The four categories are further refined as shown in Table 1:

[0021] Table 1 Specific partitioning strategy

[0022]

[0023] Furthermore, step (3) specifically includes the following steps:

[0024] If the observation point is located on the grid point, equations (3) and (4) are satisfied:

[0025] (x c -lon_s)%d x =0(3)

[0026] (y c -lat_s)%d y =0(4)

[0027] Then construct the grid points (P0,…,P c-1 ,P c ,P c+1 ,…,P c+k ,…P n ) line of sight, and calculate the elevation value H of each line of sight and the grid intersection point by linear interpolation ci , H ci The actual elevation value H of each grid point ai Compare and judge its visibility; if P c To P i Construct a sight line if and only if on the same sight line (P0,…,P i-1 ) are visible and H ci ≥H ai Then P iIf the grid points are in the same row / column as the observation point, they are visible, otherwise they are not visible. Since these grid points are in the same row / column as the observation point and will not block the subsequent rows / columns, the next row / column is considered visible and the initial reference line RL is constructed. When the observation point is not on the grid point, the visibility of the grid points in the row / column closest to the observation point is accurately calculated through linear interpolation, and the initial reference line is constructed.

[0028] The construction of the reference line first considers the grid points of the first row / column of the observation point analysis as visible, and converts the horizontal coordinate x, vertical coordinate y, and elevation h of each grid point in the geocentric coordinate system into the proximity p, direction d, and elevation e in the PDE space; the conversion formula is as follows:

[0029]

[0030]

[0031]

[0032] Adjacent grid points are converted into continuous polylines in PDE space. The PDERL algorithm retains the slope a and the d value at the end of each polyline to store the information of close-range grid points, where LR records the information of close-range grid points and LC represents the grid point information of the current analysis row / column. The visibility of LC is to find the d value interval [d0,…d c-1 ,d c ,d c+1 ,d n ] corresponds to the broken line, and based on a and d, the value of e is calculated and the value of e is c The value of e that corresponds to the largest value of d max The values ​​are compared to determine visibility and updated. Symbol (8) indicates visibility, otherwise it is not visible:

[0033] e c -e max ≥0(8)

[0034] Furthermore, step (4) specifically includes the following steps:

[0035] First, according to the grid coordinates of the first column / row analyzed by the video surveillance point, the starting and ending coordinates are determined in combination with the dividing line equation; secondly, the grid diagonal and the dividing line are combined to calculate the adjustment value required for the start and end coordinates of moving one row / column outward compared to the previous row / column. The starting adjustment value is d s , the final adjustment value is d e ; The fixed change of each column / row coordinate is 1 and L1 and L2 are both straight lines, then the change value of y is shown in formula (9):

[0036] Δy=kΔx(9)

[0037] Where k represents the slope of the dividing line equation;

[0038] Therefore, d s With d e The calculation formulas are shown in equations (10) and (11):

[0039] d s =tanβ(10)

[0040] d e =tanα(11)

[0041] When the segmentation line of the monitoring point is completely aligned with the diagonal line of the grid, there is no need to modify the adjustment value. If the segmentation line shows an increasing trend, the adjustment value should be rounded up, as shown in formula (12):

[0042]

[0043] Among them, ys i Represents the starting grid y coordinate of the current analysis column, ys i-1 Represents the starting grid y coordinate of the previous column;

[0044] For the dividing line with a decreasing trend, its adjustment value should be rounded down, as shown in formula (13):

[0045]

[0046] Among them ye i Represents the y coordinate of the ending grid of the current analysis column, ye i-1 Represents the ending grid y coordinate of the previous column.

[0047] Furthermore, step (5) specifically includes the following steps:

[0048] In order to ensure that the coordinates of the row or column after adding the adjustment value will not exceed the boundary of the visual analysis area, the present invention introduces a callback strategy. When the coordinates of a row or column exceed the visual analysis area due to adjustment, the coordinates are adjusted to the boundary value of the visual analysis area. This strategy ensures that the analysis area represented by the last column or row will not be lost due to the adjustment value;

[0049] The grid position of the segmentation line in the starting direction is compared with the boundary area. If the starting grid position exceeds the boundary area, the analysis is terminated to reduce the invalid search of non-target areas and improve the analysis efficiency.

[0050] Furthermore, step (6) specifically includes the following steps:

[0051] If there are multiple regions, the visual results of each region are merged, and finally the visibility of the region is maintained through a two-dimensional visual array. Based on the visual array, a raster visibility file is established, and the video monitoring points are saved as shapefile files with latitude and longitude, viewing distance, starting angle, ending angle, and viewpoint height attributes, and the results are visualized in GIS.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention proposes a partitioning strategy to realize that the visible area formed by any field of view angle can be analyzed by using the SPDERL algorithm for visibility, and solve the problems of misjudgment of visibility of edge points and search redundancy caused by the partitioning strategy. By optimizing the construction of the initial reference line, a large number of errors are avoided when the observation point is located at the grid point. Finally, the field of view angle of the physical characteristic of video surveillance and the PDERL algorithm are combined through the partitioning strategy to realize the fast approximate field of view analysis of video surveillance. The algorithm of the present invention is for video surveillance and other equipment whose analysis range needs to change dynamically according to physical characteristics. It realizes an effective balance between speed and accuracy in the field of field of view analysis. At the same time, it has important reference value for field of view analysis scenarios that require fast response and have high requirements for accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 It is the overall technical flow chart of the method of the present invention;

[0056] Figure 2 This is a DEM 3D terrain diagram;

[0057] Figure 3 This is a schematic diagram of the video surveillance analysis area;

[0058] Figure 4 It is a schematic diagram of the partition strategy on the lower right;

[0059] Figure 5 It is a schematic diagram of the right partitioning strategy and the left and right partitioning strategy;

[0060] Figure 6 Optimize the schematic for the initial reference line;

[0061] Figure 7 Schematic diagram of converting geocentric coordinate system to PDE coordinate system;

[0062] Figure 8It is a schematic diagram of the start and end boundary indexes and the extended values;

[0063] Fig. 9 This is a schematic diagram of the start and end coordinate callback strategy;

[0064] Fig.10 Optimize the schematic diagram for searching the termination condition;

[0065] Fig.11 Schematic diagram for visualization of the results;

[0066] Fig.12 This is the algorithm speed comparison result diagram;

[0067] Fig.13 This is a comparison chart of algorithm accuracy;

[0068] Fig.14 Mean values ​​of the aggregated error points of each algorithm under different terrains and viewpoint heights. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0070] The purpose of the present invention is to provide a video surveillance view analysis method based on an improved PDERL algorithm to achieve fast approximate view analysis of video surveillance.

[0071] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0072] The invention provides a video surveillance view analysis method based on an improved PDERL algorithm. Figure 1 The flow chart of the visual field analysis provided by the present invention comprises:

[0073] like Figure 1 The flowchart of the improved algorithm of video surveillance view analysis based on PDERL is shown, including:

[0074] Step S1: Read the DEM data downloaded from the Copernicus official website and create a video surveillance point feature in its coverage area, which contains attributes such as field of view angle and viewing distance.

[0075] Step S1 specifically includes:

[0076] Step S11: Obtain terrain file data and perform data processing

[0077] In this step, the present invention uses three DEM data, corresponding to three types of terrain: plain, hilly, and mountainous; each DEM data is composed of 3600*3600 grid points. The elevation value is obtained by the corresponding grid point coordinates, and it is determined whether there are abnormal values ​​(less than 0 or greater than the maximum height) and the abnormal values ​​are processed. The final result is a 3D terrain map as shown in the figure. Figure 2 shown.

[0078] Step S12: Input the video monitoring point parameter information and calculate the visible area

[0079] In this step, the present invention obtains the video monitoring point parameter information (viewing distance r, longitude and latitude, scanning start angle α, scanning end angle β, elevation value H a ), and construct the visible area according to its viewing distance and field of view. Figure 3 As shown in the figure, the video surveillance latitude and longitude are combined with the geographic conversion parameters of the DSM file to convert the grid point coordinate values ​​(x c ,y c ), and determine the rectangular boundary area composed of min_x, max_x, min_y, max_y with the viewing distance r. The boundary dividing line equations L1 and L2 of the visible area can be obtained by (x c ,y c ), α, β are determined, and the actual analysis area is composed of the part of the region where the angle range formed by β-α intersects with the boundary area.

[0080] Step S2: Adopt a specific partitioning strategy based on the angle range and analysis area, including:

[0081] like Figure 4 As shown in the figure, when the starting angle α∈[0,90), the ending angle β∈[180,270), and the angle range γ∈[180,270), the analysis range is divided into Region I on the lower side of the video surveillance and Region II on the right. The analysis direction of Region I is from west to east and from north to south; the analysis direction of Region II is from south to north and from west to east. Taking Region I as an example, from west to east represents the search direction for the current row, and from north to south represents the overall search direction of Region I. This division strategy allows each partition to utilize vertical grid line or horizontal grid line information. Other situations are as follows Figure 5 shown.

[0082] Step S3: Different strategies are adopted to construct the initial reference line according to whether the observation point is located at a grid point, including:

[0083] Step S31: determine whether the observation point is located at a grid point and select different initial reference line construction strategies.

[0084] like Figure 6 As shown, if the observation point is located on the grid point, equations (1) and (2) are satisfied:

[0085] (x c -lon_s)%d x =0(1)

[0086] (y c -lat_s)%d y =0(2)

[0087] Then construct the grid points (P0,…,P c-1 ,P c ,P c+1 ,…,P c+k ,…P n ) line of sight, and calculate the elevation value H of each line of sight and the grid intersection point by linear interpolation ci , H ci The actual elevation value H of each grid point ai Compare and judge its visibility; if P c To P i Construct a sight line if and only if on the same sight line (P0,…,P i-1 ) are visible and H ci ≥H ai Then P i Visible, otherwise invisible; since these grid points are in the same row / column as the observation point and will not block the subsequent rows / columns, the next row / column is considered visible and the initial reference line RL is constructed;

[0088] When the observation point is not on a grid point, the visibility of the nearest row / column grid point to the observation point is accurately calculated through linear interpolation, and the initial reference line is constructed.

[0089] Step S32: Initial reference line construction.

[0090] The initial reference line construction first considers the grid points in the first row / column of the observation point analysis as visible, and converts the horizontal coordinate x, vertical coordinate y, and elevation h of each grid point in the geocentric coordinate system into the proximity p, direction d, and elevation e in the PDE space; the conversion formula is as shown in Equation (3), Equation (4), and Equation (5):

[0091]

[0092]

[0093]

[0094] like Figure 7As shown in the figure, adjacent grid points are converted into continuous polylines in the PDE space. The PDERL algorithm retains the slope a and the d value at the end of each polyline to store the information of close-range grid points, where LR records the information of close-range grid points and LC represents the grid point information of the current analysis row / column. The visibility of LC is to find the d value interval [d0,…d c-1 ,d c ,d c+1 ,d n ] corresponds to the broken line, and based on a and d, the value of e is calculated and the value of e is c The value of e that corresponds to the largest value of d max The values ​​are compared to determine visibility and updated. If equation (6) is satisfied, it means it is visible, otherwise it is not visible:

[0095] e c -e max ≥0(6)

[0096] Step S4: Determine the start and end boundary coordinates within the visible range according to the segmentation line equation and the grid coordinates of the video monitoring points, specifically including:

[0097] Step S41: Determine the start and end dividing line equations according to the video monitoring point coordinates and the start and end dividing line slopes.

[0098] like Figure 8 As shown, the video surveillance grid point coordinates are (x c ,y c ), the video surveillance point in the figure below is not on the grid point, and the elevation value cannot be directly obtained by accessing the array through the coordinates. In this case, the grid point closest to the point is used as the center coordinate for analysis and is recorded as (x1, y1). The starting column coordinate of the first column is start_y, the ending column coordinate is end_y, and the column start adjustment value is d s , the column end adjustment value is d e The equation of the starting dividing line is shown in formula (7):

[0099] L2:y2=tanβ*(xx c )+y c (7)

[0100] The termination dividing line equation is shown in formula (8):

[0101] L1:y1=tanα*(xx c )+y c (8)

[0102] Step S42: The visible area determines the search direction according to the partition strategy table.

[0103] The visual analysis area is the right side of the video surveillance, from south to north and from west to east. This side analysis is to increase the x coordinate from the right side of the video surveillance point column by column. The horizontal coordinate of the starting column is x1+1. Substituting the horizontal coordinate into L1 and L2 can obtain the start and end coordinates. The start and end coordinates of the subsequent columns do not need to be calculated in this way. The x coordinate of each column is fixedly increased by 1 and L1 and L2 are both straight lines. The change value of y is shown in formula (9):

[0104] Δy=kΔx (9)

[0105] Where k represents the slope of the dividing line equation.

[0106] The corresponding adjustment value of L1 and L2 is d e d s , the calculation formula is shown in formula (10) and formula (11):

[0107] d s =tanβ(10)

[0108] d e =tanα(11)

[0109] Step S43: start and end grid coordinate adjustment strategy.

[0110] After adding ds and de, the start and end grid coordinates of each column are not in the grid point, so they need to be rounded. In order to make the near-layer grid point information available to the far-layer grid point information, de along the positive direction of the y-axis is rounded up, and ds along the negative direction of the y-axis is rounded down. Therefore, the calculation of the start and end grid coordinates of the subsequent columns is shown in equations (12) and (13):

[0111]

[0112]

[0113] Among them, ys i Represents the starting grid y coordinate of the current analysis column, ys i-1 Represents the starting grid y coordinate of the previous column; i Represents the y coordinate of the ending grid of the current analysis column, ye i-1 Represents the ending grid y coordinate of the previous column.

[0114] Step S5: Optimizing the search for grid points in the visible area, specifically including:

[0115] Step S51: To ensure that the coordinates of the row or column after adding the adjustment value do not exceed the boundary of the visual analysis area, the present invention introduces a callback strategy. When the coordinates of a row or column exceed the visual analysis area due to adjustment, the coordinates are adjusted to the boundary value of the visual analysis area. This strategy ensures that the analysis area represented by the last column or row will not be lost due to the adjustment value.

[0116] As shown Fig. 9 in the figure, the analysis area is on the south side, using the horizontal grid line information. The coordinates of the first grid point in the starting row are (s x , s y ), and the adjustment value of the starting dividing line is d s . It can be seen from the figure that the starting coordinates of the second row are s x - d s , and this value is less than the boundary min_x of the visible area. If the starting coordinates of the subsequent rows are assigned as s x , and the actual analysis area is RegionⅡ, the visibility of the grid points in RegionⅠ will be lost. To prevent the loss of the visibility information of the grid points in RegionⅠ, the coordinate callback strategy is used to adjust the coordinate values: the starting coordinates of the subsequent rows are all assigned as the left boundary value of the visible area, that is, min_x, so as to ensure that the visibility information of the grid points in RegionⅠ and RegionⅡ can be determined.

[0117] Step S52: To optimize the grid point search, by comparing the grid position where the dividing line is located in the starting direction with the boundary area, if the starting grid position exceeds the boundary area, the analysis is terminated, so as to reduce the ineffective search for non-target areas and improve the analysis efficiency.

[0118] As shown Fig.10 in the figure, the analysis area is located on the southwest side of the video monitoring point, the row search direction is from north to south, and the overall search direction is from east to west. The starting coordinates of the first column are (s x , s y ), and the PDERL algorithm will analyze from the column where s x is located to the column where min_x is located, but it will analyze the grid points in the non-target area. Therefore, a new restriction is added to the search condition. The starting coordinates corresponding to L1 are recorded as start_y. If it exceeds the boundary area min_y, that is, start_y < min_y, the search is terminated.

[0119] Step S6: Combine the two-dimensional visible arrays of each sub-region and visualize the results

[0120] Combine the visible results of each region (if there are multiple regions). Finally, the visibility of the region is maintained by a two-dimensional visible array. This array is centered on the video monitoring point and has no longitude and latitude information. Therefore, it is necessary to find out the longitude and latitude information of the starting point of this array, establish the corresponding grid visibility file, and save the video monitoring point as a shapefile file with attributes such as longitude and latitude, viewing distance, starting angle, ending angle, and viewpoint height, and visualize the results in GIS, as Fig.11 shown in the figure.

[0121] To better implement the effects of the present invention, the following experiments are carried out:

[0122] The algorithm of the present invention is implemented in Python language. The experimental area is three DEMs (N34E114, N41E119 and N28E097) in Copernicus DEM-, and its DSM resolution is 30m. The CPU of the experimental test computer is Intel(R) Core(TM) i7-10875HCPU@2.30GHz, memory is 16G, and the operating system is 64bitWindows10.

[0123] 1. Algorithm speed comparison

[0124] The SPDERL algorithm of the present invention is compared with the Xdraw approximate algorithm in speed. The fixed radius is 10000m. The latitude and longitude of the observation point, the starting angle, and the ending angle are randomly selected in three types of terrain. The starting observation height is 0m and the ending height is 5000m. The experiment is repeated 50 times with an interval of 50m each time to calculate the average time consumption of the present algorithm / the time consumption of the Xdraw algorithm. The experimental results are shown in Fig.12 shown.

[0125] As can be seen from the figure, no matter in which terrain area, the time used by the algorithm of the present invention is less than half of the time of the Xdraw algorithm. In mountainous terrain, as the viewpoint height increases, the time consumed by the SPDERL algorithm also increases; in plain and hilly terrain, when the viewpoint height is close to the ground, the average time consumption ratio of the algorithm of the present invention and the Xdraw algorithm is less than 0.2, and the average time consumption ratio fluctuates around 0.35 as the viewpoint height increases.

[0126] 2. Algorithm Accuracy Comparison

[0127] This experiment compares the accuracy of the algorithm of the present invention with that of the R3 algorithm. In three types of terrain, random positions, random radii, and random longitude and latitude were selected. The initial observation height was 0, and the experiment was repeated 50 times, with the height increasing by 50m until it reached 5000. The experimental results are shown in Figure 2. Fig.13 shown.

[0128] As can be seen from the figure, the overall average error rate of the algorithm of the present invention is less than 5%, and the overall trend is that the error rate gradually decreases with the increase of the viewpoint height. Among them, the error rate in the plain area is relatively high in the low viewpoint height area, and the average error rate fluctuates around 0.005 with the increase of the viewpoint height. After 3000 meters, the average error rate stabilizes below 0.005; the error rate in mountainous terrain reaches a peak at an altitude of about 1000 meters and about 2000 meters, and the overall error rate at other heights is relatively low and stable; the error rate in hilly terrain is the highest in the first 1000 meters of viewpoint height, and the error rate in other height areas is relatively low and stable.

[0129] 3. Error point aggregation

[0130] Taking the PDERL algorithm as a benchmark, the algorithm SPDERL proposed in the present invention is compared with the XDraw algorithm and the XPDERL algorithm for error point aggregation. The present invention uses the x-NAE indicator based on the x-neighborhood area error (x-NE) for verification. The x-NAE refers to the area where 90% of the grid points within the x grid distance around the central grid point and its surroundings have visible errors. The degree of error point aggregation is measured by counting the x-NAE occurrence ratio (x-NAER). The x-neighborhood area error rate (X-NAER) is calculated by dividing the number of occurrences of the X-neighborhood area error (X-NAE) by the total number of grid points in the analysis area. This experiment uses PDERL, XPDERL, XDraw and SPDERL to select fixed radius, random field of view angle, and random viewpoint position, and conducts 1000 experiments on three DEMs from 3m to 75m above the ground with an interval of 3m each time. Fig.14 And Table 2 shows the experimental results:

[0131] Table 2. The sum of the mean errors of the clustered points in each neighborhood, and the ratio of the SPDERL algorithm in this paper to the XPDERL algorithm.

[0132]

[0133] Overall, among the three types of terrain, the error aggregation in the hilly area is significantly higher than that in the plain and mountainous areas. This may be due to the complexity of the hilly terrain, which causes the viewshed analysis errors to be more concentrated. The error aggregation in the plain area is the lowest, which may be because the terrain is relatively flat and the line of sight is less obstructed. At the same time, with the increase of the neighborhood range, the error aggregation ratio in all terrain types shows a downward trend, indicating that the SPDERL algorithm is more sensitive to neighborhood changes than the XPEDERL algorithm, and its error point aggregation degree is also more relaxed. Specifically, in the mountainous terrain, the average value of the highest aggregation error point of the SPDERL algorithm of the present invention is 0.004, which is much smaller than 0.035 of the XPDERL algorithm and 0.06 of the XDraw algorithm, and the overall error value of the present algorithm is also lower than that of the XPDERL algorithm and much lower than that of the XDraw algorithm. In the plain terrain, the degree of aggregation of the error points of the three algorithms is less than that of the other two terrains. The average value of the highest error point of the algorithm of the present invention is only 0.000102, and there is no error point aggregation problem after the 2-neighborhood. In hilly terrain, the error point aggregation degree of the three algorithms is the highest among the three terrains, among which the highest error point average of the algorithm of the present invention reaches 0.0041, but it is more relaxed than the XPDERL algorithm and the XDraw algorithm. In terms of error point aggregation performance, the algorithm SPDERL proposed in the present invention significantly surpasses the XDraw algorithm and is better than the XPDERL algorithm.

[0134] Through the above experiments, the algorithm of the present invention achieves an effective balance between speed and accuracy in the field of field of view analysis for devices such as video surveillance, which require the analysis range to change dynamically according to physical properties. At the same time, it has important reference value for field of view analysis scenarios that require fast response and have high requirements on accuracy.

[0135] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and scope of application. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A video surveillance view analysis method based on an improved PDERL algorithm, characterized in that: The following steps are included: (1) Obtain DSM data and video surveillance parameters under three types of terrain, and establish a two-dimensional array of visible range based on the parameters; (2) Determine the partitioning strategy based on the field of view and visible area; (3) Optimize the initial reference line by selecting different strategies based on the location of the observation points; (4) Determine the start and end boundary coordinates within the visible range based on the segmentation line equation and the grid coordinates of the video monitoring points; (5) Optimization of grid point search within the visible area; (6) Merge the two-dimensional visual arrays of each sub-region and visualize the results.

2. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (1) specifically includes the following steps: Read the DSM (Digital Surface Model) elevation value and remove the abnormal value, and store it in the two-dimensional array H. Then, according to the DSM data geographic transformation parameters (upper left corner x coordinate, horizontal resolution, rotation parameter, upper left corner y coordinate, rotation parameter, vertical resolution), the DSM elevation array is uniformly stored with the lower left corner as the origin. The starting longitude and latitude of the terrain file is (lon_s, lat_s), and the horizontal and vertical grid point spacing is d x ,d y ; Read the parameters in the monitoring attribute data table, including the sight distance r, monitoring point P c The longitude and latitude (lon c ,lat c ), scanning start angle α, scanning end angle β and elevation value H a , convert the video surveillance latitude and longitude into grid point coordinates (x c ,y c ); Based on the monitoring attribute parameters, a sector area with the monitoring point as the center and the viewing distance r as the radius is first determined. The angular range of this area is determined by the scanning start angle α and the scanning end angle β. The two boundary dividing lines L1 and L2 correspond to the monitoring point P respectively. c And the straight lines with slopes of tanα and tanβ are shown in equation (1) and equation (2): L1:y1=tanα*(x-x c )+y c (1) L2:y2=tanβ*(x-x c )+y c (2)。 3. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (2) specifically includes the following steps: Specific zoning strategies are used for different area ranges. The minimum angle is 0 degrees, indicating the north direction, and the angle increases clockwise to a maximum of 360 degrees. The angle range is calculated based on the input starting and ending angles and is divided into four categories: [0,90], [90,180], [180,270], [270,360], and [0,360]. The four categories are further refined as shown in Table 1: Table 1 Specific partitioning strategy 4. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (3) specifically includes the following steps: If the observation point is located on the grid point, equations (3) and (4) are satisfied: (x c -lon_s)%d x =0(3) (y c -lat_s)%d y =0(4) Then construct the grid points (P0,…,P c-1 ,P c ,P c+1 ,…,P c+k ,…P n ) line of sight, and calculate the elevation value H of each line of sight and the grid intersection point by linear interpolation ci , H ci The actual elevation value H of each grid point ai Compare and judge its visibility; if P c To P i Construct a sight line if and only if on the same sight line (P0,…,P i-1 ) are visible and H ci ≥H ai Then P i If the grid points are in the same row / column as the observation point, they are visible, otherwise they are not visible. Since these grid points are in the same row / column as the observation point and will not block the subsequent rows / columns, the next row / column is considered visible and the initial reference line RL is constructed. When the observation point is not on the grid point, the visibility of the grid points in the row / column closest to the observation point is accurately calculated through linear interpolation, and the initial reference line is constructed. The construction of the reference line first considers the grid points of the first row / column of the observation point analysis as visible, and converts the horizontal coordinate x, vertical coordinate y, and elevation h of each grid point in the geocentric coordinate system into the proximity p, direction d, and elevation e in the PDE space; the conversion formula is as follows: Adjacent grid points are converted into continuous polylines in PDE space. The PDERL algorithm retains the slope a and the d value at the end of each polyline to store the information of close-range grid points, where LR records the information of close-range grid points and LC represents the grid point information of the current analysis row / column. The visibility of LC is to find the d value interval [d0,…d c-1 ,d c ,d c+1 ,d n ] corresponds to the broken line, and based on a and d, the value of e is calculated and the value of e is c The value of e that corresponds to the largest value of d max The values ​​are compared to determine visibility and updated. Symbol (8) indicates visibility, otherwise it is not visible: And c -And max ≥0(8) 。 5. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (4) specifically includes the following steps: First, according to the grid coordinates of the first column / row analyzed by the video surveillance point, the starting and ending coordinates are determined in combination with the dividing line equation; secondly, the adjustment value required for the start and end coordinates corresponding to moving one row / column outward compared to the previous row / column is calculated in combination with the position relationship between the grid diagonal and the dividing line. The starting adjustment value is d s , the final adjustment value is d e ; The fixed change of each column / row coordinate is 1 and L1 and L2 are both straight lines, then the change value of y is shown in formula (9): Δy=kΔx(9) Where k represents the slope of the dividing line equation; Therefore, d s With d e The calculation formulas are shown in equations (10) and (11): d s =tanβ(10) d e =tanα(11) When the segmentation line of the monitoring point is completely aligned with the diagonal line of the grid, there is no need to modify the adjustment value. If the segmentation line shows an increasing trend, the adjustment value should be rounded up, as shown in formula (12): Among them, ys i Represents the starting grid y coordinate of the current analysis column, ys i-1 Represents the starting grid y coordinate of the previous column; For the dividing line with a decreasing trend, its adjustment value should be rounded down, as shown in formula (13): Among them ye i Represents the y coordinate of the ending grid of the current analysis column, ye i-1 Represents the ending grid y coordinate of the previous column.

6. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (5) specifically includes the following steps: When the coordinates of a row or column exceed the visible analysis area due to adjustment, the coordinates are adjusted to the boundary value of the visible analysis area. This strategy ensures that the analysis area represented by the last column or row will not be lost due to the adjustment of the value; The grid position of the segmentation line in the starting direction is compared with the boundary area. If the starting grid position exceeds the boundary area, the analysis is terminated to reduce the invalid search of non-target areas and improve the analysis efficiency.

7. The video surveillance view analysis method based on the improved PDERL algorithm according to claim 1 is characterized in that: Step (6) specifically includes the following steps: If there are multiple sub-areas, the visual results of each area are merged, and finally the visibility of the area is maintained through a two-dimensional visual array. Based on the visual array, a raster visibility file is established, and the video monitoring points are saved as shapefile files with latitude and longitude, viewing distance, starting angle, ending angle, and viewpoint height attributes, and the results are visualized in GIS.

Images