Construction method and system of cooperative control model, electronic equipment and storage medium
By building a collaborative control model, combining vehicle detection, behavior recognition and density estimate models, the problem of inefficient traffic management in highway environments is solved, deep understanding of vehicle behavior and intelligent decision-making, and system reliability and stability.
Patent Information
- Application Number
- CN202410960083.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-17
- Publication Date
- 2024-11-12
AI Technical Summary
Traditional traffic management systems are inefficient in highway environments, and manual monitoring is difficult to respond in a timely manner, resulting in frequent traffic congestion and accidents. The application of a single technology is not enough to achieve comprehensive intelligence and automation of intelligent traffic systems.
Build a collaborative control model, determine vehicle identification and driving trajectory by obtaining continuous image data, build vehicle detection and tracking models, map it to point cloud coordinate system to determine vehicle behavior, combine vehicle density estimate model, and integrate multiple models to achieve vehicle behavior in-depth understanding and prediction.
It realizes intelligent and automated decision-making support for vehicle behavior, reduces manual intervention, and improves the reliability and stability of the system.
Smart Images

Figure CN118942051A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of highways, and in particular to a method, system, electronic device and storage medium for constructing a collaborative control model. Background Art
[0002] Traditional traffic management systems rely on manual monitoring and manual operation, which can be used when traffic volume is low and road conditions are simple, but its efficiency is significantly reduced when facing highway environments. Manual monitoring not only has problems of fatigue and misjudgment, but also makes it difficult to respond accurately and promptly in the first place, which leads to increased traffic congestion at toll stations and frequent traffic accidents. However, to achieve comprehensive intelligence and automation of intelligent transportation systems, it is far from enough to rely solely on a single technology application.
[0003] Therefore, in order to achieve a comprehensive understanding and prediction of vehicle behavior, highway toll stations need to coordinate and integrate different technologies to leverage the advantages of various technologies, conduct in-depth analysis and prediction of vehicle behavior from multiple angles, and thus provide intelligent support for driving guidance decisions. Summary of the invention
[0004] The purpose of the embodiments of the present invention is to provide a method, system, electronic device and storage medium for constructing a collaborative control model, which are used to fully or at least partially solve the technical problems existing in the above-mentioned prior art.
[0005] In order to achieve the above object, an embodiment of the present invention provides a method for constructing a collaborative control model, which is applied to a toll island at a toll station on a highway, comprising:
[0006] Acquire continuous image data of vehicles entering the toll island over a period of time, and determine the vehicle identification and driving trajectory based on the continuous image data;
[0007] Based on the driving trajectory and vehicle identification, vehicles entering the toll island, toll plaza and ramp are tracked to build a vehicle detection and tracking model;
[0008] Mapping continuous image data to a point cloud coordinate system, constructing point cloud data corresponding to the vehicle in each image, and determining the vehicle's behavior based on the point cloud data to build a behavior recognition model;
[0009] Construct a vehicle density estimation model based on the number of vehicles in the image and the area corresponding to the image;
[0010] The vehicle detection and tracking model, the behavior recognition model and the vehicle density estimation model are integrated to obtain a collaborative control model.
[0011] Optionally, obtaining continuous image data of vehicles entering the toll island within a period of time, and determining the vehicle identification and driving trajectory according to the continuous image data, includes:
[0012] Acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification according to the continuous image data;
[0013] The running track reference points are set based on the continuous image data, and the running track reference points set in each image are arranged and connected according to the time series to obtain the driving track of the vehicle.
[0014] Optionally, construct a vehicle detection and tracking model according to the following formula:
[0015]
[0016] In the formula, IoU(A,B) represents the similarity metric, A Overlap represents the vehicle bounding box in the previous frame, A Union represents the vehicle bounding box in the current frame; d represents the distance between the vehicle and the edge of the toll island, which is used to determine whether the vehicle is approaching or entering the toll island. 1 ,y 1 ) and (x 2 ,y 2 ) are the coordinates of two points on the toll island boundary line, and (x, y) are the coordinates of the vehicle.
[0017] Optionally, the continuous image data is mapped to a point cloud coordinate system, point cloud data corresponding to the vehicle in each image is constructed, and the behavior of the vehicle is determined based on the point cloud data to construct a behavior recognition model, including:
[0018] The continuous image data is mapped to the point cloud coordinate system, and based on the point cloud data corresponding to the set running trajectory reference point, the point cloud data corresponding to the vehicle in each image is constructed, and the vehicle's behavior is determined according to the point cloud data corresponding to the vehicle in each image to construct a behavior recognition model.
[0019] Optionally, construct a behavior recognition model according to the following formula:
[0020]
[0021] Where N represents the number of point cloud data corresponding to the input vehicle, y i represents the behavior category of the point cloud data i corresponding to the vehicle, p i Represents the probability that the predicted point cloud data i belongs to the positive class.
[0022] Optionally, a vehicle density estimation model is constructed based on the number of vehicles in the image and the area corresponding to the image, including:
[0023] Acquire image data within a distance range of the toll island, toll plaza and ramp, and perform edge cropping on the image to obtain a target image;
[0024] Determine the area corresponding to the target image, and obtain the number of vehicles in the target image based on thermodynamic principles;
[0025] The vehicle density is estimated based on the number of vehicles in the target image and the area corresponding to the target image.
[0026] Optionally, a vehicle density estimation model is constructed according to the following formula:
[0027]
[0028] Where D represents the vehicle density, N represents the number of vehicles detected in the image, A_img represents the pixel area of the image, and R represents the resolution of the image. Indicates the actual area of the image.
[0029] On the other hand, the present invention also provides a collaborative control model construction system, which is applied to a toll station toll island on a highway, comprising:
[0030] An acquisition unit, used to acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification and driving trajectory based on the continuous image data;
[0031] A vehicle detection and tracking model building unit, used to track vehicles entering the toll island, toll plaza and ramp based on the driving trajectory and vehicle identification, so as to build a vehicle detection and tracking model;
[0032] A behavior recognition model building unit, used for mapping continuous image data to a point cloud coordinate system, building point cloud data corresponding to the vehicle in each image, and determining the behavior of the vehicle based on the point cloud data to build a behavior recognition model;
[0033] A vehicle density estimation model building unit, used to build a vehicle density estimation model according to the number of vehicles in the image and the area corresponding to the image;
[0034] A fusion unit is used to fuse the vehicle detection and tracking model, the behavior recognition model and the vehicle density estimation model to obtain a collaborative control model.
[0035] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for constructing a collaborative control model as described above when executing the program.
[0036] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for constructing a collaborative control model described above are implemented.
[0037] Through the above technical solution, a collaborative control model is built based on vehicle detection and tracking algorithms, behavior recognition algorithms, and density estimation algorithms, which can achieve a deep understanding and prediction of vehicle behavior. This makes driving guidance decisions more intelligent and automated, reduces the need for human intervention, and improves the reliability and stability of the system.
[0038] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:
[0040] Figure 1 is an implementation flow chart of a method for constructing a collaborative control model provided by an embodiment of the present invention;
[0041] Figure 2 It is a structural diagram of a system for building a collaborative control model provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0042] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.
[0043] See also Figure 1 As shown, it is an implementation flow chart of a method for constructing a collaborative control model provided by an embodiment of the present invention, including the following execution steps:
[0044] Step 100: Acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification and driving trajectory based on the continuous image data.
[0045] It should be noted that the continuous image data includes environmental information of the toll island, toll plaza and ramp, traffic information and equipment operation status information, including environmental feature data, traffic feature data and equipment operation status feature data.
[0046] Specifically, when executing step 100, the following steps may be specifically performed:
[0047] S1000: Acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification according to the continuous image data.
[0048] In some implementations, the location of the license plate in the image may be determined using connected domain analysis, which may be performed in the following steps:
[0049] First, the collected image is preprocessed, which usually includes grayscale, filtering and denoising steps to eliminate irrelevant information and noise in the image and highlight the characteristics of the license plate. The edge detection algorithm (such as Sobel operator, Canny edge detection, etc.) is used to detect the edge of the preprocessed image to highlight the edge characteristics of the license plate area. Since the character color and background color of the license plate area change dramatically in grayscale, edge detection can effectively highlight the license plate area. The image after edge detection is binarized to separate the license plate area from the background area. In the binarization process, a suitable threshold can be set so that the license plate area and the background area are clearly distinguished in grayscale value. Connected domain analysis is a key step in determining the location of the license plate. The bwlabel function in MATLAB is used to mark the continuous pixel blocks (i.e., connected domains) in the binary image. Since the license plate is a relatively large connected domain, the license plate area can be screened out by setting the size, shape and other features of the connected domain.
[0050] Specifically, the connected domain analysis can be divided into the following steps:
[0051] Select a pixel marked as 1 as the starting point, and add the connected pixels to the object by analyzing its four-connectivity or eight-connectivity (i.e., the adjacent pixels in the upper, lower, left, right, or diagonal directions of the pixel). Iterate the loop until no new connected pixels are added to the current object, that is, the extraction of the connected domain currently marked as 1 is completed, and repeat the above steps until all pixels marked as 1 are assigned to the corresponding connected domain. In the process of connected domain analysis, the connected domain that meets the characteristics of the license plate can be screened out by setting the size, aspect ratio and other parameters of the connected domain. At the same time, since the color of the characters and the background color of the license plate area change dramatically in grayscale, the license plate area is usually shown as a larger, color-consistent connected domain in the connected domain analysis. After screening out the connected domain that meets the characteristics of the license plate, the specific position of the license plate in the image is determined by calculating the center position, area and other parameters of the connected domain. At the same time, the license plate can be rotated, scaled and other transformations as needed to correct the tilt and distortion of the license plate.
[0052] S1001: Setting running track reference points based on the continuous image data, arranging and connecting the running track reference points set in each image in a time series, and obtaining the driving track of the vehicle.
[0053] In some embodiments, each image is preprocessed, including but not limited to grayscale, noise reduction, contrast enhancement and other steps to highlight the difference between the vehicle and the background, and the target detection algorithm SSD is used to detect the vehicle target in each image to output the vehicle's bounding box, that is, the position information of the vehicle in the image. In the detected vehicle bounding box, one or more fixed points are selected as reference points of the running track. It should be understood that these points can be the center of mass of the vehicle, the center point of the front and rear license plates, etc. In this example, the center of mass of the vehicle is selected as the reference point. For the vehicle target detected in each image, the position information of its reference point is extracted, including the horizontal coordinate (x) and the vertical coordinate (y) of the reference point in the image coordinate system, and a timestamp is added to the reference point position information in each image to mark the corresponding time point. The timestamp can be the time of image acquisition, or it can be a relative time calculated according to the frequency and order of image acquisition. According to the time sequence, the running track reference points set in each image are arranged, that is, the reference point position information at different time points is organized in time sequence. Then, a linear interpolation algorithm is used to generate a smooth running track according to the time sequence of the reference point position information. Finally, the generated driving trajectory is converted into the map coordinate system and drawn on the map or in the visualization interface using a graphics drawing library.
[0054] Step 101: Based on the driving trajectory and vehicle identification, vehicles entering the toll island, toll plaza and ramp are tracked to build a vehicle detection and tracking model.
[0055] Specifically, the vehicle detection and tracking model can be constructed according to the following formula:
[0056]
[0057] In the formula, IoU(A,B) represents the similarity metric, A Overlap represents the vehicle bounding box in the previous frame, A Union represents the vehicle bounding box in the current frame; d represents the distance between the vehicle and the edge of the toll island, which is used to determine whether the vehicle is approaching or entering the toll island. 1 ,y 1 ) and (x 2 ,y 2 ) are the coordinates of two points on the toll island boundary line, and (x, y) are the coordinates of the vehicle.
[0058] Step 102: Map the continuous image data to a point cloud coordinate system, construct point cloud data corresponding to the vehicle in each image, and determine the behavior of the vehicle based on the point cloud data to construct a behavior recognition model.
[0059] Specifically, when executing step 102, the following steps may be specifically performed: mapping continuous image data to a point cloud coordinate system, and constructing point cloud data corresponding to the vehicle in each image based on the point cloud data corresponding to the set running trajectory reference point, and determining the behavior of the vehicle based on the point cloud data corresponding to the vehicle in each image to construct a behavior recognition model.
[0060] Furthermore, the image and point cloud data are registered to ensure the time synchronization and spatial calibration between the camera and LiDAR sensor. The two are registered by the vehicle bounding box in the image and the point cloud data of LiDAR. This usually involves steps such as projection transformation, image segmentation and point cloud segmentation to ensure the consistency of the vehicle area in the image and point cloud data. One or more reference points (such as the center of mass of the vehicle, the center point of the front and rear license plates, etc.) are set within the detected vehicle bounding box. Based on the vehicle bounding box in the image and the set reference points, the corresponding point cloud data is extracted from the registered point cloud data to represent the position and shape of the vehicle in 3D space. Features are extracted from the point cloud data corresponding to each image. These features may include the shape, size, speed, acceleration, direction, etc. of the vehicle. A series of vehicle behaviors are defined, such as acceleration, deceleration, turning, parking, etc. These behaviors can be defined and classified according to actual needs. The vehicle behaviors corresponding to each image are manually or automatically annotated as a training set. The goal of the model is to predict the corresponding behavior based on the input vehicle point cloud data. The trained behavior recognition model is applied to new continuous image data to identify the behavior of each vehicle at each time point.
[0061] In some implementations, a behavior recognition model may be constructed according to the following formula:
[0062] Assume that there is a point P whose coordinates in the image coordinate system are (u, v) and whose coordinates in the point cloud coordinate system are (X w ,Y w ,Z w ), transform the coordinates to the same coordinate system:
[0063]
[0064] Where T represents the transformation matrix.
[0065] Assume that the pixel coordinates of the vehicle bounding box in the image are (x min ,y min ,x max ,y max ), the corresponding point set in the point cloud data is P, and the point cloud belonging to the vehicle can be determined based on the coordinates of each point in the bounding box and the point cloud data:
[0066] p vehicle ={p∈P|x min≤p x ≤x max ,y min ≤p y ≤y max};
[0067] In the formula, p x , p y is the coordinate of point p in the point cloud coordinate system.
[0068]
[0069] Where N represents the number of point cloud data corresponding to the input vehicle, y i represents the behavior category of the point cloud data i corresponding to the vehicle, p i Represents the probability that the predicted point cloud data i belongs to the positive class.
[0070] Step 103: Construct a vehicle density estimation model based on the number of vehicles in the image and the area corresponding to the image.
[0071] Specifically, when executing step 103, the following steps may be specifically performed:
[0072] S1030: Acquire image data within a distance range of the toll island, toll plaza, and ramp, and perform edge cropping on the image to obtain a target image.
[0073] S1031: Determine the area corresponding to the target image, and obtain the number of vehicles in the target image based on thermodynamic principles.
[0074] S1032: Estimate vehicle density based on the number of vehicles in the target image and the area corresponding to the target image.
[0075] In some implementations, a vehicle density estimation model may be constructed according to the following formula:
[0076]
[0077] Where D represents the vehicle density, N represents the number of vehicles detected in the image, A_img represents the pixel area of the image, and R represents the resolution of the image. Indicates the actual area of the image.
[0078] Step 104: The vehicle detection and tracking model, the behavior recognition model, and the vehicle density estimation model are integrated to obtain a collaborative control model.
[0079] By building a collaborative control model based on vehicle detection and tracking algorithms, behavior recognition algorithms, and density estimation algorithms, we can achieve a deep understanding and prediction of vehicle behavior, which makes driving guidance decisions more intelligent and automated, reduces the need for human intervention, and improves the reliability and stability of the system.
[0080] See also Figure 2 FIG. 1 is a schematic diagram of a structure of a collaborative control model construction system provided by an embodiment of the present invention, which is applied to a toll station toll island on a highway, and includes:
[0081] The acquisition unit 200 is used to acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification and driving trajectory according to the continuous image data;
[0082] A vehicle detection and tracking model building unit 201 is used to track vehicles entering the toll island, toll plaza and ramp based on the driving trajectory and vehicle identification to build a vehicle detection and tracking model;
[0083] A behavior recognition model building unit 202 is used to map the continuous image data to a point cloud coordinate system, build point cloud data corresponding to the vehicle in each image, and determine the behavior of the vehicle based on the point cloud data to build a behavior recognition model;
[0084] A vehicle density estimation model building unit 203 is used to build a vehicle density estimation model according to the number of vehicles in the image and the area corresponding to the image;
[0085] The fusion unit 204 is used to fuse the vehicle detection and tracking model, the behavior recognition model and the vehicle density estimation model to obtain a collaborative control model.
[0086] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for constructing a collaborative control model described in any one of the above embodiments are implemented.
[0087] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for constructing a collaborative control model described in any of the above embodiments are implemented.
[0088] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0089] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0090] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0091] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0092] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0093] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0094] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0095] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0096] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A method for constructing a collaborative control model, applied to a toll island at a toll station on a highway, characterized in that: include: Acquire continuous image data of vehicles entering the toll island over a period of time, and determine the vehicle identification and driving trajectory based on the continuous image data; Based on the driving trajectory and vehicle identification, vehicles entering the toll island, toll plaza and ramp are tracked to build a vehicle detection and tracking model; Mapping continuous image data to a point cloud coordinate system, constructing point cloud data corresponding to the vehicle in each image, and determining the vehicle's behavior based on the point cloud data to build a behavior recognition model; Construct a vehicle density estimation model based on the number of vehicles in the image and the area corresponding to the image; The vehicle detection and tracking model, the behavior recognition model and the vehicle density estimation model are integrated to obtain a collaborative control model.
2. The method for constructing a collaborative control model according to claim 1, characterized in that: Acquiring continuous image data of vehicles entering the toll island within a period of time, and determining the vehicle identification and driving trajectory according to the continuous image data, including: Acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification according to the continuous image data; The running track reference points are set based on the continuous image data, and the running track reference points set in each image are arranged and connected according to the time series to obtain the driving track of the vehicle.
3. The method for constructing a collaborative control model according to claim 1, characterized in that: The vehicle detection and tracking model is constructed according to the following formula: In the formula, IoU(A, B) represents the similarity metric, A Overlap represents the vehicle bounding box in the previous frame, A Union represents the vehicle bounding box in the current frame; d represents the distance between the vehicle and the edge of the toll island, (x1, y1) and (x2, y2) are the coordinates of two points on the boundary line of the toll island, and (x, y) are the coordinates of the vehicle.
4. The method for constructing a collaborative control model according to claim 1, characterized in that: Map the continuous image data to the point cloud coordinate system, construct the point cloud data corresponding to the vehicle in each image, and determine the vehicle's behavior based on the point cloud data to build a behavior recognition model, including: The continuous image data is mapped to the point cloud coordinate system, and based on the point cloud data corresponding to the set running trajectory reference point, the point cloud data corresponding to the vehicle in each image is constructed, and the vehicle's behavior is determined according to the point cloud data corresponding to the vehicle in each image to construct a behavior recognition model.
5. The method for constructing a collaborative control model according to claim 1, characterized in that: The behavior recognition model is constructed according to the following formula: Where N represents the number of point cloud data corresponding to the input vehicle, y i represents the behavior category of the point cloud data i corresponding to the vehicle, p i Represents the probability that the predicted point cloud data i belongs to the positive class.
6. The method for constructing a collaborative control model according to claim 1, characterized in that: A vehicle density estimation model is constructed based on the number of vehicles in the image and the area corresponding to the image, including: Acquire image data within a distance range of the toll island, toll plaza and ramp, and perform edge cropping on the image to obtain a target image; Determine the area corresponding to the target image, and obtain the number of vehicles in the target image based on thermodynamic principles; The vehicle density is estimated based on the number of vehicles in the target image and the area corresponding to the target image.
7. The method for constructing a collaborative control model according to claim 1, characterized in that: The vehicle density estimation model is constructed according to the following formula: Where D represents the vehicle density, N represents the number of vehicles detected in the image, A_img represents the pixel area of the image, and R represents the resolution of the image. Indicates the actual area of the image.
8. A collaborative control model construction system, applied to toll islands at toll stations on highways, characterized in that: include: An acquisition unit, used to acquire continuous image data of vehicles entering the toll island within a period of time, and determine the vehicle identification and driving trajectory based on the continuous image data; A vehicle detection and tracking model building unit, used to track vehicles entering the toll island, toll plaza and ramp based on the driving trajectory and vehicle identification, so as to build a vehicle detection and tracking model; A behavior recognition model building unit, used for mapping continuous image data to a point cloud coordinate system, building point cloud data corresponding to the vehicle in each image, and determining the behavior of the vehicle based on the point cloud data to build a behavior recognition model; A vehicle density estimation model building unit, used to build a vehicle density estimation model according to the number of vehicles in the image and the area corresponding to the image; A fusion unit is used to fuse the vehicle detection and tracking model, the behavior recognition model and the vehicle density estimation model to obtain a collaborative control model.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for constructing a collaborative control model as described in any one of claims 1 to 7 are implemented.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for constructing a collaborative control model as described in any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Road condition analyzing method based on Gestalt psychology criterion
CN109615874A
Traffic jam degree multi-dimensional analysis method based on deep learning
CN110992693A
Multi-lane roundabout vehicle cooperative control method in intelligent network connection vehicle environment
CN112837534A
Cloud component monitoring coordination method and system
CN114550483A
Vehicle feature identification method and system based on point cloud data
CN117284320A
