Vehicle track missing complementing tracking method and system

Abstract

The invention discloses a vehicle track missing complementing tracking method and system. The method comprises the following steps: acquiring a target vehicle passing track and passing area road information; retrieving all possible passing paths of the target vehicle according to the missing road section and the passing area road information, and retrieving all reference vehicle tracks matched with the possible passing paths in the historical passing data of the missing road section; and calculating a possible passage path which most conforms to the missing road section track by utilizing a backtracking passage probability method and combining with the reference vehicle track, and complementing the target vehicle passage track by utilizing the possible passage path. According to the method, the road rule information and the road traffic condition of the GIS map are combined, and the probability model is combined with the existing historical vehicle traffic data to calculate the path with the maximum probability of traffic, so that path completion is performed, high calculation precision can be ensured under the condition of lack of monitoring, the complete track of the suspected vehicle is obtained, and the real-time performance of the suspected vehicle is improved. Therefore, travel analysis of the suspected vehicle can be carried out.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a vehicle track missing completion tracking method and system. Background Art

[0002] The existing vehicle tracking methods mainly use the video images captured by the road checkpoint camera to capture the passing car pictures and identify the license plate number in the pictures to track the location of the suspect car and restore the driving trajectory. Or extract the vehicle features to perform big data analysis and screening to find the corresponding specific vehicle trajectory.

[0003] However, in the process of restoring vehicle trajectories by the prior art, there are often problems of trajectory interruptions and omissions, that is, there is a period of time and road section in the middle where the vehicle's passing trajectory cannot be detected, but the interrupted area may be reached by several optional roads, and the existing checkpoint video processing system cannot cover all areas. There is also a certain rate of missed detection and false detection in places with camera checkpoints, which cannot accurately identify all passing vehicles, making it impossible to track the appearance and path of a specific vehicle, which brings omissions to the analysis and restoration of the vehicle's driving trajectory route.

[0004] Therefore, a method is now needed that can complete the vehicle's driving trajectory in the absence of monitoring data. Summary of the invention

[0005] To solve the above problems, the present invention provides a vehicle trajectory completion and tracking method and system, which uses the road conditions and road traffic information of missing sections in the GIS map combined with historical paths for matching and probability calculation, thereby solving the problem that the prior art cannot accurately track the trajectory of a specific vehicle when monitoring data is missing.

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

[0007] A vehicle trajectory missing completion tracking method comprises the following steps:

[0008] S1. Obtain the target vehicle's travel trajectory and road information in the travel area, extract the missing sections in the target vehicle's travel trajectory and the travel time T of the target vehicle passing through the missing sections g ;

[0009] S2. According to the missing sections and the road information of the pass area, retrieve all possible pass paths for the target vehicle. The starting point of the possible pass path is the starting point of the missing section of the target vehicle. Calculate the pass time of each possible pass path and filter out the pass time exceeding T. g possible travel paths;

[0010] S3. Obtaining the historical traffic data of the missing road section, the historical traffic data of the missing road section includes several historical vehicle trajectories. For each possible pass path, the historical vehicle trajectory with the highest degree of overlap with the possible pass path is retrieved from the historical traffic data of the missing road section according to the possible pass path, and the historical vehicle trajectory with the highest degree of overlap is the reference vehicle trajectory of the possible pass path;

[0011] S4. Use the backtracking probability method and the reference vehicle trajectory to calculate the possible path that best fits the missing section trajectory, and then add the possible path that best fits the missing section trajectory to the missing section of the target vehicle's trajectory to complete the target vehicle's trajectory. The completed target vehicle trajectory is a complete trajectory.

[0012] Furthermore, in step S3, the degree of overlap is calculated by overlapping and comparing the historical vehicle trajectory with the possible pass path, counting the roads that the two paths pass through together, and calculating the proportion of the roads that the two paths pass through together to the possible pass path. The proportion is the degree of overlap between the historical vehicle trajectory and the possible pass path.

[0013] Furthermore, the road information of the passing area includes GIS vector data of the passing area. In step S2, all possible passing paths of the target vehicle are retrieved, and the specific implementation method is as follows:

[0014] S21. Generate a traffic grid map based on the traffic area GIS vector data, and calculate the traffic area connectivity index and the traffic area road density;

[0015] S23. Using the traffic grid map, the traffic area connectivity index and the traffic area road density, retrieve all possible traffic paths between the starting point and the end point of the missing section of the target vehicle.

[0016] Furthermore, the calculation formula of the traffic area connectivity index is:

[0017]

[0018] Where N represents the number of road nodes in the traffic area, m i Indicates the number of edges adjacent to the i-th node, n 断 is the number of dead-end roads, n T is the number of T-shaped intersections, n + is the number of cross intersections, n 五 is the number of five-way intersections.

[0019] Furthermore, the calculation method of the road density in the pass area is: calculate the density value of the neighborhood of each road according to the kernel function, superimpose the density values ​​of all roads to obtain the pass area density surface, associate the density surface with the pass grid map, and extract the density value in each grid, which is the road density in the pass area.

[0020] Furthermore, the kernel function of the road neighborhood density value is:

[0021]

[0022] Among them, Xi is the location of the i-th road, h is the search radius, and n is the total number of roads in the neighborhood.

[0023] Furthermore, in step S23, the retrieval of all possible travel paths between the starting point and the end point of the missing section of the target vehicle is specifically implemented as follows: a directed graph is generated based on the traffic grid graph, each node in the directed graph is a road node in the traffic area, and the starting point of the missing section is used as the starting point. All nodes in the directed graph are traversed according to the depth-first method, and all routes passing through the end point of the missing section are counted, and duplicate routes are eliminated. The remaining routes are all possible travel paths between the starting point and the end point of the missing section of the target vehicle.

[0024] Furthermore, in step S2, the travel time of each possible travel path is calculated, and its specific implementation method is: the travel time of each possible travel path is calculated according to the travel speed, congestion, road saturation, and traffic light travel time of each section within the target vehicle travel time period.

[0025] Further, in step S4, the implementation method of using the backtracking travel probability method in combination with the reference vehicle trajectory to calculate the possible travel path that best matches the missing section trajectory is:

[0026] The specific formula of the backtracking probability is:

[0027]

[0028] Among them, P(βi) is the probability of a possible passable path among all possible passable paths, βi is the i-th possible passable path, P(α|βi) is the proportion of the possible passable path βi in the corresponding reference vehicle trajectory, and P(βi|α) represents the probability that the possible passable path βi is a missing section path. The possible passable path with the highest probability is the possible passable path that best matches the missing section trajectory.

[0029] Through the above technical scheme, the present invention has the following beneficial effects: the present invention combines the road rule information and road traffic conditions of the GIS map, and uses the probability model combined with the existing historical vehicle traffic data to calculate the path with the highest probability of passage, thereby completing the path. It can ensure high calculation accuracy even in the absence of monitoring, and obtain the complete trajectory of the suspected vehicle, so as to conduct travel analysis of the suspected vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The present invention is a schematic diagram of the overall process of a vehicle trajectory missing completion tracking method.

[0031] Figure 2 Schematic diagram of the structure of a vehicle trajectory missing completion tracking system in an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] Example 1

[0035] See also Figure 1 A vehicle trajectory missing completion tracking method comprises the following steps:

[0036] S1. Obtain the target vehicle's travel trajectory and road information in the travel area, extract the missing sections in the target vehicle's travel trajectory and the travel time T of the target vehicle passing through the missing sections g ;

[0037] S2. According to the missing sections and the road information of the pass area, retrieve all possible pass paths for the target vehicle. The starting point of the possible pass path is the starting point of the missing section of the target vehicle. Calculate the pass time of each possible pass path and filter out the pass time exceeding T. g possible travel paths;

[0038] S3. Obtaining the historical traffic data of the missing road section, the historical traffic data of the missing road section includes several historical vehicle trajectories. For each possible pass path, the historical vehicle trajectory with the highest degree of overlap with the possible pass path is retrieved from the historical traffic data of the missing road section according to the possible pass path, and the historical vehicle trajectory with the highest degree of overlap is the reference vehicle trajectory of the possible pass path;

[0039] S4. Use the backtracking probability method and the reference vehicle trajectory to calculate the possible path that best fits the missing section trajectory, and then add the possible path that best fits the missing section trajectory to the missing section of the target vehicle's trajectory to complete the target vehicle's trajectory. The completed target vehicle trajectory is a complete trajectory.

[0040] In an optional embodiment, in step S3, the degree of overlap is calculated by overlapping and comparing the historical vehicle trajectory with the possible pass path, counting the roads that the two paths pass through together, and calculating the proportion of the roads that the two paths pass through together to the possible pass path, and the proportion is the degree of overlap between the historical vehicle trajectory and the possible pass path.

[0041] It also includes obtaining target vehicle information, the target vehicle information includes vehicle model, license plate number and other information, and the traffic area GIS vector data includes road traffic rules, traffic restriction rules, road grade, road attributes, number of lanes, road width, speed limit and height limit, types of vehicles allowed to pass and other information.

[0042] By detecting the traffic data, the point information of the path interruption section is obtained, so as to retrieve the road from the source of the trajectory interruption point to the range of the detected trajectory recovery point.

[0043] In an optional embodiment, the pass area road information includes the pass area GIS vector data. In step S2, all possible pass paths of the target vehicle are retrieved, and the specific implementation method is as follows:

[0044] S21. Generate a traffic grid map based on the traffic area GIS vector data, and calculate the traffic area connectivity index and the traffic area road density;

[0045] S23. Using the traffic grid map, the traffic area connectivity index and the traffic area road density, retrieve all possible traffic paths between the starting point and the end point of the missing section of the target vehicle.

[0046] In an optional embodiment, the calculation formula of the traffic area connectivity index is:

[0047]

[0048] Where N represents the number of road nodes in the traffic area, m iIndicates the number of edges adjacent to the i-th node, n 断 is the number of dead-end roads, n T is the number of T-shaped intersections, n + is the number of cross intersections, n 五 is the number of five-way intersections.

[0049] When calculating, it is necessary to break the original intersection lines in the GIS map and refine the road data to the node level again, then create a network dataset, count the number of edges of the roads connected to each node, and then use the attribute table to find the number of intersections to calculate the connectivity index of the traffic area. The connectivity index is one of the indicators for evaluating the rationality of the road network structure. It is an important indicator that reflects the networking and reliability of the road network. It is usually used to measure the maturity of the road network. By calculating the connectivity index, it is convenient for subsequent road prediction.

[0050] In an optional embodiment, the road density in the pass area is calculated as follows: the density value of the neighborhood of each road is calculated according to the kernel function, the density values ​​of all roads are superimposed to obtain the density surface of the pass area, the density surface is associated with the pass grid map, and the density value in each grid is extracted, which is the road density in the pass area.

[0051] Road density refers to the ratio of the total length of all roads in a region to the total area of ​​the region. The greater the road density, the denser the road network, and the target vehicle's trajectory can be predicted and the road travel time can be calculated based on the road density. The denser the road, the longer it takes for the target vehicle to travel and the more possible routes it can choose.

[0052] In an optional embodiment, the kernel function of the road neighborhood density value is:

[0053]

[0054] Among them, Xi is the location of the i-th road, h is the search radius, and n is the total number of roads in the neighborhood.

[0055] In an optional embodiment, in step S23, the retrieval of all possible travel paths between the starting point and the end point of the missing section of the target vehicle is specifically implemented as follows: a directed graph is generated based on the traffic grid graph, each node in the directed graph is a road node in the traffic area, and the starting point of the missing section is used as the starting point. All nodes in the directed graph are traversed according to the depth-first method, and all routes passing through the end point of the missing section are counted, and duplicate routes are eliminated. The remaining routes are all possible travel paths between the starting point and the end point of the missing section of the target vehicle.

[0056] In the specific calculation, a directed graph is created based on the starting point, end point, roads and obstacles, and the next layer is traversed from the starting point according to the depth-first principle. When selecting the next node, the probability of the target vehicle selecting a node in the next layer is calculated based on the road neighborhood density value and road connectivity of the node neighborhood, thereby calculating the optimal route that the target vehicle may choose.

[0057] In an optional embodiment, in step S2, the travel time of each possible travel path is calculated by calculating the travel time of each possible travel path based on the travel speed, congestion, road saturation, and traffic light travel time of each road section within the target vehicle travel time period.

[0058] In an optional embodiment, in step S4, the method of using the backtracking travel probability method in combination with the reference vehicle trajectory to calculate the possible travel path that best matches the missing road section trajectory is as follows:

[0059] The specific formula of the backtracking probability is:

[0060]

[0061] Among them, P(βi) is the probability of a possible passable path among all possible passable paths, βi is the i-th possible passable path, P(α|βi) is the proportion of the possible passable path βi in the corresponding reference vehicle trajectory, and P(βi|α) represents the probability that the possible passable path βi is a missing section path. The possible passable path with the highest probability is the possible passable path that best matches the missing section trajectory.

[0062] P(α|βi) is the proportion of the possible pass path βi in the corresponding reference vehicle trajectory, specifically referring to the matching degree between the corresponding retrieved trajectory and the path βi when the path βi is used to retrieve the reference vehicle trajectory.

[0063] After calculating the possible path with the highest probability, it is matched with the target vehicle trajectory to confirm that the possible path can match the missing starting point of the target vehicle trajectory. If it cannot be matched, the next possible path with the highest probability is selected in sequence until a path that fully matches the target vehicle trajectory is found, and then the target vehicle trajectory is completed to ensure the accuracy of trajectory prediction.

[0064] Example 2

[0065] See also Figure 2 , a vehicle trajectory missing completion tracking system, comprising:

[0066] The information acquisition module is used to obtain the target vehicle's passing trajectory and the road information of the passing area, extract the missing sections in the target vehicle's passing trajectory and the passing time T of the target vehicle passing the missing sections g ;

[0067] The path calculation module is used to retrieve all possible paths for the target vehicle based on the missing sections and the road information of the pass area. The starting point of the possible path is the starting point of the missing section of the target vehicle. The travel time of each possible path is calculated, and the paths with a travel time exceeding T are screened out. g possible travel paths;

[0068] A historical matching module is used to obtain the historical traffic data of the missing road section, wherein the historical traffic data of the missing road section includes a plurality of historical vehicle trajectories. For each possible passable path, the historical vehicle trajectory with the highest degree of overlap with the possible passable path is retrieved from the historical traffic data of the missing road section according to the possible passable path, and the historical vehicle trajectory with the highest degree of overlap is the reference vehicle trajectory of the possible passable path;

[0069] The trajectory completion module is used to use the backtracking probability method and the reference vehicle trajectory to calculate the possible pass path that best matches the missing section trajectory, and then fill the missing section of the target vehicle's passage trajectory with the possible pass path that best matches the missing section trajectory to complete the target vehicle's passage trajectory. The completed target vehicle passage trajectory is a complete trajectory.

[0070] The embodiment disclosed in this specification is only an example of a unilateral feature of the present invention, and the protection scope of the present invention is not limited to this embodiment, and any other functionally equivalent embodiments fall within the protection scope of the present invention. For those skilled in the art, various other corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all of these changes and deformations should fall within the protection scope of the claims of the present invention.

Claims

1. A vehicle trajectory missing completion tracking method, characterized in that: The following steps are involved: S1. Obtain the target vehicle's travel trajectory and road information in the travel area, extract the missing sections in the target vehicle's travel trajectory and the travel time T of the target vehicle passing through the missing sections g ; S2. According to the missing sections and the road information of the pass area, retrieve all possible pass paths for the target vehicle. The starting point of the possible pass path is the starting point of the missing section of the target vehicle. Calculate the pass time of each possible pass path and filter out the pass time exceeding T. g possible travel paths; S3. Obtaining the historical traffic data of the missing road section, the historical traffic data of the missing road section includes several historical vehicle trajectories. For each possible pass path, the historical vehicle trajectory with the highest degree of overlap with the possible pass path is retrieved from the historical traffic data of the missing road section according to the possible pass path, and the historical vehicle trajectory with the highest degree of overlap is the reference vehicle trajectory of the possible pass path; S4. Use the backtracking probability method and the reference vehicle trajectory to calculate the possible path that best fits the missing section trajectory, and then add the possible path that best fits the missing section trajectory to the missing section of the target vehicle's trajectory to complete the target vehicle's trajectory. The completed target vehicle trajectory is a complete trajectory.

2. A vehicle trajectory missing completion tracking method according to claim 1, characterized in that: In step S3, the degree of overlap is calculated by overlapping and comparing the historical vehicle trajectory with the possible pass path, counting the roads that the two paths pass through together, and calculating the proportion of the roads that the two paths pass through together to the possible pass path. The proportion is the degree of overlap between the historical vehicle trajectory and the possible pass path.

3. A vehicle trajectory missing completion tracking method according to claim 1, characterized in that: The road information of the passing area includes GIS vector data of the passing area. In step S2, all possible passing paths of the target vehicle are retrieved, and the specific implementation method is as follows: S21. Generate a traffic grid map based on the traffic area GIS vector data, and calculate the traffic area connectivity index and the traffic area road density; S23. Using the traffic grid map, the traffic area connectivity index and the traffic area road density, retrieve all possible traffic paths between the starting point and the end point of the missing section of the target vehicle.

4. A vehicle trajectory missing completion tracking method according to claim 3, characterized in that: The calculation formula of the traffic area connectivity index is: Where N represents the number of road nodes in the traffic area, m i Indicates the number of edges adjacent to the i-th node, n 断 is the number of dead-end roads, n T is the number of T-shaped intersections, n + is the number of cross intersections, n 五 is the number of five-way intersections.

5. A vehicle track missing completion tracking method according to claim 4, characterized in that: The calculation method of the road density in the pass area is: calculate the density value of the neighborhood of each road according to the kernel function, superimpose the density values ​​of all roads to obtain the pass area density surface, associate the density surface with the pass grid map, and extract the density value in each grid, which is the road density in the pass area.

6. A vehicle track missing completion tracking method according to claim 5, characterized in that: The kernel function of the road neighborhood density value is: Among them, Xi is the location of the i-th road, h is the search radius, and n is the total number of roads in the neighborhood.

7. A vehicle track missing completion tracking method according to claim 6, characterized in that: In step S23, the retrieval of all possible travel paths between the starting point and the end point of the missing section of the target vehicle is specifically implemented as follows: a directed graph is generated based on the traffic grid graph, each node in the directed graph is a road node in the traffic area, and the starting point of the missing section is used as the starting point. All nodes in the directed graph are traversed according to the depth-first method, and all routes passing through the end point of the missing section are counted, and duplicate routes are eliminated. The remaining routes are all possible travel paths between the starting point and the end point of the missing section of the target vehicle.

8. A vehicle track missing completion tracking method according to claim 7, characterized in that: In step S2, the travel time of each possible travel path is calculated, and its specific implementation method is: the travel time of each possible travel path is calculated according to the travel speed, congestion, road saturation, and traffic light travel time of each road section within the target vehicle travel time period.

9. A vehicle track missing completion tracking method according to claim 1, characterized in that: In step S4, the method of using the backtracking travel probability method in combination with the reference vehicle trajectory to calculate the possible travel path that best matches the missing section trajectory is as follows: The specific formula of the backtracking probability is: Among them, P(βi) is the probability of a possible passable path among all possible passable paths, βi is the i-th possible passable path, P(α|βi) is the proportion of the possible passable path βi in the corresponding reference vehicle trajectory, and P(βi|α) represents the probability that the possible passable path βi is a missing section path. The possible passable path with the highest probability is the possible passable path that best matches the missing section trajectory.

10. A vehicle track missing completion tracking system, characterized in that: include: The information acquisition module is used to obtain the target vehicle's passing trajectory and the road information of the passing area, extract the missing sections in the target vehicle's passing trajectory and the passing time T of the target vehicle passing the missing sections g ; The path calculation module is used to retrieve all possible paths for the target vehicle based on the missing sections and the road information of the pass area. The starting point of the possible path is the starting point of the missing section of the target vehicle. The travel time of each possible path is calculated, and the paths with a travel time exceeding T are screened out. g possible travel paths; A history matching module is used to obtain the missing section historical traffic data, the missing section historical traffic data includes a number of historical vehicle trajectories, for each possible pass path, the missing section historical traffic data is retrieved according to the possible pass path, the historical vehicle trajectory with the highest degree of overlap with the possible pass path, the historical vehicle trajectory with the highest degree of overlap is the reference vehicle trajectory of the possible pass path; The trajectory completion module is used to use the backtracking probability method and the reference vehicle trajectory to calculate the possible pass path that best matches the missing section trajectory, and then fill the missing section of the target vehicle's passage trajectory with the possible pass path that best matches the missing section trajectory to complete the target vehicle's passage trajectory. The completed target vehicle passage trajectory is a complete trajectory.

Images