Target monitoring method, system and device and storage medium
By constructing and combining the algorithm MFGTM and optimization algorithm FTO in a wireless video sensor network, the problem of being unable to build a full-view monitoring fence is solved, the construction success rate and efficiency are improved, and multi-objective efficiency is achieved Merger and network lifespan extend.
Patent Information
- Application Number
- CN202411057859.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-11-19
- Estimated Expiration
- Not applicable · inactive patent
Smart Images

Figure CN118984504A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wireless video sensor network technology, specifically to a target monitoring method, system, device, and storage medium. Background Technology
[0002] Target fence coverage is an emerging technology in wireless sensor network (WSN) fence coverage. It primarily studies the construction of ring-shaped fences around large monitoring targets within the WSN coverage area, thereby enabling effective detection of intrusion detection target behavior. Existing networks used for target fence construction mainly include undirected sensor networks and directed sensor networks.
[0003] In undirected sensor networks, nodes use a disk-based perception model. When constructing a target fence, the perception direction of the nodes does not need to be considered; only whether the nodes effectively surround the monitored target is required. In directed sensor networks, the perception area of a node is generally modeled as a fan-shaped region. Wireless Video Sensor Networks (WVSNs) are a typical application of directed sensor networks, where the sensing nodes have video acquisition capabilities. Due to the different node perception models, fence construction in this network environment differs from that in undirected sensor networks. To ensure that the fence in a WVSN can capture a frontal image of the monitored object, a definition of full-view coverage for WVSNs is given. This definition is combined with strong and weak fence coverage techniques to propose a full-view strong and weak fence coverage model and detection method. To address the WVSN full-view strong fence coverage problem, a mesh partitioning method is used to divide the WVSN coverage area into equal-sized grids, and then the WVSN full-view fence is constructed using the full-view mesh. For the WVSN full-view weak fence coverage problem, the W-GraProj algorithm for constructing WVSN full-view weak fences is proposed, improving fence construction efficiency and extending network lifetime. The optimal full-view weak fence coverage of WVSN has been proven to be an NP-hard problem; a heuristic algorithm for WVSN weak fence construction, GTPFWBC, is designed to further improve fence construction efficiency.
[0004] However, regardless of whether it is a strong or weak fence with a full field of view, although it achieves full field of view coverage at the fence's coverage area, it cannot achieve the purpose of surrounding the monitored object because the fence is not closed. Summary of the Invention
[0005] To address the shortcomings of existing technologies in constructing a full-view monitoring fence around the monitored object to detect intrusion behavior, this invention proposes a target monitoring method, system, device, and storage medium. By proposing the algorithm MFGTM for constructing and merging the minimum full-view grid target fence in WVSN and the algorithm FTO for optimizing the coverage nodes of the full-view target fence, the problems existing in the prior art are solved.
[0006] A target monitoring method includes the following steps:
[0007] Obtain the WVSN coverage area corresponding to multiple monitoring objects;
[0008] The MFGTM algorithm is used to search for the minimum full-view grid target fence (MFGT) for all monitored objects, and monitored objects with MFGT are merged to obtain the optimal full-view target fence coverage for multiple monitored objects.
[0009] Each monitoring object is monitored based on the optimal full-view target fence coverage;
[0010] The process of searching for the minimum full-view grid target fence (MFGT) for all monitored objects using the MFGTM algorithm, and merging monitored objects with MFGT to obtain the optimal full-view target fence coverage for multiple monitored objects, specifically includes the following steps:
[0011] The WVSN coverage area is divided into grids, and all grids with full viewing angle are marked;
[0012] A breadth-first search algorithm is used to search the non-full-view grid area around the monitored object after grid division; if the non-full-view grid area is not open, the innermost full-view target fence of the monitored object is found along its boundary.
[0013] The Graham algorithm is used to solve for the convex hull of the innermost full-view target fence of the monitored object. The shortest full-view grid path between each pair of adjacent vertices on the convex hull is spliced to obtain the MFGT of the monitored object; the MFGT represents the target fence that uses the fewest full-view grids.
[0014] The monitored objects with MFGT are combined, and a new MFGT for each combination is generated using the Graham algorithm. The MFGT with the fewest full-view grids is used as the optimal full-view target fence coverage for multiple monitored targets.
[0015] Furthermore, the grid division of the WVSN coverage area is specifically carried out by using a square grid model to perform gridding processing on the WVSN coverage area. The processing includes dividing the irregular full-view coverage area into a regular grid.
[0016] Furthermore, the method of using a breadth-first search algorithm to search for the non-full-view grid area around the monitored object after grid division specifically involves iterating through all non-full-view grid neighbors of the non-full-view grid until no new non-full-view grids appear; wherein, the non-full-view grid neighbor grids only include non-full-view grids above, below, to the left and right of the non-full-view grid, and do not include non-full-view grids diagonally opposite the non-full-view grid.
[0017] Furthermore, the optimal full-view target fence coverage is NP-hard; where, if the sensor set S of the WVSN is S = {S1, S2, ..., S...} n If a subset S' of a monitoring object has a minimum full-view grid fence B covering the entire view of the monitored object, but there is no subset Z of S such that Z covers the entire view of B and |Z|<|S'|, then the set S' is an optimal full-view fence coverage of the monitored object.
[0018] Furthermore, it also includes using the FTO algorithm to optimize node coverage of the merged full-view target fence.
[0019] The present invention also includes a target monitoring system, comprising:
[0020] The acquisition module is used to acquire the WVSN coverage area corresponding to multiple monitoring objects;
[0021] The MFGT search module is used to search for the minimum full-view grid target fence MFGT of all monitored objects using the MFGTM algorithm, and merge monitored objects with MFGT to obtain the optimal full-view target fence coverage for multiple monitored objects.
[0022] The monitoring module is used to monitor each monitoring object based on the optimal full-view target fence coverage;
[0023] The MFGT search module includes:
[0024] Mesh division unit, used to divide the WVSN coverage area into meshes and mark all full-view meshes;
[0025] The innermost full-view target fence acquisition module is used to search the non-full-view grid area around the monitored object after grid division using a breadth-first search algorithm; if the non-full-view grid area is not open, the innermost full-view target fence of the monitored object is found along its boundary.
[0026] The MFGT acquisition unit is used to solve the convex hull of the innermost full-view target fence of the monitored object using the Graham algorithm, and to splice the shortest full-view grid path between each pair of adjacent vertices on the convex hull to obtain the MFGT of the monitored object; the MFGT represents the target fence that uses the fewest full-view grids.
[0027] Selected units are used to combine monitoring objects with MFGT. The new MFGT for each combination is calculated using the Graham algorithm. The MFGT with the fewest full-view grids is selected as the optimal full-view target fence coverage for multiple monitoring targets.
[0028] The present invention also includes a target monitoring computer device, comprising: a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the target monitoring method.
[0029] Furthermore, a readable storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, are used to perform the steps of the target monitoring method.
[0030] This invention provides a target monitoring method, system, device, and storage medium, which have the following beneficial effects:
[0031] This invention divides the WVSN coverage area into equally sized grids using the MFGTM algorithm, labels all full-view grids, constructs a minimum full-view grid target fence for each monitoring target, merges multiple full-view target fences, and selects the MFGT with the fewest full-view grids as the optimal full-view target fence coverage for multiple monitoring targets. This method effectively constructs independent full-view target fences for multiple monitoring targets in WVSN and enables efficient merging. Compared to the TBC method, this invention's target fence construction method has a higher success rate, higher construction efficiency, and better fence merging effect, and also exhibits better universality. Attached Figure Description
[0032] Figure 1 This is a schematic diagram showing full-view coverage in an embodiment of the present invention;
[0033] Figure 2 This is a schematic diagram of a wireless video sensor model in an embodiment of the present invention;
[0034] Figure 3 This is a schematic diagram of the WVSN deployment model in an embodiment of the present invention;
[0035] Figure 4 This is a schematic diagram of the grid division of the full-view coverage area in an embodiment of the present invention;
[0036] Figure 5 This is a schematic diagram of a minimum full-view grid target fence for the monitored object in an embodiment of the present invention; Figure 5 (a) is a schematic diagram of a fence with a target in all views; Figure 5(b) is a schematic diagram of the innermost all-view target fence; Figure 5 (c) is a schematic diagram of the minimum full-view grid target fence;
[0037] Figure 6 This is a neighbor diagram of a non-full-view mesh in an embodiment of the present invention; Figure 6 (a) is a schematic diagram of all non-full-view mesh neighbors of the non-full-view mesh; Figure 6 (b) is a schematic diagram of breaking through the boundary of the full view;
[0038] Figure 7 This is a schematic diagram of the angular space division of the full-view grid in an embodiment of the present invention;
[0039] Figure 8 This is a flowchart of the MFGTM algorithm in an embodiment of the present invention;
[0040] Figure 9 This is a schematic diagram of the FTO algorithm flow in an embodiment of the present invention;
[0041] Figure 10 This is a schematic diagram illustrating the effect of the MFGTM algorithm in generating fences and merging fences in a WVSN with 3000 video sensor nodes in an embodiment of the present invention. Figure 10 (a) is a schematic diagram of the coverage effect of a single target and a single fence; Figure 10 (b) is a schematic diagram of the merging effect of multiple target fences;
[0042] Figure 11 This is a schematic diagram comparing the success rate of fence construction in embodiments of the present invention;
[0043] Figure 12 This is a schematic diagram comparing the grid efficiency of constructing fences in an embodiment of the present invention;
[0044] Figure 13 This is a schematic diagram comparing the node efficiency of constructing the fence in an embodiment of the present invention;
[0045] Figure 14 This is a schematic diagram comparing the grid efficiency of fence merging in an embodiment of the present invention;
[0046] Figure 15 This is a schematic diagram comparing the optimization ratio of the FTO algorithm in an embodiment of the present invention;
[0047] Figure 16 This is a schematic diagram comparing the optimization effect of the FTO algorithm on merging fences in an embodiment of the present invention;
[0048] Figure 17 This is a flowchart of a target monitoring method according to an embodiment of the present invention. Detailed Implementation
[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0050] This invention proposes a target monitoring method. First, the WVSN coverage area is divided into equal-sized grids using the MFGTM algorithm. Then, all full-view grids are labeled. Next, a minimum full-view grid target fence is constructed for each monitored target. Then, multiple full-view target fences are merged. Finally, the FTO algorithm is used to optimize the sensor deployment scheme within the merged fence, minimizing the number of sensors covering the merged fence. Figure 17 As shown, the specific steps include:
[0051] S1. Obtain the WVSN coverage area corresponding to the monitored object.
[0052] S2. The minimum full-view target fence (MFGT) for all monitored objects is searched using the MFGTM algorithm and merged. The design of the full-view target fence merging algorithm for wireless video sensor networks addresses the issue that in special scenarios such as nuclear radiation or heavy pollution, manual deployment of WVSNs is not feasible. Therefore, random deployment of WVSNs is often achieved through methods such as aerial seeding, where sensors remain stationary once deployed. Since a certain number of nodes are required in the WVSN to achieve full-view target fence coverage of the monitored objects, this invention assumes that the number of nodes deployed in the WVSN can meet the requirements for constructing a full-view target fence. First, the MFGTM algorithm for WVSN is proposed to search for and merge the MFGTs of all monitored objects in the WVSN. Then, the FTO algorithm for full-view target fences is proposed to optimize the merged MFGTs, further reducing the number of video sensors in the full-view target fence surrounding the monitored objects. The MFGTM algorithm first starts with the non-full-view grid near the monitored object's location, using a breadth-first search to find the non-full-view grid region around the monitored object. If the non-full-view grid region is not open, then the innermost full-view target fence of the monitored object is found along its boundary. The Graham algorithm is used to solve for the convex hull of the innermost full-view target fence of the monitored object, and then the shortest full-view grid paths between each pair of adjacent vertices on the convex hull are concatenated to form the MFGT of the monitored object. Monitored objects with MFGTs are combined, and the new MFGT of each combination is calculated using the Graham algorithm. The combination scheme with the fewest full-view grids is retained as the fence merging result.
[0053] S2.1 Mesh Generation of the WVSN Full-View Coverage Area. In WVSN, the full-view coverage area formed by the collaboration of multiple video sensors is an irregular convex shape. Since the boundaries of this convex shape can be both straight and curved, achieving full-view target fence coverage using these convex-shaped coverage areas presents a complex problem. To reduce the difficulty of constructing the full-view target fence, the WVSN coverage area is meshed, dividing the irregular full-view coverage area into regular grids. This significantly reduces the difficulty of constructing the full-view target fence. This invention uses a square mesh model to divide the network. Figure 4 What is shown is Figure 1 The image shows the grid division results for the full-view coverage area, with the gray grid representing the full-view coverage area. Clearly, this gridding method will lose some full-view coverage area (mainly at the edges of irregular full-view coverage areas), but this error will gradually decrease as the grid size decreases. Full-view coverage: If at any point, regardless of the visitor's facing direction, at least one video sensor in the WVSN can capture a frontal image of the visitor, then that point is said to have full-view coverage. For example... Figure 1 As shown, the area within the thick wireframe where the visitor is located is covered by the entire field of view, and the visitor was only captured by the sensor in a frontal image.
[0054] The principle of WVSN minimum full-view grid target fence construction: After the WVSN coverage area is gridded, if the full-view grid can tightly surround the monitored object, then a full-view target fence must exist around the monitored object. Figure 5 As shown in (a). From the perspective of target fence construction efficiency, the fewer grids a full-view target fence has, the higher its network resource utilization. Here, the target fence using the fewest full-view grids is called the minimum full-view grid target-barrier (MFGT). As a general rule, the target fence composed of full-view grids closer to the inner layer should have fewer grids. Figure 5 (b) The loop formed by the black grid is the innermost full-view target fence that surrounds the monitored object. Obviously, it is not the target fence with the fewest grids, because the shape enclosed by the innermost full-view target fence is not a convex shape. The inward concave boundary actually increases the number of grids. Figure 5 The yellow polygon in (b) is the convex hull constructed based on the center of the grid in the innermost full-view target fence. The vertices of the polygon are the convex hull vertices, which are the grids that the shortest full-view target fence must pass through. Thus, the problem of constructing the full-view target fence coverage of the detection object in WVSN can be transformed into solving the convex hull of the innermost full-view target fence. Figure 5The red grid in (c) constitutes the MFGT of the monitored object, which is formed by stitching together the shortest full-view grid paths between adjacent vertices on the convex hull. In terms of the number of grids alone, the target fence is the shortest because the convex hull polygon has the property of having the shortest perimeter.
[0055] To facilitate the study of the multi-target fence merging problem in WVSN (Wide View System), this invention presents the concepts and models of WVSN and multi-target fence. The perception model of the video sensor is as follows: Figure 2 As shown, a fan-shaped area with an initial angle α, an included angle β, and a sensing radius r is defined. Its communication radius R ≥ 2r. Each sensor can obtain its own coordinate position (x, y) through a built-in positioning device. The WVSN's coverage area is a rectangular planar area with length L and width W. Within this area, n video sensors and m monitored objects are randomly and uniformly deployed. Each monitored object has its own safe distance d. i (i = 1, 2, ..., m), such as Figure 3 As shown. In WVSN, all sensors and monitored targets are static objects, and all parameters remain unchanged once deployed. All sensors in WVSN are isomorphic, with identical parameters except for initial coordinates and starting angles, while all monitored targets are heterogeneous, with different initial positions and safety distances.
[0056] S2.2. After dividing the monitoring object into grids using a breadth-first search algorithm, the non-full-view grid regions around the monitoring object must first be identified. Before searching the innermost full-view target fence of the monitoring object, the non-full-view grid regions near the monitoring object must first be searched using a breadth-first search algorithm. Figure 5 (a) The white grid area surrounded by a gray grid. The search rule is to iterate through all non-full-view grid neighbors of the non-full-view grid, such as... Figure 6 As shown in (a), this continues until no new non-full-view meshes appear or the boundary of the WVSN is found (i.e., the full-view meshes around the monitored object are not closed, and the target fence does not exist). Here, the definition of neighbor meshes only includes the non-full-view meshes above, below, to the left, and to the right of the non-full-view mesh, excluding the non-full-view meshes diagonally opposite it. Otherwise, non-full-view meshes might exploit neighbor relationships to break through the full-view boundary, such as... Figure 6 As shown in (b).
[0057] WVSN Optimal Full-View Target Fence Requirements Analysis: Resource constraints are a current unavoidable drawback of WVSN. In order to save resources and extend network lifespan, WVSN should use as few video sensors as possible when building a full-view target fence for monitored objects.
[0058] Optimal all-view target fence coverage: If the sensor set S of WVSN is S = {S1, S2, ..., S...} nIf a subset S' of a monitoring object fully covers a minimum full-view grid fence B, but there is no subset Z of S such that Z fully covers B and |Z|<|S'|, then the set S' is called an optimal full-view fence coverage of the monitoring object.
[0059] The optimal full-view target fence coverage of a monitored object in WVSN is NP-hard. The proof is as follows: Assume that in WVSN, the MFGT of a monitored object contains p full-view grids. For ease of proof, each full-view grid is divided into q equal angular spaces with its center as the origin, denoted as a. i (i=1,2,…,p×q), such as Figure 7 As shown.
[0060] If a grid has full-view coverage, then the union of the effective angular regions of all video sensors covering it must completely cover [0, 2π). In other words, if the union of the effective angular regions of all video sensors covering a grid completely covers [0, 2π), then this grid must have full-view coverage. However, it's worth noting that video sensors covering a full-view grid are often not optimal and are highly likely to have redundancy. This is the main reason why MFGT needs to be optimized.
[0061] Let the universal set U = {a1, a2, a3, ..., a...} p×q Based on the concept of the effective angle region of a video sensor, it can be known that any visual sensor S in a WVSN... i Effective angle area It is a subset of U, if sensor S i If no grid in the MFGT is covered, then
[0062] As defined, if a subset S' of the sensor set S of a WVSN represents an optimal full-view target fence coverage for a certain monitored object, then... Where z = |S'|, S k ∈S'. Therefore, the optimal full-view target fence coverage problem for monitored objects in WVSN is actually a set coverage problem. Since set coverage problems are NP-hard, the optimal full-view target fence coverage problem for monitored objects in WVSN is also NP-hard.
[0063] S2.3. Use Graham's algorithm to solve for the convex hull of the innermost full-view target fence of the monitored object. Then, stitch together the shortest full-view mesh paths between each pair of adjacent vertices on the convex hull to obtain the MFGT of the monitored object, such as... Figure 8 As shown, the detailed steps of the MFGT algorithm are as follows:
[0064] The WVSN uses the sensor set S and the monitored object set T as inputs.
[0065] (1) Divide the coverage area of WVSN into grids, store all grids in grid set G, and label the full-view coverage grids in G.
[0066] (2) Scan the set of monitored objects T with t as the variable. If the scan is finished, execute (5); otherwise execute (3).
[0067] (3) Perform a breadth-first search on the blank grid area around t based on the neighbor relationship and store it in the blank grid set TempTB. If TempTB extends to the boundary of the WVSN coverage area or is an empty set, execute (2); otherwise execute (4).
[0068] (4) Use Graham's algorithm to find the convex hull MFGT of the TempTB full-view mesh boundary and store it in the single-target fence set MFGTS.
[0069] (5) Combine the single-target fences in MFGTS and store the combination scheme in COM.
[0070] (6) Scan COM with c as the variable. If the scan is finished, execute (8); otherwise, execute step 7.
[0071] (7) Use Graham's algorithm to find the convex hull of the combination scheme, realize the merging of the target fences, and store the merging result of the shortest full-view target fence into MFGT.
[0072] (8) Output MFGT.
[0073] S3. An FTO algorithm for WVSN is proposed, which optimizes the merged MFGT to further reduce the number of video sensors in the full-view target fence surrounding the monitored object. FTO Algorithm: Since optimal full-view target fence coverage of the monitored object in WVSN is an NP-hard problem, a polynomial-time heuristic algorithm is designed to find the suboptimal solution of the optimal full-view target fence. Due to the high-density deployment of WVSN, the video sensors covering the MFGT of the monitored object generally have significant redundancy. Generally, video sensors with more full-view grids covering the MFGT should contribute more to the MFGT coverage. Therefore, this invention sorts the video sensors covering the MFGT according to their coverage contribution from largest to smallest, prioritizing the retention of video sensors with larger coverage contributions while ensuring the MFGT function is not compromised. This is the basic idea behind the FTO algorithm. The sensor set S of WVSN and the minimum full-view grid target fence MFGT are used as inputs to the FTO algorithm, such as... Figure 9 As shown, it includes:
[0074] (1) Scan S with s as the variable. If the scan ends, execute (4); otherwise, execute (2).
[0075] (2) Scan MFGT with g as the variable. If the scan ends, execute step 1; otherwise, execute (3).
[0076] (3) Store s in TS, increment the contribution value s.contribution of node s by 1, store g in the covering mesh set s.cg of node s, store s in the covering node set g.cs of mesh g, and execute (2).
[0077] (4) Sort the sensors in TS by node contribution value s.contribution from smallest to largest.
[0078] (5) Scan TS with s as the variable. If the scan ends, execute (8); otherwise, execute (6).
[0079] (6) Use Tem_TS to store the temporary result after TS is removed.
[0080] (7) If Tem_TS cannot fully cover MFGT to execute (5), then copy Tem_TS to TS and execute (5).
[0081] (8) Output TS.
[0082] S4. Simulation Experiments and Data Analysis.
[0083] This invention uses Matlab 2012a to simulate the MFGTM and FTO algorithms. Wireless video sensors within the WVSN monitoring area are randomly and uniformly deployed. The experimental data is the average value obtained from running the algorithm 30 times, and a comparative analysis was conducted with the TBC algorithm. The relevant parameter settings of the simulation experimental platform are shown in Table 1.
[0084] Table 1 Simulation Experiment Platform Parameter Settings
[0085]
[0086]
[0087] The simulation experiments deployed 1000 to 3500 video sensors within the WVSN coverage area, starting from 1000 and increasing in increments of 100, up to 3500. To reduce the impact of random deployment on the algorithm's performance, each fixed number of WVSNs had 30 different deployment schemes. Figure 10This image shows the effect of the MFGTM algorithm generating fences and merging fences in a WVSN with 3000 video sensor nodes. The white grid represents a non-full-view grid, while the remaining five grids (gray, red, yellow, blue, and black) represent full-view grids. The solid circles and rings in red, yellow, and blue represent three monitored targets and their safe zones, respectively. Figure 10 (a) is a schematic diagram of the coverage effect of a single target and a single fence; Figure 10 (b) is a schematic diagram of the merging effect of multiple target fences.
[0088] Analysis of experimental data shows that the success rate of fence construction using the MFGTM algorithm is ( Figure 11 ) and fence construction efficiency ( Figure 12 , Figure 13 All of these are superior to the TBC algorithm. The success rate of fence construction is about 23 percentage points higher than that of the TBC algorithm. The mesh efficiency and node efficiency of the MFGTM algorithm for fence construction are about 20.2% and 10.8% higher than those of the TBC algorithm, respectively.
[0089] In terms of fence merging efficiency, the MFGTM algorithm is not as good as the TBC algorithm. Figure 14 As shown. This is largely due to the relatively high efficiency of the MFGTM algorithm in constructing a single fence. In other words, the MFGTM algorithm proposed in this invention can construct a single fence for a single target with high efficiency, resulting in a smaller mesh compression ratio during fence merging. This also proves the rationality of the MFGTM algorithm from another perspective.
[0090] For the FTO algorithm, its optimization ratio for fences is generally below 50%, and its optimization effect decreases as the number of deployed nodes increases. Figure 15 As shown, the FTO algorithm shows almost no difference in the degree of fence optimization achieved by the MFGTM and TBC algorithms, such as... Figure 15 , Figure 16 As shown, this demonstrates that the FTO algorithm has good universality.
[0091] This invention addresses the problem of merging full-view target fences in WVSN (Wireless Visual System), proposing the MFGTM algorithm for merging target fences with minimal full-view mesh size and the FTO algorithm for optimizing full-view target fences. Simulation results show that the MFGTM algorithm can effectively construct individual full-view target fences for multiple monitoring targets in WVSN and then merge them. The MFGTM algorithm outperforms the TBC algorithm in terms of single-target fence construction success rate, construction efficiency, and fence merging effect. The FTO algorithm achieves a fence optimization ratio of approximately below 50% and exhibits good versatility. Both MFGTM and FTO effectively reduce the number of sensors required to construct target fences, lower network energy consumption, and further extend the mesh lifespan.
[0092] Based on the same inventive concept, this invention also proposes a target monitoring system, comprising:
[0093] The acquisition module is used to acquire the WVSN coverage area of the monitored object.
[0094] The MFGT search module is used to search for the minimum full-view grid target fence MFGT of all monitored objects using the MFGTM algorithm, and merge monitored objects with MFGT to obtain the optimal full-view target fence coverage for multiple monitored objects.
[0095] The monitoring module is used to monitor the target object based on the optimal all-view target fence coverage.
[0096] The MFGT search module includes:
[0097] Mesh division unit, used to divide the WVSN coverage area into meshes.
[0098] The innermost full-view target fence acquisition module is used to search the non-full-view grid area around the monitored object after the grid is divided using a breadth-first search algorithm; if the non-full-view grid area is not open, the innermost full-view target fence of the monitored object is found along its boundary.
[0099] The MFGT acquisition unit is used to solve the convex hull of the innermost full-view target fence of the monitored object using the Graham algorithm, and to stitch together the shortest full-view grid path between each pair of adjacent vertices on the convex hull to obtain the MFGT of the monitored object.
[0100] Selected units are used to combine monitored objects with MFGT. The new MFGT for each combination is calculated using the Graham algorithm. The MFGT with the fewest full-view grids is selected as the optimal full-view target fence coverage for multiple monitored targets.
[0101] Based on the same inventive concept, the present invention also proposes a target monitoring computer device, comprising: a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the target monitoring method.
[0102] Based on the same inventive concept, the present invention also proposes a readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, are used to perform the steps of a target monitoring method.
[0103] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A target monitoring method, characterized in that: The following steps are involved: Obtain WVSN coverage areas corresponding to multiple monitoring objects; The MFGTM algorithm is used to search for the minimum full-view grid target fence MFGT of all monitoring objects, and the monitoring objects with MFGT are merged to obtain the optimal full-view target fence coverage of multiple monitoring objects. Each monitoring object is monitored according to the optimal full-view target fence coverage; The MFGTM algorithm is used to search for the minimum full-view grid target fence MFGT of all monitoring objects, and the monitoring objects with MFGT are merged to obtain the optimal full-view target fence coverage of multiple monitoring objects. The specific process includes the following steps: Grid the WVSN coverage area and mark all full-view grids; The breadth-first algorithm is used to search the non-full-view grid area around the monitored object after grid division; if the non-full-view grid area is not open, the innermost full-view target fence of the monitored object is searched along its boundary; The convex hull of the innermost full-view target fence of the monitored object is solved using the Graham algorithm, and the shortest full-view grid paths between each pair of adjacent vertices on the convex hull are spliced to obtain the MFGT of the monitored object; the MFGT represents the target fence using the least number of full-view grids; The monitoring objects with MFGT are combined, and the new MFGT of each combination is generated by the Graham algorithm. The MFGT with the least number of full-view grids is used as the optimal full-view target fence coverage for multiple monitoring targets.
2. A target monitoring method according to claim 1, characterized in that: The WVSN coverage area is gridded by adopting a square grid model to perform gridding processing on the WVSN coverage area, and the processing process includes dividing the irregular full-view coverage area into regular grids.
3. A target monitoring method according to claim 1, characterized in that: The breadth-first algorithm is used to search the non-full-view grid area around the monitored object after grid division, and it is specifically performed by looping through all non-full-view grid neighbors of the non-full-view grid until no new non-full-view grid appears; wherein the non-full-view grid neighbor grid only includes the non-full-view grids above, below, left, and right of the non-full-view grid, and does not include the non-full-view grid diagonally opposite to the non-full-view grid.
4. A target monitoring method according to claim 1, characterized in that: The optimal full-view target barrier coverage is NP-hard; if the sensor set S of the WVSN is S = {S1, S2, ..., S n A subset S' of} covers a minimum full-view grid target fence B of the monitored object in full view, but there is no subset Z of S such that Z covers B in full view and |Z|<|S'|, then the set S' is an optimal full-view target fence coverage of the monitored object.
5. A target monitoring method according to claim 1, characterized in that: It also includes using the FTO algorithm to optimize the node layout in the merged optimal full-view target fence.
6. A target monitoring system, characterized in that: include: An acquisition module, used to acquire WVSN coverage areas corresponding to multiple monitoring objects; MFGT search module, used to search for the minimum full-view grid target fence MFGT of all monitoring objects through MFGTM algorithm, and merge the monitoring objects with MFGT to obtain the optimal full-view target fence coverage of multiple monitoring objects; A monitoring module, for monitoring each monitoring object according to the optimal full-view target fence coverage; Wherein, the MFGT search module includes: A grid division unit is used to divide the WVSN coverage area into grids and mark all full-view grids; The innermost full-view target fence acquisition module is used to use a breadth-first algorithm to search the non-full-view grid area around the monitored object after grid division; if the non-full-view grid area is not open, the innermost full-view target fence of the monitored object is searched along its boundary; The MFGT acquisition unit is used to solve the convex hull of the innermost full-view target fence of the monitored object using the Graham algorithm, and to splice the shortest full-view grid paths between each pair of adjacent vertices on the convex hull to obtain the MFGT of the monitored object; the MFGT represents the target fence using the least number of full-view grids; The selection unit is used to combine the monitoring objects with MFGT, and the new MFGT of each combination is calculated by Graham algorithm. The MFGT with the least number of full-view grids is used as the optimal full-view target fence coverage of multiple monitoring targets.
7. A target monitoring computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory, wherein the processor implements the steps of the target monitoring method according to any one of claims 1 to 5 when executing the computer program.
8. A readable storage medium, characterized in that: The readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor, they are used to execute the steps of the target monitoring method according to any one of claims 1 to 5.
