Anti-interference method and system for high-precision global snapshot, and medium
By optimizing camera deployment through digital twin technology and deep learning networks, the accuracy problem of traditional surveillance cameras in shooting high-speed moving objects and dynamic background interference is solved, achieving a highly efficient anti-interference capture effect.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-04-10
AI Technical Summary
Traditional surveillance cameras are prone to deformation when capturing fast-moving objects, affecting the clarity of details. Dynamic background interference can lead to false alarms and missed detections. Existing optimization solutions are costly and difficult to predict the impact of new types of interference.
A high-fidelity digital twin is constructed using digital twin technology to simulate camera deployment schemes and interference scenarios, dynamically adjust the deployment scheme, and build a dual-branch deep learning network and a generative adversarial network to generate an anti-interference capture model.
It improves the accuracy of global image capture, reduces false alarms and missed detections, optimizes camera deployment schemes, and reduces costs and time.
Smart Images
Figure CN121842352A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of high-precision global snapshot, and in particular to an anti-interference method and system for high-precision global snapshot, and a medium. BACKGROUND
[0002] The basis of high-precision snapshot is image acquisition hardware. Traditional surveillance cameras often use rolling shutter sensors, which are prone to distortion ( "jelly effect" ) when shooting high-speed moving objects, affecting detail clarity. Global shutter sensors can expose all pixels simultaneously, perfectly freezing high-speed motion moments, and ensuring image distortion-free. Combined with ultra-high resolution ( such as 4K, 8K ) sensors, it can cover the global scene while still extracting local high-definition details through digital zoom, providing a hardware foundation for realizing "seeing the whole picture and seeing the details". Nowadays, physical world camera layout optimization, algorithm parameter tuning, emergency exercise cost high, long cycle, and difficult to predict the impact of new interference, and dynamic background interference such as leaf shaking, water reflection, light and shadow changes easily lead to false alarm, and the target is partially or severely obscured, which may result in missed detection. SUMMARY
[0003] The present application overcomes the shortcomings of the prior art and provides an anti-interference method and system for high-precision global snapshot, and a medium.
[0004] To achieve the above purpose, the technical solution adopted by the present application is as follows: The present application provides an anti-interference method for high-precision global snapshot, comprising the following steps: Collecting multi-source heterogeneous information in the monitoring area, and constructing a high-fidelity digital twin of the monitoring area based on digital twin technology; Obtaining a camera layout scheme for the monitoring area, simulating the coverage rate and blind area under the camera layout scheme in the monitoring area based on the high-fidelity digital twin of the monitoring area, and evaluating the snapshot performance under simulated interference conditions; Adjusting the layout scheme of the monitoring area dynamically according to the snapshot performance under simulated interference conditions and the coverage rate and blind area under the layout scheme in the monitoring area; Laying out based on the dynamically adjusted layout scheme of the monitoring area, constructing a snapshot model, and generating a composite image based on the snapshot model.
[0005] Further, in the anti-interference method for high-precision global snapshot, multi-source heterogeneous information in the monitoring area is collected, and a high-fidelity digital twin of the monitoring area is constructed based on digital twin technology, specifically: Collect three-dimensional geographic information, building information, traffic flow information, light weather information and camera data information in the monitoring area; According to the building information, traffic flow information, light weather information and camera data information, a building model, a traffic flow model, a light weather model and a virtual model of all cameras are constructed by using digital twin technology; The virtual model of the building model, the traffic flow model, the light weather model and all cameras is estimated and constructed by using the three-dimensional geographic information, forming a high-fidelity digital twin of the monitoring area.
[0006] Further, in the anti-interference method of high-precision global snapshot, the camera layout scheme of the monitoring area is obtained, and the coverage and blind area under the camera layout scheme of the monitoring area are simulated based on the high-fidelity digital twin of the monitoring area, specifically including: Obtain the camera layout scheme of the monitoring area, and obtain the initial coverage area of each camera according to the camera layout scheme of the monitoring area; Simulate the actual coverage area of the camera based on the high-fidelity digital twin of the monitoring area, and obtain the actual coverage area of each camera; According to the initial coverage area of each camera and the actual coverage area of each camera, the blind area of each camera is calculated.
[0007] Further, in the anti-interference method of high-precision global snapshot, the snapshot performance under the simulated interference is evaluated, specifically: Collect the camera performance data of the camera under each interference scene, set a plurality of interference scenes by using digital twin technology, and generate a camera twin simulation scene by using digital twin technology based on the interference scene and the camera performance data of the camera under each interference scene; Obtain the camera twin simulation scene of each timestamp, and construct a camera dynamic twin simulation scene according to the camera twin simulation scene of each timestamp; Obtain the historical interference factor data in the current target monitoring area, and obtain the maximum interference factor data according to the historical interference factor data in the current target monitoring area; Based on the maximum interference factor data input into the camera dynamic twin simulation scene for dynamic simulation, the snapshot performance data under the simulated interference is obtained.
[0008] Further, in the anti-interference method of high-precision global snapshot, the layout scheme of the monitoring area is dynamically adjusted according to the snapshot performance under the simulated interference and the coverage and blind area under the layout scheme of the monitoring area, specifically: updating the coverage rate and the blind area based on the coverage rate and the blind area under the layout scheme of the monitoring area in the case of simulating interference, and obtaining the updated coverage rate and blind area; introducing a genetic algorithm, setting the number of generations of the genetic algorithm, taking the updated coverage rate and blind area as the population, setting the threshold range of the coverage rate and blind area indicators, and performing genetic algorithm on the population generation by generation; obtaining the latest population, and judging whether the coverage rate and blind area corresponding to the latest population are within the threshold range of the coverage rate and blind area; when the coverage rate and blind area corresponding to the latest population are within the threshold range of the coverage rate and blind area, the layout scheme of the current monitoring area is laid out according to the layout scheme of the current monitoring area; when the coverage rate and blind area corresponding to the latest population are not within the threshold range of the coverage rate and blind area, the layout scheme of the current monitoring area is adjusted until the coverage rate and blind area corresponding to the latest population are within the threshold range of the coverage rate and blind area.
[0009] Further, in the anti-interference method of high-precision global snapshot, the layout scheme of the dynamically adjusted monitoring area is laid out, and a snapshot model is constructed, and a synthetic image is generated based on the snapshot model, specifically: based on the layout scheme of the dynamically adjusted monitoring area, a double-branch deep learning network is constructed, one branch is a spatio-temporal feature extraction branch, and a Transformer is used to capture the spatio-temporal consistency features between continuous frames, and the dynamic mode of the background is learned; the other branch is a salient target detection branch, which is dedicated to the static salient features of the current frame, and a joint attention mechanism is introduced, which can focus attention on the regions that violate the spatio-temporal consistency and the spatial salient regions in complex scenes, and suppress the response of the periodic dynamic background; an adversarial generative network is used for data enhancement, a generator generates various extreme occlusion, light and shadow interference, and weather interference synthetic images, and a discriminator continuously enhances its anti-interference and feature robustness in the confrontation with the generator, and forms a snapshot model; when in the case of extreme occlusion, light and shadow interference, and weather interference, the snapshot model is used to synthesize based on the currently collected images to form a synthetic image.
[0010] The second aspect of the application provides an anti-interference system for high-precision global snapshot, comprising a memory and a processor, the memory comprising an anti-interference method for high-precision global snapshot program, the anti-interference method for high-precision global snapshot program being executed by the processor to realize the steps of any one of the anti-interference methods for high-precision global snapshot.
[0011] The third aspect of the present application provides a computer readable storage medium comprising a high-precision global snapshot anti-interference method program, which, when executed by a processor, implements the steps of any of the high-precision global snapshot anti-interference methods.
[0012] The present application solves the defects in the background art and has the following beneficial effects: The present application collects multi-source heterogeneous information in the monitoring area, constructs a high-fidelity digital twin of the monitoring area based on digital twin technology, and then obtains a camera layout scheme for the monitoring area. Based on the high-fidelity digital twin of the monitoring area, the coverage and blind area under the camera layout scheme in the monitoring area are simulated, and the snapshot performance under simulated interference is evaluated. The layout scheme of the monitoring area is dynamically adjusted according to the snapshot performance under simulated interference and the coverage and blind area under the layout scheme in the monitoring area. Finally, the layout is performed based on the dynamically adjusted layout scheme of the monitoring area, a snapshot model is constructed, and a composite image is generated based on the snapshot model. The present application improves the precision of global snapshot by improving digital twin from visualization to simulation, prediction, and optimization, providing decision support and pre-rehearsal platform for physical snapshot systems. BRIEF DESCRIPTION OF DRAWINGS
[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings of other embodiments according to these drawings without creative labor.
[0014] Figure 1 The overall flowchart of the high-precision global snapshot anti-interference method is shown. Figure 2 The system block diagram of the high-precision global snapshot anti-interference system is shown. DETAILED DESCRIPTION
[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings of other embodiments according to these drawings without creative labor.
[0016] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the scope of protection of the present application is not limited by the specific embodiments disclosed below.
[0017] As Figure 1 shown, the first aspect of the present application provides an anti-interference method for high-precision global snapshot, comprising the following steps: Collecting multi-source heterogeneous information in the monitoring area, and constructing a high-fidelity digital twin of the monitoring area based on digital twin technology; Obtaining a camera layout scheme for the monitoring area, simulating the coverage and blind area under the camera layout scheme in the monitoring area based on the high-fidelity digital twin of the monitoring area, and evaluating the snapshot performance under simulated interference; According to the snapshot performance under simulated interference and the coverage and blind area under the layout scheme in the monitoring area, dynamically adjusting the layout scheme of the monitoring area; Based on the dynamically adjusted layout scheme of the monitoring area, the layout is performed, and a snapshot model is constructed, and a synthetic image is generated based on the snapshot model.
[0018] Further, in the anti-interference method for high-precision global snapshot, multi-source heterogeneous information in the monitoring area is collected, and a high-fidelity digital twin of the monitoring area is constructed based on digital twin technology, specifically: Collecting three-dimensional geographic information, building information, traffic flow information, lighting and meteorological information, and camera data information in the monitoring area; It should be noted that the three-dimensional geographic information includes the longitude and latitude information of the building, the longitude and latitude information of the position of the camera, etc., the traffic flow information includes the real-time traffic condition, the number of vehicles driving information, etc., and the camera data information includes the model, the parameter, the field of view angle and other parameters.
[0019] According to the building information, traffic flow information, lighting and meteorological information, and camera data information, a building model, a traffic flow model, a lighting and meteorological model, and a virtual model of all cameras are constructed by using digital twin technology; The three-dimensional geographic information is used to construct the estimated real model of the building model, the traffic flow model, the lighting and meteorological model, and the virtual model of all cameras, forming a high-fidelity digital twin of the monitoring area.
[0020] It should be noted that in the digital twin environment, a large number of people, vehicles and their behaviors are imported or generated, and various scenes such as heavy rain, thick fog, night, construction shielding, etc. are simulated. The same snapshot algorithm as the actual system is run in the virtual environment to perform sand table deduction. Thus, the coverage and blind area under different camera layout schemes can be quickly tested; the algorithm parameters can be adjusted to evaluate their performance under simulated interference.
[0021] Further, in the anti-interference method of high-precision global snapshot, the camera layout scheme of the monitoring area is obtained, and the coverage and blind area of the camera layout scheme of the monitoring area are simulated based on the high-fidelity digital twin of the monitoring area. Specifically, it comprises: Obtaining the camera layout scheme of the monitoring area, and obtaining the initial coverage area of each camera according to the camera layout scheme of the monitoring area; Simulating the actual coverage area of the camera based on the high-fidelity digital twin of the monitoring area, and obtaining the actual coverage area of each camera; According to the initial coverage area of each camera and the actual coverage area of each camera, the blind area of each camera is calculated.
[0022] It should be noted that the actual coverage area of each camera and the blind area of the camera can be obtained by the method.
[0023] Further, in the anti-interference method of high-precision global snapshot, the snapshot performance under simulated interference is evaluated, specifically: Collecting camera performance data of the camera under each interference scene, setting a plurality of interference scenes by using digital twin technology, generating camera twin simulation scenes based on the interference scenes and the camera performance data of the camera under each interference scene by using digital twin technology; It should be noted that the interference scene includes rainstorm, heavy fog, night, construction shielding and the like.
[0024] Obtaining the camera twin simulation scene of each timestamp, and constructing the camera dynamic twin simulation scene according to the camera twin simulation scene of each timestamp; Obtaining historical interference factor data in the current target monitoring area, and obtaining the maximum interference factor data according to the historical interference factor data in the current target monitoring area; Based on the maximum interference factor data input into the camera dynamic twin simulation scene for dynamic simulation, the snapshot performance data under simulated interference is obtained.
[0025] It should be noted that the snapshot performance data includes clarity, contrast, resolution and the like, and the maximum interference factor data is obtained according to the historical interference factor data in the current target monitoring area, so that the maximum interference factor data is input into the camera dynamic twin simulation scene for dynamic simulation, so that high-precision snapshot can be realized under any interference factor data (such as severe weather, such as rainstorm day, haze day, etc.), Further, in the anti-interference method of high-precision global snapshot, the layout scheme of the monitoring area is dynamically adjusted according to the snapshot performance under the simulated interference condition and the coverage and blind area under the layout scheme of the monitoring area, specifically: The coverage and blind area under the layout scheme of the monitoring area are updated based on the snapshot performance under the simulated interference condition, and the updated coverage and blind area are obtained; The genetic algorithm is introduced, the genetic generation number is set, the updated coverage and blind area are used as the population, the coverage and blind area threshold index range is set, and the population is genetically one generation after another; The latest population is obtained, and it is judged whether the coverage and blind area corresponding to the latest population are within the coverage and blind area threshold index range; When the coverage and blind area corresponding to the latest population are within the coverage and blind area threshold index range, the layout scheme of the current monitoring area is laid out; When the coverage and blind area corresponding to the latest population are not within the coverage and blind area threshold index range, the layout scheme of the current monitoring area is adjusted until the coverage and blind area corresponding to the latest population are within the coverage and blind area threshold index range.
[0026] It should be noted that the layout scheme of the monitoring area can be optimized by the method, so as to further improve the high-precision snapshot of the target monitoring area.
[0027] Further, in the anti-interference method of high-precision global snapshot, the layout scheme of the monitoring area is dynamically adjusted based on the dynamically adjusted layout scheme of the monitoring area, and a snapshot model is constructed, and a synthetic image is generated based on the snapshot model, specifically: The layout scheme of the monitoring area is laid out based on the dynamically adjusted layout scheme of the monitoring area, and a double-branch deep learning network is constructed, one branch is a spatio-temporal feature extraction branch, and the other branch is a salient target detection branch. The Transformer is used to capture the spatio-temporal consistency features between consecutive frames and learn the dynamic mode of the background; The joint attention mechanism is introduced to focus attention on the area that violates the spatio-temporal consistency and the spatial salient area in the complex scene, and the response of the periodic dynamic background is suppressed; When in the case of extreme occlusion, light and shadow interference, weather interference, the snapshot model is used to synthesize based on the current collected image to form a synthesized image.
[0028] It should be noted that when the dynamic background interference such as leaf shaking, water wave reflection, light and shadow change easily leads to false alarm and the phenomenon of missing detection when the target is partially or severely occluded, a double-branch deep learning network is designed. One branch is the "spatiotemporal feature extraction branch", which uses 3D CNN or Transformer to capture the spatiotemporal consistency features between consecutive frames, and learns the dynamic mode of the background (such as the period of shaking). The other branch is the "salient target detection branch", which focuses on the static salient features of the current frame. The joint attention mechanism is introduced, which enables the network to learn to "focus" on the areas that violate the spatiotemporal consistency (i.e. moving targets) and spatially salient areas in complex scenes, and suppress the response of periodic dynamic backgrounds. An adversarial generative network (GAN) is used for data augmentation. The generator generates various synthetic images of extreme occlusion, light and shadow interference, and weather interference, and the discriminator (i.e. the snapshot model) continuously enhances its anti-interference and feature robustness in the confrontation with the generator.
[0029] It should be noted that by regarding the background as a learnable dynamic pattern rather than noise, and actively attacking the model's weaknesses through adversarial training, the method can develop strong anti-interference ability, thereby avoiding the phenomenon of false alarm caused by dynamic background interference such as leaf shaking, water wave reflection, light and shadow change, and missing detection when the target is partially or severely occluded.
[0030] In addition, the method also includes: Using 5G network slicing and multi-access edge computing technology, a camera end side, edge side and cloud side are constructed, a lightweight model is run in the camera end side, the lightweight model is responsible for preliminary detection and region interception, and the video stream and metadata of the region of interest are transmitted; The edge side receives ROI streams from multiple end devices, runs high-precision recognition, trajectory association and relay tracking algorithms, and exchanges data between edge nodes through 5G links with a preset time delay; The cloud side is responsible for large-scale data storage, long-term trajectory mining, global model training and distribution; At the same time, according to the network congestion state, edge server load and task urgency, the computing task is placed on the camera end side, edge side or cloud side to optimize the transmission path.
[0031] It should be noted that the slice of the communication network and the calculation offloading depth are integrated into the snapshot business flow, the joint optimization of network resources and calculation resources is realized, and the real-time of global linkage is ensured. Wherein according to the network congestion state, the edge server load and the task urgency, the calculation task is placed at the camera end side, the edge side or the cloud side, such as when the network congestion state of the camera end side is not ideal, it can be placed at the edge side or the cloud side, and such as when the data of the camera end side is relatively urgent, the cloud side, so as to facilitate tracking.
[0032] In addition, the method further comprises: When the target monitoring area is a privacy area, deploying a commercial Wi-Fi or 5G small base station device in the privacy area, and analyzing channel state information by using the commercial Wi-Fi or 5G small base station device; According to the channel state information, the movement, breathing and micro-motion of the human body are analyzed, and whether an abnormal behavior occurs in the area is judged; When it is perceived that an abnormal behavior occurs, the camera outside the privacy area is linked to focus on controlling and snapshotting the exit and the channel; When it is perceived that no abnormal behavior occurs, the data acquisition frequency of the commercial Wi-Fi or 5G small base station device is reduced.
[0033] It should be noted that the commercial Wi-Fi or 5G small base station device is deployed inside the privacy area, but it is not used for communication, but is used for analyzing channel state information (CSI), and the CSI is extremely sensitive to the movement, breathing and micro-motion of the human body in the environment. By using an AI model to analyze the disturbance mode of the CSI, whether there is a person, the number of people and whether an abnormal behavior such as falling and violent struggle occurs in the area can be judged without obtaining any visual image. The method solves the contradiction between privacy and security, and provides effective supplementary information for the visual blind area.
[0034] As shown in Figure 2 The second aspect of the application provides an anti-interference system for high-precision global snapshotting, comprising a memory and a processor, the memory comprising an anti-interference method program for high-precision global snapshotting, and the anti-interference method program for high-precision global snapshotting being executed by the processor to realize the steps of any one of the anti-interference methods for high-precision global snapshotting.
[0035] The third aspect of the application provides a computer-readable storage medium comprising an anti-interference method program for high-precision global snapshotting, and the anti-interference method program for high-precision global snapshotting being executed by a processor to realize the steps of any one of the anti-interference methods for high-precision global snapshotting.
[0036] It should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative, for example, the division of the units is merely a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the various components shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0037] The units described above as separate components can or can not be physically separate, and the components shown as units can or can not be physical units; they can be located in one place or distributed on multiple network units; and some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.
[0038] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be realized in the form of hardware or hardware plus software functional unit.
[0039] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the above-mentioned program can be stored in a computer readable storage medium, and the program executes the steps including the above-mentioned method embodiments when executed; and the above-mentioned storage medium includes mobile storage device, read-only memory (ROM), random access memory (RAM), magnetic disc or optical disc and various storage program codes.
[0040] Alternatively, the above-mentioned integrated unit of the present application can be stored in a computer readable storage medium if it is realized in the form of software functional module and sold or used as an independent product. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of software product, which is stored in a storage medium and includes a plurality of instructions for making a computer device (which can be a personal computer, a server or a network device) execute all or part of the method of the embodiments of the present application. The above-mentioned storage medium includes mobile storage device, ROM, RAM, magnetic disc or optical disc and various storage program codes.
[0041] The above merely illustrates the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A high-precision global image capture anti-interference method, characterized in that, Includes the following steps: Collect multi-source heterogeneous information in the monitoring area and construct a high-fidelity digital twin of the monitoring area based on digital twin technology; Obtain the camera deployment scheme of the monitoring area, simulate the coverage and blind spots under the camera deployment scheme based on the high-fidelity digital twin of the monitoring area, and evaluate the capture performance under simulated interference conditions. Based on the capture performance under simulated interference conditions and the coverage and blind spot dynamic adjustment scheme of the monitoring area under the deployment scheme of the monitoring area; The monitoring area is deployed based on a dynamically adjusted deployment plan, and a capture model is constructed. A synthetic image is then generated based on the capture model.
2. The anti-interference method for high-precision global image capture according to claim 1, characterized in that, Collect multi-source heterogeneous information from the monitoring area and construct a high-fidelity digital twin of the monitoring area based on digital twin technology, specifically: Collect 3D geographic information, building information, traffic flow information, lighting and meteorological information, and camera data in the monitored area; Based on the building information, traffic flow information, lighting and meteorological information, and camera data, digital twin technology is used to construct building models, traffic flow models, lighting and meteorological models, and virtual models of all cameras; Using the three-dimensional geographic information, a preliminary on-site model is constructed based on the building model, traffic flow model, lighting and meteorological model, and virtual models of all cameras, forming a high-fidelity digital twin of the monitored area.
3. The anti-interference method for high-precision global image capture according to claim 1, characterized in that, Obtain the camera deployment plan for the monitored area, and simulate the coverage and blind spots under the camera deployment plan based on a high-fidelity digital twin of the monitored area, specifically including: Obtain the camera deployment plan for the monitored area, and obtain the initial coverage area of each camera based on the camera deployment plan for the monitored area; The actual coverage area of the cameras is simulated using a high-fidelity digital twin of the monitored area to obtain the actual coverage area of each camera. The blind zone of each camera is calculated based on the initial coverage area and the actual coverage area of each camera.
4. The anti-interference method for high-precision global image capture according to claim 1, characterized in that, The performance of image capture under simulated interference conditions was evaluated, specifically as follows: Collect camera performance data under various interference scenarios, set up several interference scenarios using digital twin technology, and dynamically simulate the interference scenarios and camera performance data under each interference scenario using digital twin technology to generate camera twin simulation scenarios. Obtain the camera twin simulation scene for each timestamp, and construct a dynamic camera twin simulation scene based on the camera twin simulation scene for each timestamp; Obtain historical interference factor data in the current target monitoring area, and obtain the maximum interference factor data based on the historical interference factor data in the current target monitoring area; The maximum interference factor data is input into the camera's dynamic twin simulation scene to perform dynamic simulation and obtain capture performance data under simulated interference conditions.
5. The anti-interference method for high-precision global image capture according to claim 1, characterized in that, Based on the image capture performance under simulated interference conditions, the coverage rate under the deployment scheme in the monitoring area, and the dynamic adjustment scheme for the monitoring area based on blind spots, the specific details are as follows: Based on the capture performance under simulated interference, the coverage and blind spots of the monitoring area under the deployment scheme are updated to obtain the updated coverage and blind spots. A genetic algorithm is introduced, and the genetic generation is set using the genetic algorithm. The updated coverage and blind zone are used as the population. The coverage and blind zone threshold index ranges are set, and the population is genetically inherited generation by generation. Obtain the latest generation of population and determine whether the coverage and blind zone corresponding to the latest generation of population are within the range of the coverage and blind zone threshold indicators. When the coverage and blind zone corresponding to the latest generation of population are within the range of the coverage and blind zone threshold indicators, the current monitoring area shall be deployed according to the current deployment plan. When the coverage and blind zone corresponding to the latest generation of the population are not within the range of the coverage and blind zone threshold indicators, the deployment scheme of the current monitoring area is adjusted until the coverage and blind zone corresponding to the latest generation of the population are within the range of the coverage and blind zone threshold indicators.
6. The anti-interference method for high-precision global image capture according to claim 1, characterized in that, The monitoring area is deployed according to a dynamically adjusted layout plan, and a capture model is constructed. A composite image is then generated based on the capture model. Specifically: Based on the dynamically adjusted deployment scheme of the monitoring area, a dual-branch deep learning network is constructed. One branch is for spatiotemporal feature extraction, and a Transformer is used to capture the spatiotemporal consistency features between consecutive frames to learn the dynamic patterns of the background. The other branch is the salient object detection branch, which is dedicated to the static salient features of the current frame. It introduces a joint attention mechanism, and the dual-branch deep learning network learns to focus attention on regions that violate spatiotemporal consistency and spatially salient regions in complex scenes, and suppress the response of periodic dynamic background. Data augmentation is achieved by using a generative adversarial network. The generator produces synthetic images with various extreme occlusions, light and shadow interferences, and weather interferences. The discriminator continuously strengthens its anti-interference and feature robustness in adversarial interactions with the generator, thus forming a capture model. When faced with extreme occlusion, light and shadow interference, or weather interference, the capture model uses the captured image to synthesize a composite image based on the currently acquired image.
7. A high-precision global capture anti-interference system, characterized in that, The device includes a memory and a processor. The memory includes a program for an anti-interference method for high-precision global capture. When the processor executes the program for the anti-interference method for high-precision global capture, it implements the steps of the anti-interference method for high-precision global capture as described in any one of claims 1-6.
8. A computer-readable storage medium, characterized in that, The method includes a high-precision global capture anti-interference method program, which, when executed by a processor, implements the steps of the high-precision global capture anti-interference method as described in any one of claims 1-6.