A spatial difference evolution-based intelligent control method and device for a perception camera

Abstract

The application provides a kind of intelligent control method and device of perception camera based on spatial difference evolution, including creating scene layer to research area, generating ground target point not on obstacle based on scene layer according to preset density, selecting candidate point of camera to research area, constructing all candidate points to all ground target points of line of sight, carrying out visibility analysis to line of sight to obtain preliminary selected visible point, selecting final selected visible point as visible ground target point from preliminary selected visible point based on visible distance and angle of view of camera, setting camera control quantity, selecting best combination based on visible ground target point in all candidate points using spatial adaptive differential evolution algorithm, and taking best combination as camera control scheme of research area.The application is based on actual scene, selects sampling point at fixed interval on ground, selects camera control using spatial differential evolution algorithm, can minimize blind area, and realizes maximum monitoring coverage.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of geographic information science, and particularly relates to a method and device for intelligent control of a perception camera based on spatial differential evolution. BACKGROUND

[0002] With the rapid development of information technology and network technology, the monitoring system is no longer applied in the communication, transportation, security and other industries, and it is gradually developing towards other industries and the public, and occupies an increasingly important position in the construction of modern smart cities. The social public security situation is becoming increasingly complex, and public security problems are constantly highlighted. All of these urgently require the development of a video monitoring system that focuses on proactive prevention. Therefore, the reasonable layout and installation position of the camera are problems that we need to consider first. The traditional camera installation process is usually selected by experienced personnel based on experience, which may result in many problems, such as the existence of monitoring dead angles, and excessive overlap of camera coverage.

[0003] Under limited available resources, we always want to obtain the maximum range of actual monitoring effective area, that is, to achieve the maximum proportion of coverage of the target area using a fixed number of cameras, so as to maximize the utilization and avoid waste of resources. Although the traditional video monitoring installation method has a certain rationality, it is not a quantitative analysis of the possible monitoring dead angles and overlap of coverage. Moreover, there are many point positions that can be selected, and it is unrealistic to adjust the monitoring after installation according to the effect, which is time-consuming and laborious, so this greatly reduces the utilization rate of the camera.

[0004] The video sensor deployment optimization problem largely aims to achieve maximum coverage with the fewest possible video sensors. Previous research approaches can be broadly categorized into discretization and continuous methods. Discretization, as the name suggests, involves discretizing the target area to be covered by the cameras into a set of target points, and then using a set of candidate points for camera placement. The analysis then examines whether these target points are within the camera's coverage area, connecting them with the candidate points, and finally performing combination optimization to select the optimal combination that maximizes coverage of the target area. Key discretization-related research includes Angella et al.'s optimization using a voting strategy in 3D scenes, Conci and Zhao et al.'s optimization of camera placement in 2D scenes using binary integer programming algorithms, Gupta et al.'s use of a non-dominated sorting genetic algorithm with an elitist strategy, and Mini et al.'s use of artificial bee colony optimization and particle swarm optimization to solve sensor deployment problems, followed by heuristic scheduling. The camera perception model is typically fan-shaped. Bairagi et al. designed a II-based non-dominated sorting heuristic genetic algorithm to solve the problem of minimizing energy consumption and maximizing coverage area for 3D video sensor nodes deployed on 2D target areas.

[0005] Continuous deployment primarily treats the area to be covered by the camera and the candidate locations where the camera can be placed as a continuous space, and then adjusts the parameters of the video sensor through optimization methods. Researchers such as Ma used simulated annealing in a three-dimensional space for deployment optimization under continuous deployment conditions; while researchers such as Bouyagoub used a hill-climbing algorithm in a two-dimensional virtual space to address the problem.

[0006] Zhang Yanan summarized some common obstacles that obstruct the view, representing them with a grid and using the A* search algorithm to solve the camera deployment optimization problem. Ugur et al. identified key points in the area, classifying the camera deployment optimization problem as a 0 / 1 optimization problem, neglecting many important practical factors. This approach is clearly not realistic. ZHONG et al. used a more realistic data model compared to the methods mentioned above, using 3D models of buildings, etc. Furthermore, their imaging model did not simply use 2D sectors or circles, but rather truncated pyramids, and considered five influencing factors including architecture and landscape. However, they did not use intelligent algorithms for candidate point selection, but rather a more traditional exhaustive method. The drawback of this method is obvious: it is impractical when there are many candidate points and combinations.

[0007] The problem of camera deployment is essentially to minimize the unmonitored area. In addition to reasonable deployment, dead angles can be supplemented by other means. Milosavljevic et al. use augmented reality (AR) and the integration of three-dimensional (3D) geographic information systems (GIS) and video surveillance systems.

[0008] In summary, from the above analysis, it can be seen that the camera deployment planning problem has been valued by the relevant field, and the relevant theory and method research has also made corresponding progress, but there are still some problems that cannot be ignored. For example, in recent years, most researchers focus on the parameters of the camera and other factors, and pay insufficient attention to the joint operation problem of multiple cameras. The actual camera and geographic scene coupling and the application in the actual geographic environment are not enough. The actual geographic data is often simply regarded as a two-dimensional geometric scene, and there is basically no simulation based on a real three-dimensional scene. The obstacles are also simulated as relatively simple squares. In general, further development is still needed, and the current research still lacks specificity and practicality. SUMMARY

[0009] According to the deficiencies of the prior art, the purpose of the present application is to provide a spatial differential evolution-based intelligent camera deployment method and device, which is based on actual scenes, selects sampling points at fixed intervals on the ground, and uses a spatial differential evolution algorithm to calculate the camera distribution point combination that can achieve the maximum monitoring coverage range under the condition of a fixed number of cameras in the candidate points where cameras can be installed.

[0010] To solve the above technical problems, the technical scheme adopted by the present application is:

[0011] A spatial differential evolution-based intelligent camera deployment method, comprising:

[0012] Creating a scene layer for the research area;

[0013] Generating ground target points that are not on obstacles based on the scene layer according to a preset density;

[0014] Selecting candidate points for cameras in the research area;

[0015] Constructing the line of sight from all candidate points to all ground target points;

[0016] Performing a visibility analysis on the line of sight to obtain preliminary selected visible points;

[0017] Selecting final selected visible points as visible ground target points from the preliminary selected visible points based on the visible distance and viewing angle of the camera;

[0018] The number of cameras to be deployed is set, and the best combination is selected from all candidate points based on visible ground target points using a spatial adaptive differential evolution algorithm. The best combination is then used as the camera deployment scheme for the study area.

[0019] Furthermore, the method for generating ground target points that are not located on obstacles based on a preset density is as follows:

[0020] Two-dimensional data points are generated on the scene layer according to a preset density. The two-dimensional data points are then upgraded to three-dimensional data points based on the ground data of the digital surface model of the study area. The three-dimensional data points that are on obstacles due to the upgrade are removed according to the elevation value. The three-dimensional data points that need to be observed on the ground are retained as ground target points.

[0021] Furthermore, the method of selecting the optimal combination using the spatial adaptive differential evolution algorithm for all candidate points is as follows:

[0022] Construct a controlled population consisting of several chromosomes, each encoded by all candidate points;

[0023] Establish a fitness function to obtain the fitness function value of the chromosome;

[0024] Chromosomes are updated and the population is iteratively controlled through mutation operations and / or crossover selection;

[0025] The optimal chromosome is selected based on the fitness function value.

[0026] Furthermore, the method for establishing the fitness function is as follows:

[0027] The formula for calculating ground target point coverage is as follows:

[0028]

[0029] |C|=m

[0030] Among them, w i is the weight corresponding to the i-th visible ground target point, and m is the total number of all visible ground target points. It is the first The weights corresponding to each ground target point are given by p1, p2, p3, ..., pc, where n is the total number of ground target points, and C is the set obtained by combining the visible ground target points of all selected cameras. m For all visible ground target points, It is the first The set of visible ground target points corresponding to each camera, where q is the total number of cameras;

[0031] Obtain ground target point coverage redundancy The calculation formula is:

[0032]

[0033] wherein D set is a set of all visible ground target points repeatedly in two or more cameras, |D| is the number of elements in the D set, is the kth visible ground target point repeatedly in two or more cameras, for judging whether the visible ground target point repeatedly in two or more cameras appears in , if it appears, it is 1, otherwise it is 0;

[0034] The standard deviation of the ground target point coverage repetition is obtained, and the calculation formula is:

[0035]

[0036] The maximum average angle of the repeatedly covered ground target point Oritation is obtained, and the calculation formula is:

[0037]

[0038] wherein θ r is the plane angle between any two cameras repeatedly covering the same visible ground target point, and it is required to ensure that the angle value range is [0°, 180°], x lat , x lon is the coordinate of the x point converted from latitude and longitude to radian, y lat , y lon is the coordinate of the y point converted from latitude and longitude to radian, z lat , z lon is the coordinate of the z point converted from latitude and longitude to radian, the y point is a ground target point, and the z point and the x point are the position coordinates of two cameras covering the y point at the same time, the max(θ r ) function is to take the maximum value of all angles.

[0039] Based on the ground target point coverage, the ground target point coverage repetition, the standard deviation of the ground target point coverage repetition, and the maximum average angle of the repeatedly covered ground target point, the fitness function is obtained, and the calculation formula is:

[0040]

[0041] wherein k1 and k2 are the weights of coverage and respectively, and k3 and k4 are the weights of (1-Oritation) and σ respectively.

[0042] Further, the method for updating the chromosome through the mutation operation is specifically:

[0043] The mutation operation employs spatially adaptive differential evolution, as shown in the following formula:

[0044]

[0045] Among them, v l It is the l-th mutated individual, chm best It is the optimal solution for the current population, chm best2 This represents the suboptimal solution for the current population, where r1 and r2 are two distinct random integers representing the population size. r1 This represents the random individual corresponding to r1, chm r2 rand represents the random individual corresponding to r2. p [0,1] represents the random real number taken in the p-th round of mutation, curround is the current iteration number, and R m R m2 R is the initial probability of selecting this term. n R n2 The optimal probability chosen is the shrinkage parameter, and ⊙ represents the probability of element-wise multiplication. far It is a remote item.

[0046] Furthermore, the method for updating chromosomes through crossover is as follows:

[0047] By incorporating an adaptive crossover operation with random perturbations, borrowing from the harmony search algorithm, the formula is as follows:

[0048]

[0049] Where j is a gene segment of a certain length, dim is the latitude (i.e., the number of solutions to be found), and NP is the population size. For the first Gene fragments of new individuals produced by crossover. It is a gene segment of an individual in the current population. For the first The gene fragments of each mutant individual, where rand(j) is a randomly selected gene fragment, M R To preserve the probability, C R For crossover probability, Indicates the first The random real number obtained by the round cross operation;

[0050] The formula for calculating each gene in rand(j) is:

[0051] ran = randint[2, L-1];

[0052] Among them, randj (t) represents the tth gene locus in rand(j), chm j (t) represents the tth gene locus in the gene segment of the current population individual, randint[0,1] represents a random number taking 0 or 1, and randint[2, L-1] represents a random number taking 2 to L-1

[0053] The adjustment parameter is adjusted by an adaptive adjustment mode, and the specific formula is as follows:

[0054]

[0055] Wherein, MR s (N) represents the retention probability and the crossover probability of the s th individual in the N th generation, ω s (N) represents the retention probability and the crossover probability of the s th individual in the N th generation, ω s = (f avg -f s ) / f avg , and α and β are the adjustment intervals of the retention probability and the crossover probability respectively, f avg is the average fitness value of the current population, and f s is the fitness value of the s th individual, and MR1(N) and CR1(N) in the N th generation are obtained according to the Logistic function.

[0056] A spatial difference evolution-based intelligent camera control device for perception, comprising:

[0057] A scene layer creation module for creating a scene layer for a study area;

[0058] A ground target point acquisition module for generating ground target points not on obstacles based on the scene layer according to a preset density;

[0059] A candidate point acquisition module for selecting candidate points of cameras in the study area;

[0060] A line of sight construction module for constructing lines of sight from all candidate points to all ground target points;

[0061] A preliminary selected visible point acquisition module for performing a line-of-sight analysis on the lines of sight to obtain preliminary selected visible points;

[0062] A final selected visible point acquisition module for filtering out final selected visible points as visible ground target points from the preliminary selected visible points based on the visible distance and the viewing angle of the camera;

[0063] A camera control scheme acquisition module for setting the number of camera controls, selecting the best combination based on the visible ground target points in all candidate points using a spatial adaptive differential evolution algorithm, and taking the best combination as the camera control scheme of the study area.

[0064] An electronic device comprises a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the spatial difference evolution-based perception camera intelligent deployment method described above when executing the program.

[0065] A non-transitory computer-readable storage medium has a computer program stored thereon, and the computer program implements the spatial difference evolution-based perception camera intelligent deployment method described above when executed by a processor.

[0066] A computer program product comprises a computer program, and the computer program implements the spatial difference evolution-based perception camera intelligent deployment method described above when executed by a processor.

[0067] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0068] The spatial difference evolution-based perception camera intelligent deployment method provided by the present application is based on the scene of a research area, comprehensively considers various limiting factors such as monitoring blind area, obstacle shielding, line-of-sight radiation range, coverage overlap degree, and effective monitoring area, and combines the specific parameter characteristics of the camera, selects the final selected visible point from the initial selected visible point as the visible ground target point based on the visual distance and the visual angle of the camera, sets the number of camera deployments, and selects the best combination based on the spatial adaptive differential evolution algorithm of all candidate points. Compared with the traditional experience method, the present embodiment can select the relatively optimal combination from the candidate point combination of millions of levels, and is not easy to fall into local optimization, thereby providing reliable guarantee for the governance, development, and emergency of each region. BRIEF DESCRIPTION OF DRAWINGS

[0069] The accompanying drawings, which are included to provide a further understanding of the present application, constitute a part of this application. The illustrative embodiments of the present application and their description serve to explain the present application. They do not, however, limit the present application. In the drawings:

[0070] Figure 1 is a flow chart of the urban monitoring camera intelligent deployment planning method based on the binary coding differential evolution algorithm of the present application;

[0071] Figure 2 is a three-dimensional data graph of a research area in the specific embodiment of the present application;

[0072] Figure 3 is a three-dimensional ground point distribution schematic diagram after discretization;

[0073] Figure 4 is a candidate point position of camera installation selected by an artificial;

[0074] Figure 5is a schematic view of the result of constructing a line of sight from each candidate point to a ground point;

[0075] Figure 6 is a DSM data map of the reference area;

[0076] Figure 7 is a result table of population coding;

[0077] Figure 8 is a comparison chart of the convergence results of ten times of artificial cloth points, genetic algorithm and spatial adaptive difference algorithm;

[0078] Figure 9 is an evolution curve chart of the first time of simultaneously performing genetic algorithm and spatial adaptive algorithm;

[0079] Figure 10 is an evolution curve chart of the first time of simultaneously performing genetic algorithm and spatial adaptive algorithm;

[0080] Figure 11 is a point distribution chart of the optimal convergence result;

[0081] Figure 12 is a visibility schematic view of the optimal convergence result. DETAILED DESCRIPTION

[0082] The technical solutions in the embodiments of the present application will be clearly and completely described in combination with the drawings in the embodiments of the present application.

[0083] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the purpose of facilitating the description of the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application. In addition, the terms "first", "second" and the like are only for the purpose of description and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" and the like can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0084] First embodiment

[0085] In recent years, especially in urban areas, the normal operation and efficient management of social life today cannot be separated from video monitoring. But in the process of using the camera, it is easy to appear the picture is not clear, obstacles block the line of sight, the invalid area of the monitoring area is too large and other problems, the camera should play a major role to bring a lot of convenient efficiency is greatly reduced. Comprehensive analysis of space, cost and placement and other factors, the more ideal situation is to use the least camera to achieve the maximum effective area of observation range. But just by human pure feeling to select and place the camera position, the error will be too large, the effective monitoring rate of video monitoring will be very low.

[0086] Therefore, based on the above assumptions, the first embodiment proposes to improve the specific situation on the basis of intelligent optimization algorithm, to realize the maximum monitoring coverage of the camera by using genetic algorithm under the condition of fixed number of cameras, as shown in Figure 1 , including:

[0087] Step S1: create a scene layer for the study area;

[0088] Step S2: generate ground target points not on obstacles based on the scene layer according to the preset density;

[0089] Step S3: select candidate points for the camera in the study area;

[0090] Step S4: construct the line of sight from all candidate points to all ground target points;

[0091] Step S5: perform visibility analysis on the line of sight to obtain preliminary selected visible points;

[0092] Step S6: filter out the final selected visible points as visible ground target points from the preliminary selected visible points based on the visible distance and angle of view of the camera;

[0093] Step S7: set the number of camera control, based on the visible ground target points, select the best combination of all candidate points using a spatial self-adaptive differential evolution algorithm, and take the best combination as the camera control scheme of the study area.

[0094] The method for intelligently deploying a perception camera based on spatial differential evolution provided in this embodiment considers various limiting factors such as monitoring blind spots, obstacle blockage, line-of-sight radiation range, coverage overlap, and effective monitoring area, in combination with the specific parameter characteristics of the camera, and selects the final selected visible point from the initial selected visible point as the visible ground target point based on the visual distance and angle of view of the camera, sets the number of camera deployments, and selects the optimal combination based on the spatial adaptive differential evolution algorithm of all candidate points. Compared with the traditional experience method, this embodiment can select the relatively optimal combination from the candidate point combination of millions, and is not prone to local optimization, providing reliable guarantee for the governance, development, and emergency of various regions.

[0095] The above steps are described in detail below.

[0096] In step S1 of this embodiment, a scene layer is created: taking part of a community as an example, the original 3D tile data is converted into an integrated mesh scene layer (slpk format) using the scene layer creation tool in the geographic processing tool of ArcGIS Pro software, and the result is displayed in the local scene. The data of part of a community is obtained by a UAV, as shown in FIG. 1. Figure 2 The three-dimensional data graph of the research area in the specific embodiment of the application is shown.

[0097] In step S2 of this embodiment, the method for generating ground target points not on obstacles according to a preset density is specifically as follows:

[0098] Two-dimensional data points are generated on the scene layer according to a preset density, the two-dimensional data points are upgraded to three-dimensional data points according to the ground data in the digital surface model (DSM) format of the research area, the three-dimensional data points on the ground obstacles due to upgrading are removed according to the size of the elevation value, and the three-dimensional data points that need to be observed on the ground are retained as ground target points.

[0099] Specifically: discretization of the monitoring target area: first, two-dimensional data points of a preset density are automatically generated on the two-dimensional surface corresponding to the scene layer using python, and then the two-dimensional data is upgraded to three-dimensional data according to the ground data in the digital surface model format, that is, the two-dimensional mesh data points are attached to the surface of the three-dimensional scene. The digital surface model format data is converted from osgb data by Supermap software, as shown in FIG. 2. The ground obstacles are usually houses, trees, etc., and then the points on the houses, trees, etc. due to upgrading are removed according to the size of the z value, and only the ground target points that need to be observed are retained, a total of 7783, as shown in FIG. 3. Figure 6 Figure 3

[0100] ​​In step S3 of the embodiment, candidate points of the camera are selected: according to the experience of the city range oblique photogrammetry DSM data, appropriate camera control candidate points are selected, such as low resolution camera candidate points on both sides of the road, with a height of 3 meters, and high resolution cameras are generally distributed at the side edges of buildings.

[0101] In the embodiment, the height of the camera is set to 3 meters, and a total of 367 candidate points are selected, and the red points in the figure are the selected candidate points, as shown in the figure. Figure 4

[0102] In step S4 of the embodiment, a line of sight is constructed: a line of sight tool provided by the 3D Analyst toolbox in ArcGIS Pro is used to construct the line of sight from all candidate points to the ground target point.

[0103] In step S5 of the embodiment, a visibility analysis is performed: all lines of sight are analyzed using the 3D Analyst toolbox in ArcGIS Pro to determine whether a line of sight can pass through an obstacle and whether a target point is visible, as shown in the figure. Figure 5

[0104] In step S6 of the embodiment, visible points are selected: the actual distance between the candidate point corresponding to the camera and the ground target point is calculated in the attribute table, and the angle between the line of sight and the plumb line of the ground target point is calculated. Finally, according to the specific camera visibility distance, angle of view and other limiting conditions, points that meet the conditions are selected, specifically: when the actual distance between the candidate point and the ground target point is less than or equal to the visibility distance of the camera and the angle between the line of sight and the plumb line of the ground target point is less than or equal to the angle of view, it indicates that the ground target point is a point that meets the conditions, i.e. the ground target point is a visible ground target point; when the actual distance between the candidate point and the ground target point is greater than the visibility distance of the camera and the angle between the line of sight or the plumb line of the ground target point is greater than the angle of view, it indicates that the ground target point is not a point that meets the conditions, i.e. the ground target point is not a visible ground target point. In the embodiment, the camera parameter limit is set to a visibility distance of 50 meters and an angle of view of 75°.

[0105] In the embodiment, the preset density can be set according to actual needs, for example, 0.5 m per point or 1 m per point.

[0106] In step S7 of the embodiment, a spatial adaptive differential evolution algorithm is used to select the best combination of all candidate points based on the visible ground target points, and the method is specifically as follows:

[0107] Step S701: a control population including a plurality of chromosomes is constructed, and the chromosomes are coded by all candidate points;

[0108] ​​Step S702: Establish a fitness function to obtain the fitness function value of the chromosome;

[0109] Step S703: Update the chromosomes and iteratively deploy the population by mutation operations of spatial adaptive differential evolution and / or by introducing adaptive crossover selection with random perturbation by referencing the harmony search algorithm;

[0110] Step S704: Select the optimal chromosome based on the fitness function value.

[0111] In step S701 of this embodiment, the candidate points are encoded using binary encoding, forming a binary encoding matrix composed of 0s and 1s. The matrix size is NP*M, where NP is the population size, M is maxL*n, maxL is the number of bits in the binary encoding of the maximum candidate point index, and n is the number of available cameras. In a specific implementation of this embodiment, the candidate points are encoded using binary encoding, forming a binary encoding matrix composed of 0s and 1s. The matrix size is NP*M, where NP is the population size set to 30, M is maxL*n, maxL is the number of bits in the binary encoding of the maximum candidate point index 367 (9 bits), and n is the number of available cameras, here set to 25. Figure 7 As shown.

[0112] In this embodiment S702, the fitness function is designed by considering four dimensions simultaneously: ground coverage, ground target point coverage repetition, standard deviation of ground target point coverage repetition, and the average of the maximum included angles of the repeatedly covered ground target points.

[0113] The formula for calculating ground target point coverage is as follows:

[0114]

[0115] |C|=m

[0116] Among them, w i is the weight corresponding to the i-th visible ground target point, and m is the total number of all visible ground target points. It is the first The weights corresponding to each ground target point are given by p1, p2, p3, ..., pc, where n is the total number of ground target points, and C is the set obtained by combining the visible ground target points of all selected cameras. m For all visible ground target points, It is the first The set of visible ground target points corresponding to each camera, where q is the total number of cameras.

[0117] Obtain ground target point coverage redundancy The average number of times that a repeated covered ground target point is covered by different cameras is calculated by the formula:

[0118]

[0119] wherein D is a set of all repeated visible ground target points in two or more cameras, |D| is the number of elements in the set D, is the kth repeated visible ground target point in two or more cameras, is used to determine whether the repeated visible ground target point in two or more cameras appears in , if it appears, it is 1, otherwise it is 0.

[0120] For example, the set of all repeated visible ground target points in two or more cameras includes a first visible target point, a second visible target point and a third visible target point, the total number of selected cameras is 5, wherein the first visible target point can be repeatedly seen by 2 cameras, the second visible target point can be repeatedly seen by 3 cameras, and the third visible target point can be repeatedly seen by 2 cameras, then

[0121] The standard deviation σ of the coverage repetition of the ground target point is obtained, that is, the standard deviation of the coverage of the repeated visible ground target point by different cameras is calculated, and the calculation formula is:

[0122]

[0123] The maximum average angle of orientation Oritation of the repeated covered ground target point is obtained, and the calculation formula is:

[0124]

[0125] wherein θ r is the plane angle between any two cameras that repeatedly cover the same visible ground target point, and it is required to ensure that the angle value range is [0°, 180°], x lat , x lon are the latitude and longitude coordinates of the x point converted into radians, y lat , y lon are the latitude and longitude coordinates of the y point converted into radians, z lat , z lon are the latitude and longitude coordinates of the z point converted into radians, the point y is a ground target point, z and x are the position coordinates of two cameras that cover the point y at the same time, and max(θ r) function is to take all the maximum value of the angle, so Oritation is the average of the maximum value of the angle between two or more cameras corresponding to the ground target points covered repeatedly.

[0126] Based on the ground target point coverage, ground target point coverage, ground target point coverage, the standard deviation of the maximum angle of the repeated coverage of the ground target point, the fitness function is obtained, and the calculation formula is:

[0127]

[0128] Among them, k1, k2 are the weight of coverage and , k3, k4 are the weight of (1-Oritation) and σ, coverage and are positive attributes, that is, the evolution direction of fitness function Fitness is to hope that coverage and The evolution direction of the larger direction, and (1-Oritation) and σ are negative attributes, and their increase will make the value of fitness function Fitness smaller, and the evaluation of the individual is lower.

[0129] In the embodiment S703, the method for updating the chromosome by mutation operation is:

[0130] The mutation operation adopts the spatial adaptive differential evolution (SA-DEA / best-and-rand / 2, Spatial Adaptive Differential Evolution Algorithm) method, and the formula is as follows:

[0131]

[0132] Among them, v l is the lth mutation individual, chm best is the current population optimal solution, chm best2 is the current population suboptimal solution, r1 and r2 are two different random integers in the population size, chm r1 indicates the random individual corresponding to r1, chm r2 indicates the random individual corresponding to r2, rand p [0,1] indicates a random real number taken by the pth mutation operation, curround is the current iteration number, R m , R m2 is the starting probability of expecting to select this item, R n , R n2 is the optimal probability of expecting to select, is the contraction parameter, and the suboptimal probability of expecting to select is indicated by ⊙. Among them, chm farThe distal item refers to the distal individuals randomly selected from the tail of the current population that are performing poorly. It is usually the distal individuals randomly selected from the tail of the current population that are performing poorly.

[0133] In step S703 of this embodiment, the method for updating chromosomes through crossover is specifically as follows:

[0134] By incorporating an adaptive crossover operation with random perturbations, borrowing from the harmony search algorithm, the formula is as follows:

[0135]

[0136] Where j is a gene segment of a certain length, dim is the latitude (i.e., the number of solutions to be found), and NP is the population size. For the first Gene fragments of new individuals produced by crossover. It is a gene segment of an individual in the current population. For the first The gene fragments of each mutant individual, where rand(j) is a randomly selected gene fragment, M R To preserve the probability, C R For crossover probability, Indicates the first The random real number obtained by the round cross operation.

[0137] Since the individuals in this paper are encoded using index numbers, crossover cannot be performed point-by-point. Instead, crossover is performed with each gene point represented by L bits. Here, L represents the number of bits after the binary encoded index number. Each gene in rand(j) is:

[0138] ran = randint[2, L-1]

[0139] Among them, rand j (t) represents the t-th gene position in rand(j), chm j (t) represents the t-th gene position in the gene segment of the current population individual, randint[0,1] represents a random number of 0 or 1, and randint[2,L-1] represents a random number of 2 to L-1.

[0140] Through continuous cycles, a new crossover offspring population is eventually generated, with the same size as the parent.

[0141] The parameters are adjusted using an adaptive adjustment method, as shown in the following formula:

[0142]

[0143] Among them, MRs (N), CR s (N) denotes the retention probability and the crossover probability of the s-th individual in the N-th generation, ω s = (f avg -f s ) / f avg , α, β are the adjustment intervals of the retention probability and the crossover probability, f avg is the average fitness value of the current population, f s is the fitness value of the s-th individual, i.e. the retention and crossover probability values are adjusted according to the fitness value, the MR1(N) and CR1(N) calculation formulas of the N-th generation are related to the Logistic function, and are as follows:

[0144]

[0145] wherein c MR , c CR , α MR , α CR , b MR , b CR represent control quantities, c MR , c CR are used to control the maximum value of the Logistic function, and the value of the Logistic function tends to c MR / CR +d MR / Cr as n increases; α MR / CR controls the translation position of the Logistic function curve. Specifically, α MR / CR determines the starting position of the curve moving on the n-axis, and when α MR / CR increases, the curve moves to the left, and vice versa; b MR / CR determines the steepness of the Logistic function, which actually controls the slope of the function curve. When b MR / CR is larger, the steepness of the curve is higher, and the function changes faster; when b MR / CR is smaller, the curve is flatter, and the change is slower; finally, d MR / CR controls the minimum value of the Logistic function.

[0146] Selection operation: selecting the new population and the old population with the largest fitness value to update the old population.

[0147] In summary, the embodiment is based on the three-dimensional DSM data of the research area produced by aerial oblique photogrammetry, various limiting factors such as monitoring blind area, tree and building shielding, line-of-sight radiation range, coverage overlap, and effective monitoring area are comprehensively considered, and the intelligent comprehensive evaluation model of the network coverage of the video monitoring equipment is established in combination with the specific parameter characteristics of the video monitoring equipment, and the spatial population coding and the differential evolution operator of the spatial multi-mode differential evolution camera spatial deployment algorithm are designed. In the specific evolution process, first, the candidate points are binary coded, the population and parameters are initialized, the individual fitness function is calculated, the parent body is selected layer by layer, the hybridization and mutation operations are performed, and the parameters are adaptively optimized and adjusted at each layer to improve the efficiency and constantly update the population until convergence. Compared with the traditional experience method, the embodiment can select the relatively optimal combination from the candidate point combination of millions of levels, and the adaptive adjustment method is also adopted in the parameter aspect, so that the population can naturally evolve according to the S-shaped curve as a whole, and then the optimal scheme can be obtained more quickly, and it is not easy to fall into local optimum, which provides reliable guarantee for the governance, development and emergency of the city.

[0148] In order to verify the effectiveness of the application, the embodiment is based on real three-dimensional community data (OSGB format) of a community, various limiting factors such as monitoring blind area, tree and building shielding, line-of-sight radiation range, and effective monitoring area are comprehensively considered, the specific parameters of the video monitoring equipment are combined, the line-of-sight between the camera and the ground monitoring point is constructed by using the spatial intelligent calculation method of the genetic algorithm, whether the ground test points involved can be seen is judged by using the visibility analysis tool of ArcGIS Pro, the intelligent optimization algorithm is designed, and the best placement position of the video monitoring is obtained through multiple iterations, selection, hybridization, mutation and other processes.

[0149] Firstly, the parameter setting in the parameter adjustment operation of the algorithm of the application is R m =0.5, R n =0.3, R m2 =0.8, R n2 =0.1, alpha MR =alpha CR =6, beta MR =beta CR =0.03, c CR =0.15, d CR =0.8, alpha=

[0150] =0.2, c MR =0.6, d MR= 0.2, β = 0.05, k1 = 0.8, k2 = 0.2, k3 = 2, k4 = 1 population and iteration number settings are 30 populations 2000 iteration times respectively by relevant practitioners according to experience, binary genetic algorithm and the binary improved differential evolution proposed in the patent are tested, the number of cameras is 25, the iteration number is 2000 times, each method is run 10 times respectively, and the maximum coverage rate obtained each time is as shown in Figure 8 .

[0151] As shown in Figure 9 and Figure 10 It can be seen that the proportion of the camera coverage monitoring area of the optimal control position calculated by the method of the application is not easy to fall into a local optimal value compared with the result of the traditional genetic algorithm. Although the genetic algorithm can sometimes get better results, it is easy to stagnate in a search area for a long time, but the application is not easy to fall into a local optimum. It can also be clearly seen that the differential evolution algorithm converges much faster than the genetic algorithm.

[0152] Visualization and application: finally, the experiment filters the line in the attribute table after the viewshed analysis according to the index number of the optimal individual obtained in the above algorithm in ArcGIS Pro, and then exports it as a Shapefile file. As shown in Figure 12 , the visualization display in ArcGIS Pro is a visual diagram of the optimal convergence result. As shown in Figure 11 , it is a point distribution diagram of the optimal convergence result.

[0153] Second embodiment

[0154] The second embodiment provides a spatial differential evolution-based intelligent camera control device, comprising:

[0155] A scene layer creation module for creating a scene layer for a study area;

[0156] A ground target point acquisition module for generating ground target points not on obstacles based on the scene layer according to a preset density;

[0157] A candidate point acquisition module for selecting candidate points of cameras in the study area;

[0158] A line of sight construction module for constructing lines of sight from all candidate points to all ground target points;

[0159] A preliminary selected visible point acquisition module for performing viewshed analysis on the lines of sight to obtain preliminary selected visible points;

[0160] A final selected visible point acquisition module for filtering final selected visible points from the preliminary selected visible points as visible ground target points based on the visual distance and viewing angle of the cameras.

[0161] The camera control scheme acquisition module is configured to set the number of cameras to be controlled, select an optimal combination based on all candidate points using a spatial adaptive differential evolution algorithm, and take the optimal combination as the camera control scheme of the research area.

[0162] The third embodiment

[0163] The third embodiment provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the above-mentioned spatial differential evolution-based perception camera intelligent control method when executing the program.

[0164] The fourth embodiment

[0165] The fourth embodiment provides a non-transitory computer-readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned spatial differential evolution-based perception camera intelligent control method.

[0166] The memory in the embodiments of the application is used to store various types of data to support the operation of the electronic device. Examples of these data include: any computer program for operating on the electronic device.

[0167] The fifth embodiment

[0168] The fourth embodiment provides a computer program product, including a computer program, and the computer program is executed by a processor to implement the above-mentioned spatial differential evolution-based perception camera intelligent control method.

[0169] The method for intelligent control of a perception camera based on spatial differential evolution disclosed in the embodiments of the present application can be applied in a processor or implemented by a processor. The processor can be an integrated circuit chip with a processing capability of signals. In the implementation process, each step of the method for intelligent control of a perception camera based on spatial differential evolution can be completed by integrated logic circuits of hardware in the processor or instructions in the form of software. The processor mentioned above can be a general processor, a digital signal processor (DSP), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor can implement or execute each method, step and logic block diagram disclosed in the embodiments of the present application. The general processor can be a microprocessor or any conventional processor, etc. In combination with the steps of the method disclosed in the embodiments of the present application, the hardware decoding processor can be directly embodied to execute the steps, or the hardware and software modules in the decoding processor can be combined to execute the steps. The software module can be located in a storage medium, which is located in a memory. The processor reads information in the memory and combines the hardware to complete the steps of the method for intelligent control of a perception camera based on spatial differential evolution provided in the embodiments of the present application.

[0170] In exemplary embodiments, the electronic device can be implemented by one or more application specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), FPGAs, general-purpose processors, controllers, micro controllers (MCUs), microprocessors (Microprocessors), or other electronic elements for executing the aforementioned methods.

[0171] It can be understood that the memory can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a ferromagnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM). The magnetic surface memory can be a disk memory or a tape memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example but not limitation, many forms of RAM can be used, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), sync link dynamic random access memory (SLDRAM), and direct rambus random access memory (DRRAM).The memory described in the embodiments of the present application is intended to include, but not limited to, these and any other suitable types of memory.

[0172] The above embodiments are only examples of the technical solutions of the present application. The method involved in the present application is not limited to the content described in the above embodiments, but is subject to the scope defined in the claims. Any modification or supplement or equivalent replacement made by the skilled in the art on the basis of the above embodiments is within the scope claimed by the claims of the present application.

Claims

1. A method for intelligent deployment and control of sensing cameras based on spatial differential evolution, characterized in that, include: Create a scene layer for the study area; Based on the scene layer, ground target points that are not on obstacles are generated according to a preset density. Candidate camera points are selected for the study area; Construct the line of sight from all candidate points to all ground target points; Perform a line-of-sight analysis to obtain initial viewpoints; Based on the camera's viewing distance and angle, the final viewpoint is selected from the initial viewpoints as the visible ground target points; The number of cameras is set, and the best combination is selected from all candidate points based on visible ground target points using a spatial adaptive differential evolution algorithm. The best combination is then used as the camera deployment scheme for the study area. The method for selecting the optimal combination of all candidate points using the spatial adaptive differential evolution algorithm is as follows: Construct a controlled population consisting of several chromosomes, each encoded by all candidate points; Establish a fitness function to obtain the fitness function value of the chromosome; Chromosomes are updated and the population is iteratively controlled through mutation operations and / or crossover selection; Selecting the optimal chromosome based on the fitness function value; The specific methods for updating chromosomes through mutation operations are as follows: The mutation operation employs spatially adaptive differential evolution, as shown in the following formula: = = in, It is the first A mutated individual, It is the optimal solution for the current population. It is the second-best solution for the current population. and These are two distinct random integers representing the population size. express The corresponding random individual, express 2 corresponds to a random individual, Indicates the first The random real number obtained by the round mutation operation. This is the current iterative algebra. , To determine the initial probability of selecting this term, , The optimal probability to be selected is the shrinkage parameter. This indicates element-wise multiplication. It is a remote item.

2. The intelligent deployment method for sensing cameras based on spatial differential evolution according to claim 1, characterized in that, The method for generating ground target points that are not on obstacles based on a preset density is as follows: Two-dimensional data points are generated on the scene layer according to a preset density. The two-dimensional data points are then upgraded to three-dimensional data points based on the ground data of the digital surface model of the study area. The three-dimensional data points that are on obstacles due to the upgrade are removed according to the elevation value. The three-dimensional data points that need to be observed on the ground are retained as ground target points.

3. The intelligent deployment method for sensing cameras based on spatial differential evolution according to claim 1, characterized in that, The method for establishing the fitness function is as follows: Obtain ground target point coverage The calculation formula is: in, It is the first The weights corresponding to each visible ground target point. This represents the total number of all visible ground target points. It is the first The weights corresponding to each ground target point This represents the total number of all ground target points. It is the set obtained by combining the visible ground target points of all selected cameras. For all visible ground target points, It is the first The set of visible ground target points corresponding to each camera. This represents the total number of cameras; Obtain ground target point coverage redundancy The calculation formula is: in, The set is the set of all visible ground target points that appear repeatedly in two or more cameras. for The number of elements in the set For the first A visible ground target point that repeats in two or more cameras. Used to identify visible ground target points that appear repeatedly in two or more cameras. Does it appear If it appears, the value is 1; otherwise, it is 0. The standard deviation of ground target point coverage repeatability is obtained using the following formula: ； Obtain the average of the maximum included angles of overlapping ground target points. The calculation formula is: in, It refers to the angle between any two cameras that repeatedly cover the same visible ground target point, and the angle must be within the range of [0°, 180°]. , for Converting latitude and longitude coordinates of a point to radians , for Converting latitude and longitude coordinates of a point to radians , for Converting latitude and longitude coordinates of a point to radians The point is the ground target point. Point and Points are covered simultaneously The coordinates of the two cameras at the point. The function is to take the maximum value among all included angles; Based on ground target point coverage, ground target point coverage repetition, the standard deviation of ground target point coverage repetition, and the average maximum included angle of repetitive ground target points, the fitness function is obtained, and the calculation formula is as follows: in, They are and The weights, They are The weights.

4. The intelligent deployment method for sensing cameras based on spatial differential evolution according to claim 1, characterized in that, The specific method for updating chromosomes through crossover is as follows: By incorporating an adaptive crossover operation with random perturbations, borrowing from the harmony search algorithm, the formula is as follows: in, A gene segment of a certain length, NP represents the latitude, i.e., the number of solutions to be found, and the population size. For the first Gene fragments of new individuals produced by crossover. It is a gene segment of an individual in the current population. For the first Gene fragments from a mutated individual, These are randomly selected gene fragments. To preserve probability, For crossover probability, Indicates the first The random real number obtained by the round cross operation; The formula for calculating each gene in the formula is: ， ； in, express The first in One gene locus, This represents the first gene segment of an individual in the current population. One gene locus, This represents a random number that can be either 0 or 1. Indicates taking arrive random numbers in The parameters are adjusted using an adaptive method, and the specific formula is as follows: in, , Indicates the first The individual in the first The retention probability and crossover probability of each generation. These are the adjustment ranges for retention probability and crossover probability, respectively. This represents the current average fitness value of the population. For the current number The fitness value of each individual, the first generation and Obtained from the Logistic function.

5. A smart deployment device for sensing cameras based on spatial differential evolution, characterized in that, include: The scene layer creation module is used to create scene layers for the study area; The ground target point acquisition module is used to generate ground target points that are not on obstacles based on the scene layer and a preset density. The candidate point acquisition module is used to select candidate points for cameras in the study area; The line-of-sight construction module is used to construct the line-of-sight from all candidate points to all ground target points; The initial viewpoint acquisition module is used to perform line-of-sight analysis to obtain initial viewpoints; The final viewpoint acquisition module is used to select the final viewpoint as the visible ground target point from the initial viewpoint based on the camera's viewing distance and angle. The camera deployment scheme acquisition module is used to set the number of cameras to be deployed. Based on visible ground target points, it uses a spatial adaptive differential evolution algorithm to select the best combination from all candidate points and uses the best combination as the camera deployment scheme for the study area. The method for selecting the optimal combination of all candidate points using the spatial adaptive differential evolution algorithm is as follows: Construct a controlled population consisting of several chromosomes, each encoded by all candidate points; Establish a fitness function to obtain the fitness function value of the chromosome; Chromosomes are updated and the population is iteratively controlled through mutation operations and / or crossover selection; Selecting the optimal chromosome based on the fitness function value; The specific methods for updating chromosomes through mutation operations are as follows: The mutation operation employs spatially adaptive differential evolution, as shown in the following formula: = = in, It is the first A mutated individual, It is the optimal solution for the current population. It is the second-best solution for the current population. and These are two distinct random integers representing the population size. express The corresponding random individual, express 2 corresponds to a random individual, Indicates the first The random real number obtained by the round mutation operation. This is the current iterative algebra. , To determine the initial probability of selecting this term, , The optimal probability to be selected is the shrinkage parameter. This indicates element-wise multiplication. It is a remote item.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the intelligent deployment method for sensing cameras based on spatial differential evolution as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the intelligent deployment method for sensing cameras based on spatial differential evolution as described in any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent deployment method for sensing cameras based on spatial differential evolution as described in any one of claims 1 to 4.

Images