Suspicious vehicle path tracking method and system based on space-time reverse retrieval

Abstract

The invention provides a suspicious vehicle path tracking method and system based on space-time reverse retrieval. The method comprises the following steps: acquiring road network structure information related to a preset toll station and vehicle feature library information of a target suspicious vehicle, extracting path branches according to the road network structure information, matching path priorities, acquiring road section scene data corresponding to each path branch, processing to obtain vehicle speed probability density and a corrected time window, and determining a target suspicious vehicle according to the vehicle speed probability density. The method comprises the following steps: acquiring video data, processing in combination with vehicle feature library information to obtain a vehicle matching degree, and acquiring a final complete path in combination with the vehicle matching degree according to a corrected time window, thereby realizing a suspicious vehicle path tracking technology based on time-space reverse retrieval.

Description

Technical Field

[0001] The present application relates to the field of path tracking technology, and in particular to a method and system for tracking the path of a suspicious vehicle based on spatiotemporal reverse retrieval. Background Art

[0002] Tracking the paths of suspicious vehicles is crucial in traffic management and case investigation. Traditional vehicle path tracking methods often rely on manual video surveillance, which is inefficient and prone to omissions or misjudgments due to human factors. With the development of intelligent transportation systems, some tracking methods based on video analysis and road network information have emerged. However, these methods have difficulty accurately extracting effective path information when faced with complex road networks and massive amounts of data, and the matching of vehicle features and control of time windows are not accurate enough, resulting in unsatisfactory tracking results.

[0003] Most existing path tracking technologies track forward from the starting point to the end point. When the driving trajectory of a suspicious vehicle is unclear, it is easy to fall into the dilemma of too many path branches, which increases the difficulty and workload of tracking. The method based on spatiotemporal reverse retrieval, which tracks backward from known key nodes such as toll stations, can effectively narrow the tracking scope and improve tracking efficiency. However, the current related technologies have shortcomings in path branch priority matching, time window correction, and vehicle matching combination, and need further optimization.

[0004] In response to the above problems, effective technical solutions are urgently needed. Summary of the Invention

[0005] The purpose of this application is to provide a suspicious vehicle path tracking method and system based on spatiotemporal reverse retrieval. The method and system can obtain road network structure information related to preset toll stations and vehicle feature library information of target suspicious vehicles, extract path branches and match path priorities according to the road network structure information, obtain road section scene data corresponding to each path branch, process to obtain vehicle speed probability density and corrected time window, obtain video data, combine the vehicle feature library information to obtain vehicle matching degree, and obtain the final complete path according to the corrected time window combined with the vehicle matching degree, thereby realizing a technology for suspicious vehicle path tracking based on spatiotemporal reverse retrieval.

[0006] This application also provides a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval, comprising the following steps: Obtain road network structure information related to preset toll stations and vehicle feature database information of target suspicious vehicles; Extracting path branches and matching path priorities according to the road network structure information; Obtaining the road scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; Acquire video data and process it in combination with the vehicle feature database information to obtain a vehicle matching degree; The final complete path is obtained according to the corrected time window and the vehicle matching degree.

[0007] Optionally, in the suspicious vehicle path tracking method based on spatiotemporal reverse retrieval described in the present application, obtaining the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

[0008] Optionally, in the suspicious vehicle path tracking method based on spatiotemporal reverse retrieval described in the present application, the step of extracting path branches and matching path priorities based on the road network structure information includes: The path branches are obtained by processing the destination location, destination time, section length, confluence information, and connection relationship information through a preset path prediction model; Obtaining historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; Priority ranking is performed on each path branch according to the historical travel data to obtain a priority corresponding to each path branch.

[0009] Optionally, in the suspicious vehicle path tracking method based on spatiotemporal reverse retrieval described in the present application, obtaining the road segment scene data corresponding to each path branch and processing to obtain the vehicle speed probability density and the corrected time window include: Obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, real-time traffic speed data, and real-time weather data; Obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; Processing is performed according to the length of the branch road section, the vehicle speed probability density and the terminal time to obtain a time window; The time window is corrected according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

[0010] Optionally, in the suspicious vehicle path tracking method based on spatiotemporal reverse retrieval described in the present application, the acquiring of video data and processing of the vehicle feature database information to obtain a vehicle matching degree may include: Obtain video data from monitoring points on each branch of the route and extract the feature database information of the vehicle to be verified; Corresponding feature matching is performed respectively according to the to-be-verified vehicle feature library information and the vehicle feature library information, and weighted processing is performed to obtain a vehicle matching degree.

[0011] Optionally, in the suspicious vehicle path tracking method based on spatiotemporal reverse retrieval described in the present application, obtaining the final complete path based on the corrected time window in combination with the vehicle matching degree includes: Performing a threshold comparison based on the vehicle matching degree and a preset vehicle matching threshold; If the vehicle matching degree is greater than the preset vehicle matching threshold, the vehicle matching degree is marked as the target matching degree, and the corresponding target monitoring point position and time are obtained; The final complete path is generated according to the corrected time window combined with the target monitoring point position and time.

[0012] In a second aspect, the present application provides a suspicious vehicle path tracking system based on spatiotemporal reverse retrieval, the system comprising: a memory and a processor, the memory comprising a program for a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval, the program for a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval, when executed by the processor, implementing the following steps: Obtain road network structure information related to preset toll stations and vehicle feature database information of target suspicious vehicles; Extracting path branches and matching path priorities according to the road network structure information; Obtaining the road scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; Acquire video data and process it in combination with the vehicle feature database information to obtain a vehicle matching degree; The final complete path is obtained according to the corrected time window and the vehicle matching degree.

[0013] Optionally, in the suspicious vehicle path tracking system based on spatiotemporal reverse retrieval described in the present application, the step of obtaining the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

[0014] Optionally, in the suspicious vehicle path tracking system based on spatiotemporal reverse retrieval described in the present application, the step of extracting path branches and matching path priorities according to the road network structure information includes: The path branches are obtained by processing the destination location, destination time, section length, confluence information, and connection relationship information through a preset path prediction model; Obtaining historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; Priority ranking is performed on each path branch according to the historical travel data to obtain a priority corresponding to each path branch.

[0015] Optionally, in the suspicious vehicle path tracking system based on spatiotemporal reverse retrieval described in the present application, the step of obtaining the road scene data corresponding to each path branch and processing the data to obtain the vehicle speed probability density and the corrected time window includes: Obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, real-time traffic speed data, and real-time weather data; Obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; Processing is performed according to the length of the branch road section, the vehicle speed probability density and the terminal time to obtain a time window; The time window is corrected according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

[0016] From the above, it can be seen that the suspicious vehicle path tracking method and system based on spatiotemporal reverse retrieval provided by the present application obtains the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle, extracts the path branches and matches the path priority according to the road network structure information, obtains the road section scene data corresponding to each path branch, processes to obtain the vehicle speed probability density and the corrected time window, obtains video data, combines the vehicle feature library information to obtain the vehicle matching degree, and obtains the final complete path according to the corrected time window combined with the vehicle matching degree, thereby realizing the technology of suspicious vehicle path tracking based on spatiotemporal reverse retrieval.

[0017] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or understood by practicing the embodiments of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 A flowchart of a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval provided in an embodiment of the present application; Figure 2 A flowchart of matching path priorities of a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval provided in an embodiment of the present application; Figure 3 A flowchart of obtaining vehicle speed probability density and a corrected time window for a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval provided in an embodiment of the present application; Figure 4 A flowchart of obtaining vehicle matching degree in a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0021] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0022] Please refer to Figure 1 , Figure 1 This is a flow chart of a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval in some embodiments of the present application. This suspicious vehicle path tracking method based on spatiotemporal reverse retrieval is used in a terminal device, such as a computer, mobile phone terminal, etc. This suspicious vehicle path tracking method based on spatiotemporal reverse retrieval includes the following steps: S11. Obtaining road network structure information related to a preset toll station and vehicle feature database information of a target suspicious vehicle; S12, extracting path branches and matching path priorities according to the road network structure information; S13, obtaining the road section scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; S14, obtaining video data, and processing the data in combination with the vehicle feature database information to obtain a vehicle matching degree; S15. Obtain a final complete path based on the corrected time window and vehicle matching.

[0023] It should be noted that when a suspicious vehicle A is spotted at a toll station exit on a highway, its travel path needs to be determined because there are multiple paths in the road network and multiple toll stations leading to the current toll station. Conventional video retrieval methods require processing massive amounts of video data. Therefore, to address these issues, we first obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle. The road network structure information includes the destination location, destination time, section length, confluence information, and connection relationship information. The vehicle feature library information includes the license plate number, vehicle model, color, and wheel and tire model. Next, we extract path branches based on the road network structure information and match them with their priorities. We then obtain the corresponding road segment scene data for each path branch, including branch segment length, historical speed distribution, real-time traffic speed data, and real-time weather data. This data is then processed to obtain the vehicle speed probability density and the corrected time window. Video data is then acquired and combined with the vehicle feature library information to obtain the vehicle matching degree. The final complete path is then obtained based on the corrected time window and the vehicle matching degree, thus implementing a technique for tracking suspicious vehicle paths based on spatiotemporal reverse retrieval.

[0024] According to an embodiment of the present invention, obtaining the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

[0025] It should be noted that when a target suspicious vehicle is detected at a preset toll station, the information collection mechanism must be activated immediately to fully obtain two types of key data: the road network structure information associated with the preset toll station and the feature library information of the target suspicious vehicle, to provide an accurate basis for subsequent tracking and interception. Among them, the road network structure information covers multi-dimensional core elements: the terminal position clearly defines the specific location of the vehicle when it is discovered, which defines the boundary of the control range; the terminal time is the specific time when the suspicious vehicle passes through the toll station. Combined with the driving path, the reasonable travel time can be calculated to assist in judging whether the vehicle has abnormal stops or detours; the road section length is directly related to the road grade. Traffic efficiency can be estimated at each The driving time of the road section; the confluence information marks all possible branch intersections along the way to avoid missing paths during the tracking process; the connection relationship information clearly presents the connection logic of each road section, providing a topological basis for analyzing the lane change and turning routes that the vehicle may choose; and in the vehicle feature library information, the license plate number is the primary identifier for identifying the vehicle, which can be quickly associated with the vehicle registration information; the vehicle model (such as sedans, trucks, SUVs, etc.) determines its access restrictions and road selection preferences; the body color facilitates rapid positioning in complex traffic; the style, size and tire model of the vehicle's wheels as unique details can provide auxiliary identification basis when the license plate is blocked or tampered with, greatly improving tracking recognition.

[0026] Please refer to Figure 2 , Figure 2 This is a flow chart of matching path priorities in a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval in some embodiments of the present application. According to an embodiment of the present invention, extracting path branches based on the road network structure information and matching path priorities includes: S21, processing the destination location, destination time, section length, confluence information, and connection relationship information using a preset path prediction model to obtain each path branch; S22. Obtain historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; S23. Prioritize each path branch according to the historical driving data to obtain a priority corresponding to each path branch.

[0027] It should be noted that after obtaining the road network structure information of the preset toll station, the preset path prediction model is used to comprehensively process the destination location, destination time, road section length, confluence point information, and connection relationship information. By analyzing the spatial correlation and traffic constraints of each road section, the model can generate all possible path branches from the current location to the destination, covering main roads, secondary roads, and alternative routes derived from each confluence point, providing a complete path reference framework for subsequent tracking. At the same time, the historical driving data of the target suspicious vehicle must be retrieved. The core information includes branch selection records and the historical driving frequency corresponding to each path branch. Branch selection records can intuitively demonstrate the vehicle's past preference for different forks in similar road networks, such as whether it tends to turn left or go straight at a specific confluence point. The historical driving frequency quantitatively reflects the vehicle's reliance on a specific path, such as the proportion of times a provincial highway is chosen between the same starting and ending points reaching 60%. Based on this historical data, the generated path branches can be prioritized. The sorting logic is centered on historical selection patterns, marking paths that have been frequently selected by the vehicle in the past as high priority, paths that have been selected occasionally as medium priority, and paths that have never been selected as low priority. Through this sorting, tracking efficiency can be greatly improved, and law enforcement resources can be focused on the paths most likely to be selected.

[0028] Please refer to Figure 3 , Figure 3 This is a flowchart of obtaining the vehicle speed probability density and the corrected time window of a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval in some embodiments of the present application. According to an embodiment of the present invention, obtaining the road segment scene data corresponding to each path branch and processing to obtain the vehicle speed probability density and the corrected time window includes: S31, obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, the real-time traffic speed data, and the real-time weather data; S32, obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; S33, processing according to the branch road section length, vehicle speed probability density and terminal time to obtain a time window; S34. Correct the time window according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

[0029] It should be noted that in order to accurately determine the travel time of the target suspicious vehicle at each path branch, it is necessary to first obtain the corresponding road section scene data, including the length of the branch section, the historical speed distribution of the section, real-time traffic speed data and real-time weather data; based on the historical speed distribution of the section, the speed probability density can be obtained through statistical analysis. This data can reflect the possibility of different speeds appearing on the section. For example, in a certain suburban road section, the probability of a speed of 60-80 kilometers per hour is 70%, providing a probabilistic basis for predicting the travel time; then, the branch section length, speed probability density and terminal time are combined for calculation and processing to obtain the time window. Specifically, by dividing the length of the road section by the vehicle speed under different probabilities, a variety of possible travel times are obtained, and then combined with the end time to reversely calculate, the time range for vehicles to pass through the road section is framed to form an initial time window. Since real-time road conditions and weather will significantly affect the travel efficiency, the initial time window needs to be corrected using real-time traffic speed data and real-time weather data; for example, when the real-time traffic speed is 30% lower than the historical average, the time window needs to be extended; in case of heavy rain, the time range is further expanded. For example, the initial time window is set to 30-40 minutes. If it rains suddenly, In the event of heavy rain, the window needs to be adjusted to 45-60 minutes based on the vehicle speed attenuation model of historical rainy days; if there is heavy fog with visibility less than 50 meters, an elastic interval needs to be added to the window, marked as "may be delayed by 15-30 minutes"; during the correction process, a two-factor weight model needs to be established: when real-time traffic flow and weather data affect each other in the same direction (such as congestion + heavy rain), the window is superimposed and extended; when the data act in the opposite direction (such as smooth traffic + light rain), the window range is fine-tuned based on the traffic speed, and finally a more realistic corrected time window is obtained, providing a time basis for precise control.

[0030] Please refer to Figure 4 , Figure 4 This is a flow chart of obtaining a vehicle matching degree for a suspicious vehicle path tracking method based on spatiotemporal reverse retrieval in some embodiments of the present application. According to an embodiment of the present invention, obtaining video data and processing the vehicle feature database information to obtain a vehicle matching degree includes: S41, obtaining video data from monitoring points at each branch of the route and extracting information from a feature library of the vehicle to be verified; S42: performing corresponding feature matching on the to-be-verified vehicle feature library information and the vehicle feature library information, and performing weighted processing to obtain a vehicle matching degree.

[0031] It should be noted that the system collects video data from monitoring points on each path branch in sequence according to the priority of the path branches, and uses image recognition technology to extract feature information of the vehicle to be verified, including parameters such as license plate number, vehicle model, color, vehicle wheel and tire model; then, the feature library of the vehicle to be verified is matched dimension by dimension with the feature library of the target suspicious vehicle, where the license plate number is accurately matched by characters, the vehicle model is matched by contour and parameter library, the color is calibrated by spectral analysis, and the wheel and tire model is identified by combining pattern texture and size data; after the matching is completed, weights are assigned according to the importance of the features, for example, the license plate number accounts for 0.3, the vehicle model and color each account for 0.2, and the wheel and tire model account for 0.3 in total; the vehicle matching degree in the range of 0-100 is calculated through a weighted sum algorithm to achieve accurate quantitative comparison of multiple feature dimensions.

[0032] According to an embodiment of the present invention, obtaining a final complete path based on the corrected time window and vehicle matching degree includes: Performing a threshold comparison based on the vehicle matching degree and a preset vehicle matching threshold; If the vehicle matching degree is greater than the preset vehicle matching threshold, the vehicle matching degree is marked as the target matching degree, and the corresponding target monitoring point position and time are obtained; The final complete path is generated according to the corrected time window combined with the target monitoring point position and time.

[0033] It should be noted that the system will accurately compare the calculated matching degree of each vehicle with the preset vehicle matching threshold one by one; this threshold is set based on historical data and recognition accuracy requirements, and is used to filter out vehicle information that meets the basic matching standards; when the matching degree of a vehicle exceeds the preset threshold, the system will automatically mark it as the target matching degree, and quickly extract the target monitoring point location information corresponding to the matching result and the time data of the vehicle passing through the monitoring point; this information will be temporarily stored as key node data for path generation; then, the system calls the corrected time window parameters, combined with the target monitoring point location and time obtained previously, the system uses time series analysis and spatial association algorithms to connect each node in chronological order, and finally generates a complete and continuous driving path for the target suspicious vehicle.

[0034] In a second aspect, the present invention further discloses a suspicious vehicle path tracking system based on spatiotemporal reverse retrieval, comprising a memory and a processor, wherein the memory includes a suspicious vehicle path tracking method program based on spatiotemporal reverse retrieval, and when the suspicious vehicle path tracking method program based on spatiotemporal reverse retrieval is executed by the processor, the following steps are implemented: Obtain road network structure information related to preset toll stations and vehicle feature database information of target suspicious vehicles; Extracting path branches and matching path priorities according to the road network structure information; Obtaining the road scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; Acquire video data and process it in combination with the vehicle feature database information to obtain a vehicle matching degree; The final complete path is obtained according to the corrected time window and the vehicle matching degree.

[0035] It should be noted that when a suspicious vehicle A is spotted at a toll station exit on a highway, its travel path needs to be determined because there are multiple paths in the road network and multiple toll stations leading to the current toll station. Conventional video retrieval methods require processing massive amounts of video data. Therefore, to address these issues, we first obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle. The road network structure information includes the destination location, destination time, section length, confluence information, and connection relationship information. The vehicle feature library information includes the license plate number, vehicle model, color, and wheel and tire model. Next, we extract path branches based on the road network structure information and match them with their priorities. We then obtain the corresponding road segment scene data for each path branch, including branch segment length, historical speed distribution, real-time traffic speed data, and real-time weather data. This data is then processed to obtain the vehicle speed probability density and the corrected time window. Video data is then acquired and combined with the vehicle feature library information to obtain the vehicle matching degree. The final complete path is then obtained based on the corrected time window and the vehicle matching degree, thus implementing a technique for tracking suspicious vehicle paths based on spatiotemporal reverse retrieval.

[0036] According to an embodiment of the present invention, obtaining the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

[0037] It should be noted that when a target suspicious vehicle is detected at a preset toll station, the information collection mechanism must be activated immediately to fully obtain two types of key data: the road network structure information associated with the preset toll station and the feature library information of the target suspicious vehicle, to provide an accurate basis for subsequent tracking and interception. Among them, the road network structure information covers multi-dimensional core elements: the terminal position clearly defines the specific location of the vehicle when it is discovered, which defines the boundary of the control range; the terminal time is the specific time when the suspicious vehicle passes through the toll station. Combined with the driving path, the reasonable travel time can be calculated to assist in judging whether the vehicle has abnormal stops or detours; the road section length is directly related to the road grade. Traffic efficiency can be estimated at each The driving time of the road section; the confluence information marks all possible branch intersections along the way to avoid missing paths during the tracking process; the connection relationship information clearly presents the connection logic of each road section, providing a topological basis for analyzing the lane change and turning routes that the vehicle may choose; and in the vehicle feature library information, the license plate number is the primary identifier for identifying the vehicle, which can be quickly associated with the vehicle registration information; the vehicle model (such as sedans, trucks, SUVs, etc.) determines its access restrictions and road selection preferences; the body color facilitates rapid positioning in complex traffic; the style, size and tire model of the vehicle's wheels as unique details can provide auxiliary identification basis when the license plate is blocked or tampered with, greatly improving tracking recognition.

[0038] According to an embodiment of the present invention, extracting path branches and matching path priorities according to the road network structure information includes: The path branches are obtained by processing the destination location, destination time, section length, confluence information, and connection relationship information through a preset path prediction model; Obtaining historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; Priority ranking is performed on each path branch according to the historical travel data to obtain a priority corresponding to each path branch.

[0039] It should be noted that after obtaining the road network structure information of the preset toll station, the preset path prediction model is used to comprehensively process the destination location, destination time, road section length, confluence point information, and connection relationship information. By analyzing the spatial correlation and traffic constraints of each road section, the model can generate all possible path branches from the current location to the destination, covering main roads, secondary roads, and alternative routes derived from each confluence point, providing a complete path reference framework for subsequent tracking. At the same time, the historical driving data of the target suspicious vehicle must be retrieved. The core information includes branch selection records and the historical driving frequency corresponding to each path branch. Branch selection records can intuitively demonstrate the vehicle's past preference for different forks in similar road networks, such as whether it tends to turn left or go straight at a specific confluence point. The historical driving frequency quantitatively reflects the vehicle's reliance on a specific path, such as the proportion of times a provincial highway is chosen between the same starting and ending points reaching 60%. Based on this historical data, the generated path branches can be prioritized. The sorting logic is centered on historical selection patterns, marking paths that have been frequently selected by the vehicle in the past as high priority, paths that have been selected occasionally as medium priority, and paths that have never been selected as low priority. Through this sorting, tracking efficiency can be greatly improved, and law enforcement resources can be focused on the paths most likely to be selected.

[0040] According to an embodiment of the present invention, the step of obtaining the road scene data corresponding to each path branch and processing the data to obtain the vehicle speed probability density and the corrected time window includes: Obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, real-time traffic speed data, and real-time weather data; Obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; Processing is performed according to the length of the branch road section, the vehicle speed probability density and the terminal time to obtain a time window; The time window is corrected according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

[0041] It should be noted that in order to accurately determine the travel time of the target suspicious vehicle at each path branch, it is necessary to first obtain the corresponding road section scene data, including the length of the branch section, the historical speed distribution of the section, real-time traffic speed data and real-time weather data; based on the historical speed distribution of the section, the speed probability density can be obtained through statistical analysis. This data can reflect the possibility of different speeds appearing on the section. For example, in a certain suburban road section, the probability of a speed of 60-80 kilometers per hour is 70%, providing a probabilistic basis for predicting the travel time; then, the branch section length, speed probability density and terminal time are combined for calculation and processing to obtain the time window. Specifically, by dividing the length of the road section by the vehicle speed under different probabilities, a variety of possible travel times are obtained, and then combined with the end time to reversely calculate, the time range for vehicles to pass through the road section is framed to form an initial time window. Since real-time road conditions and weather will significantly affect the travel efficiency, the initial time window needs to be corrected using real-time traffic speed data and real-time weather data; for example, when the real-time traffic speed is 30% lower than the historical average, the time window needs to be extended; in case of heavy rain, the time range is further expanded. For example, the initial time window is set to 30-40 minutes. If it rains suddenly, In the event of heavy rain, the window needs to be adjusted to 45-60 minutes based on the vehicle speed attenuation model of historical rainy days; if there is heavy fog with visibility less than 50 meters, an elastic interval needs to be added to the window, marked as "may be delayed by 15-30 minutes"; during the correction process, a two-factor weight model needs to be established: when real-time traffic flow and weather data affect each other in the same direction (such as congestion + heavy rain), the window is superimposed and extended; when the data act in the opposite direction (such as smooth traffic + light rain), the window range is fine-tuned based on the traffic speed, and finally a more realistic corrected time window is obtained, providing a time basis for precise control.

[0042] According to an embodiment of the present invention, the step of acquiring video data and processing the vehicle feature database information to obtain a vehicle matching degree includes: Obtain video data from monitoring points on each branch of the route and extract the feature database information of the vehicle to be verified; Corresponding feature matching is performed respectively according to the to-be-verified vehicle feature library information and the vehicle feature library information, and weighted processing is performed to obtain a vehicle matching degree.

[0043] It should be noted that the system collects video data from monitoring points on each path branch in sequence according to the priority of the path branches, and uses image recognition technology to extract feature information of the vehicle to be verified, including parameters such as license plate number, vehicle model, color, vehicle wheel and tire model; then, the feature library of the vehicle to be verified is matched dimension by dimension with the feature library of the target suspicious vehicle, where the license plate number is accurately matched by characters, the vehicle model is matched by contour and parameter library, the color is calibrated by spectral analysis, and the wheel and tire model is identified by combining pattern texture and size data; after the matching is completed, weights are assigned according to the importance of the features, for example, the license plate number accounts for 0.3, the vehicle model and color each account for 0.2, and the wheel and tire model account for 0.3 in total; the vehicle matching degree in the range of 0-100 is calculated through a weighted sum algorithm to achieve accurate quantitative comparison of multiple feature dimensions.

[0044] According to an embodiment of the present invention, obtaining a final complete path based on the corrected time window and vehicle matching degree includes: Performing a threshold comparison based on the vehicle matching degree and a preset vehicle matching threshold; If the vehicle matching degree is greater than the preset vehicle matching threshold, the vehicle matching degree is marked as the target matching degree, and the corresponding target monitoring point position and time are obtained; The final complete path is generated according to the corrected time window combined with the target monitoring point position and time.

[0045] It should be noted that the system will accurately compare the calculated matching degree of each vehicle with the preset vehicle matching threshold one by one; this threshold is set based on historical data and recognition accuracy requirements, and is used to filter out vehicle information that meets the basic matching standards; when the matching degree of a vehicle exceeds the preset threshold, the system will automatically mark it as the target matching degree, and quickly extract the target monitoring point location information corresponding to the matching result and the time data of the vehicle passing through the monitoring point; this information will be temporarily stored as key node data for path generation; then, the system calls the corrected time window parameters, combined with the target monitoring point location and time obtained previously, the system uses time series analysis and spatial association algorithms to connect each node in chronological order, and finally generates a complete and continuous driving path for the target suspicious vehicle.

[0046] The present invention discloses a suspicious vehicle path tracking method and system based on spatiotemporal reverse retrieval. The method and system obtain road network structure information related to preset toll stations and vehicle feature library information of target suspicious vehicles, extract path branches and match path priorities based on the road network structure information, obtain road section scene data corresponding to each path branch, process to obtain vehicle speed probability density and a corrected time window, obtain video data, combine the vehicle feature library information for processing to obtain vehicle matching degree, and obtain the final complete path based on the corrected time window and the vehicle matching degree, thereby realizing a technology for suspicious vehicle path tracking based on spatiotemporal reverse retrieval.

[0047] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: 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, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0048] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0049] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0050] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware related to program instructions, and the aforementioned program may be stored in a readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0051] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as standalone products, they can also be stored on a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This software product, stored on a storage medium, includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as removable storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A suspicious vehicle path tracking method based on spatiotemporal reverse retrieval is characterized by: The following steps are involved: Obtain road network structure information related to preset toll stations and vehicle feature database information of target suspicious vehicles; Extracting path branches and matching path priorities according to the road network structure information; Obtaining the road scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; Acquire video data and process it in combination with the vehicle feature database information to obtain a vehicle matching degree; The final complete path is obtained according to the corrected time window and the vehicle matching degree.

2. The suspicious vehicle path tracking method based on spatiotemporal reverse retrieval according to claim 1 is characterized in that: The obtaining of the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

3. The suspicious vehicle path tracking method based on spatiotemporal reverse retrieval according to claim 2 is characterized in that: The extracting path branches and matching path priorities according to the road network structure information includes: The path branches are obtained by processing the destination location, destination time, section length, confluence information, and connection relationship information through a preset path prediction model; Obtaining historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; Priority ranking is performed on each path branch according to the historical travel data to obtain a priority corresponding to each path branch.

4. The suspicious vehicle path tracking method based on spatiotemporal reverse retrieval according to claim 3 is characterized in that: The acquiring of the road scene data corresponding to each path branch and processing to obtain the vehicle speed probability density and the corrected time window includes: Obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, real-time traffic speed data, and real-time weather data; Obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; Processing is performed according to the length of the branch road section, the vehicle speed probability density and the terminal time to obtain a time window; The time window is corrected according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

5. The suspicious vehicle path tracking method based on spatiotemporal reverse retrieval according to claim 4 is characterized in that: The acquiring of video data and processing the vehicle feature database information to obtain a vehicle matching degree includes: Obtain video data from monitoring points on each branch of the route and extract the feature database information of the vehicle to be verified; Corresponding feature matching is performed respectively according to the to-be-verified vehicle feature library information and the vehicle feature library information, and weighted processing is performed to obtain a vehicle matching degree.

6. The suspicious vehicle path tracking method based on spatiotemporal reverse retrieval according to claim 5 is characterized in that: The step of obtaining a final complete path based on the corrected time window and vehicle matching includes: Performing a threshold comparison based on the vehicle matching degree and a preset vehicle matching threshold; If the vehicle matching degree is greater than the preset vehicle matching threshold, the vehicle matching degree is marked as the target matching degree, and the corresponding target monitoring point position and time are obtained; The final complete path is generated according to the corrected time window combined with the target monitoring point position and time.

7. The suspicious vehicle path tracking system based on time-space reverse retrieval is characterized by: The system includes: a memory and a processor, wherein the memory includes a program of a suspicious vehicle path tracking method based on time-space reverse retrieval, and when the program of the suspicious vehicle path tracking method based on time-space reverse retrieval is executed by the processor, the following steps are implemented: Obtain road network structure information related to preset toll stations and vehicle feature database information of target suspicious vehicles; Extracting path branches and matching path priorities according to the road network structure information; Obtaining the road scene data corresponding to each path branch, and processing to obtain the vehicle speed probability density and the corrected time window; Acquire video data and process it in combination with the vehicle feature database information to obtain a vehicle matching degree; The final complete path is obtained according to the corrected time window and the vehicle matching degree.

8. The suspicious vehicle path tracking system based on time-space reverse retrieval according to claim 7 is characterized in that: The obtaining of the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle includes: If a target suspicious vehicle is found at a preset toll station, obtain the road network structure information related to the preset toll station and the vehicle feature library information of the target suspicious vehicle; The road network structure information includes the destination location, destination time, section length, confluence information and connection relationship information; The vehicle feature database information includes license plate number, vehicle model, color, vehicle wheel and tire model.

9. The suspicious vehicle path tracking system based on time-space reverse retrieval according to claim 8 is characterized in that: The extracting path branches and matching path priorities according to the road network structure information includes: The path branches are obtained by processing the destination location, destination time, section length, confluence information, and connection relationship information through a preset path prediction model; Obtaining historical driving data of the target suspicious vehicle, including branch selection records and historical driving frequencies corresponding to each path branch; Priority ranking is performed on each path branch according to the historical travel data to obtain a priority corresponding to each path branch.

10. The suspicious vehicle path tracking system based on time-space reverse retrieval according to claim 9 is characterized in that: The acquiring of the road scene data corresponding to each path branch and processing to obtain the vehicle speed probability density and the corrected time window includes: Obtaining the road segment scene data corresponding to each path branch, including the branch segment length, the historical speed distribution of the segment, real-time traffic speed data, and real-time weather data; Obtaining a vehicle speed probability density based on the historical vehicle speed distribution of the road section; Processing is performed according to the length of the branch road section, the vehicle speed probability density and the terminal time to obtain a time window; The time window is corrected according to the real-time traffic speed data and the real-time weather data to obtain a corrected time window.

Images