Method, device and positioning system for vehicle positioning using roadside acquisition equipment

Abstract

The present invention discloses a method, device, positioning system, and storage medium for vehicle positioning using roadside acquisition equipment. The method includes: obtaining image data of a vehicle collected by the roadside acquisition equipment, the image data having the acquisition time; identifying the vehicle's positioning feature parts from the image data to determine the first positioning information of the vehicle's positioning feature parts at the acquisition time; determining whether the first positioning information can match the corresponding vehicle being positioned and tracked, and if so, updating the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature parts of the corresponding vehicle; if not, treating the vehicle as a vehicle to be positioned and tracked based on the first positioning information; and using the posture of the corresponding vehicle at a certain moment as the vehicle's positioning information at that moment. The embodiments of the present invention do not rely on the vehicle's own high-precision positioning equipment and high-computing power computing platform, and can obtain the high-precision vehicle positioning required for autonomous driving.

Description

Technical Field

[0001] The present invention relates to the field of automobile communication technology, and in particular to a method, device, positioning system and storage medium for vehicle positioning using road-side acquisition equipment. Background Art

[0002] With the rapid development of autonomous vehicles, positioning has become almost a must-have feature. Currently, autonomous vehicles typically use the Global Positioning System (GPS) for positioning. However, GPS accuracy is poor or even inoperable indoors and in areas with dense buildings. Furthermore, many existing vehicles lack high-precision positioning equipment and high-performance computing platforms, making them unable to meet the high-precision positioning requirements required for autonomous driving. Summary of the Invention

[0003] The embodiments of the present invention provide a method, device, positioning system and storage medium for vehicle positioning using road-side acquisition equipment, which does not rely on the vehicle's own high-precision positioning equipment and high-computing power computing platform, and can obtain high-precision vehicle positioning required for autonomous driving.

[0004] An embodiment of the present invention provides a method for vehicle positioning using a roadside acquisition device, comprising:

[0005] Obtaining image data of the vehicle collected by a roadside collection device, the image data includes a collection time;

[0006] Identifying a characteristic location of the vehicle from the image data to determine first positioning information of the characteristic location of the vehicle at the acquisition time;

[0007] determining whether the first positioning information matches a corresponding vehicle being tracked; if so, updating the position of the corresponding vehicle using a difference between the first positioning information and the second positioning information of the positioning feature of the corresponding vehicle; if not, treating the vehicle as a tracked vehicle based on the first positioning information;

[0008] The position of the corresponding vehicle at a certain moment is used as the positioning information of the vehicle at that moment.

[0009] As an improvement to the above solution, identifying the characteristic location of the vehicle from the image data to determine the first location information of the characteristic location of the vehicle at the acquisition time includes:

[0010] identifying a characteristic location of the vehicle from the image data, and obtaining coordinate information of the characteristic location of the vehicle in the image data;

[0011] Performing perspective transformation on the coordinate information of the positioning feature portion in the image data to obtain the coordinate information of the positioning feature portion in the top view of the image data;

[0012] According to the conversion relationship between the world coordinate system and the coordinate system of the overhead view of the road-side acquisition device, the world coordinate information corresponding to the coordinate information in the overhead view is calculated, and the world coordinate information is used as the first positioning information of the characteristic part for positioning the vehicle at the time of acquisition.

[0013] As an improvement to the above solution, the method further includes:

[0014] When it is determined that any one of the positioning release conditions is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

[0015] As an improvement to the above solution, the positioning feature portion is any wheel hub;

[0016] The determining whether the first positioning information can match the corresponding vehicle being positioned and tracked includes:

[0017] Obtaining the positions of all vehicles being tracked at the time of acquisition;

[0018] Calculating third positioning information of a characteristic location of each vehicle being tracked at the acquisition time based on the position and posture of each vehicle being tracked and preset vehicle parameters; wherein the preset vehicle parameters include a wheelbase and a track width;

[0019] The third positioning information of each vehicle in the positioning tracking is associated and matched with the first positioning information to determine whether the first positioning information can match the corresponding vehicle in the positioning tracking.

[0020] As an improvement to the above solution, the first positioning information includes coordinate information of multiple points on the hub;

[0021] The correlating and matching the third positioning information of each vehicle being tracked with the first positioning information to determine whether the first positioning information can match the corresponding vehicle being tracked includes:

[0022] Constructing a cost matrix based on the third positioning information of each tracking vehicle and the first positioning information; wherein the element in the i-th row and j-th column of the cost matrix is the Euclidean distance between the third positioning information of the i-th vehicle being tracked and the coordinate information of the j-th point in the first positioning information;

[0023] Modifying the values of the elements in the cost matrix that exceed the first preset value to a second preset value to obtain a modified cost matrix;

[0024] Data association is performed based on the global nearest neighbor algorithm and the corrected cost matrix to determine whether third positioning information that matches the first positioning information can be associated; if so, it is determined that the first positioning information can be matched to the corresponding vehicle in positioning tracking; if not, it is determined that the first positioning information cannot be matched to the corresponding vehicle in positioning tracking.

[0025] As an improvement to the above solution, updating the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle includes:

[0026] Obtaining the position and prediction error covariance matrix of the corresponding vehicle at the acquisition time;

[0027] Calculating the filter gain at the acquisition moment according to a preset positioning conversion matrix between the vehicle and the positioning feature part, a preset field-side observation error covariance matrix, and the prediction error covariance matrix;

[0028] Calculating the difference according to the first positioning information, the positioning conversion matrix, and the posture at the acquisition moment;

[0029] updating the pose at the acquisition moment according to the difference and the filter gain, and predicting the pose at the next observation moment after the acquisition moment according to the updated pose at the acquisition moment;

[0030] The prediction error covariance matrix of the acquisition moment is updated according to the positioning conversion matrix and the filter gain, and the prediction error covariance matrix of the next observation moment of the acquisition moment is predicted according to the updated prediction error covariance matrix of the acquisition moment.

[0031] Another embodiment of the present invention provides a device for positioning a vehicle using a roadside acquisition device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements any one of the above methods for positioning a vehicle using a roadside acquisition device.

[0032] Another embodiment of the present invention provides a vehicle positioning system, comprising at least one roadside acquisition device and a vehicle tracking device; wherein,

[0033] The roadside acquisition device is used to collect image data of the vehicle, and the image data has the acquisition time;

[0034] The vehicle tracking device includes a first processing module and a second processing module;

[0035] The first processing module is specifically configured to:

[0036] Acquiring image data of the vehicle collected by the roadside collection device;

[0037] Identifying a characteristic location of the vehicle from the image data to determine first positioning information of the characteristic location of the vehicle at the acquisition time;

[0038] The second processing module is specifically configured to:

[0039] determining whether the first positioning information matches a corresponding vehicle being tracked; if so, updating the position of the corresponding vehicle using a difference between the first positioning information and the second positioning information of the positioning feature of the corresponding vehicle; if not, treating the vehicle as a tracked vehicle based on the first positioning information;

[0040] The position of the corresponding vehicle at a certain moment is used as the positioning information of the vehicle at that moment.

[0041] As an improvement to the above solution, the second processing module is further configured to:

[0042] When it is determined that any one of the positioning release conditions is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

[0043] As an improvement to the above solution, the first processing module is set at the road side, and the second processing module is set at the cloud side; or,

[0044] The first processing module and the second processing module are both set in the cloud.

[0045] As an improvement to the above solution, there are multiple road-end collection devices, and the multiple road-end collection devices are respectively arranged on both sides of the target road. The collection area of each road-end collection device corresponds to a different area of the target road.

[0046] Another embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the methods for vehicle positioning using a roadside acquisition device as described above.

[0047] Compared with the prior art, the method, device, positioning system and storage medium for vehicle positioning using road-side acquisition equipment provided in the embodiments of the present invention use the image data of the vehicle collected by the road-side acquisition equipment to determine the first positioning information of the vehicle's positioning feature parts at the time of collection, and when matching the corresponding vehicle in positioning and tracking, the posture of the corresponding vehicle is updated based on the difference between the first positioning information and the second positioning information of the positioning feature parts of the corresponding vehicle, or when the corresponding vehicle is not matched, the vehicle is used as the positioning and tracking vehicle based on the first positioning information to obtain the posture of the corresponding vehicle at a certain moment as the positioning information of the vehicle at that moment, thereby realizing positioning and tracking of multiple vehicles, and without relying on the high-precision positioning equipment and high-computing power computing platform of the vehicle itself, it is possible to obtain high-precision vehicle positioning that meets the requirements of autonomous driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flow chart of a method for vehicle positioning using a roadside acquisition device provided by one embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram of an application scenario of a method for vehicle positioning using a roadside acquisition device provided by an embodiment of the present invention;

[0050] Figure 3 is a schematic diagram of a vehicle coordinate system provided by one embodiment of the present invention;

[0051] Figure 4 This is a schematic structural diagram of a device for vehicle positioning using a roadside acquisition device provided by one embodiment of the present invention;

[0052] Figure 5 1 is a schematic structural diagram of a vehicle positioning system provided by one embodiment of the present invention;

[0053] Figure 6 1 is a schematic structural diagram of an optional vehicle positioning system provided in one embodiment of the present invention;

[0054] Figure 7 It is a structural diagram of another optional vehicle positioning system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] See also Figure 1 , is a flow chart of a method for vehicle positioning using a roadside acquisition device provided by an embodiment of the present invention.

[0057] An embodiment of the present invention provides a method for vehicle positioning using a roadside data acquisition device. The method can be performed by a roadside computing device, a cloud computing device, or a system consisting of both, and includes:

[0058] S11. Acquire image data of the vehicle collected by a roadside collection device, where the image data includes a collection time.

[0059] Exemplarily, the road-side acquisition device may be a wide-angle camera, and the road-side acquisition device is set on the road.

[0060] S12: Identify characteristic parts of the vehicle for positioning from the image data to determine first positioning information of the characteristic parts of the vehicle for positioning at the acquisition moment.

[0061] For example, the characteristic location part of the target vehicle may be a wheel hub, a rear taillight, or other parts of the vehicle body, which are not limited here.

[0062] S13. Determine whether the first positioning information can match the corresponding vehicle in positioning and tracking. If so, update the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle; if not, use the first positioning information as the vehicle to be positioned and tracked.

[0063] It should be noted that in order to achieve continuous tracking and positioning of multiple vehicles, this embodiment uses the first positioning information to determine whether the vehicle is the corresponding vehicle in the positioning and tracking. If so, it means that the vehicle is already in the tracking process. At this time, the error constructed according to the first positioning information and the second positioning information can be used to update the posture of the corresponding vehicle. If not, it means that the target vehicle appears for the first time. Based on the first positioning information, the vehicle is used as the positioning and tracking vehicle so that the vehicle can be effectively tracked later.

[0064] S14. Using the position of the corresponding vehicle at a certain moment as the positioning information of the vehicle at that moment.

[0065] Exemplarily, the posture includes a position and an attitude, wherein the position may be the center coordinates of the vehicle and the attitude may be the heading angle of the vehicle.

[0066] As one of the optional embodiments, the calibration parameters of the roadside acquisition device include intrinsic parameters and a conversion relationship between a world coordinate system and a top view coordinate system of the roadside acquisition device;

[0067] The identifying the characteristic part for positioning of the vehicle from the image data to determine the first positioning information of the characteristic part for positioning of the vehicle at the acquisition time includes:

[0068] S121, identifying characteristic parts for positioning of the vehicle from the image data, and obtaining coordinate information of the characteristic parts for positioning of the vehicle in the image data;

[0069] S122, performing perspective transformation on the coordinate information of the positioning feature part in the image data to obtain the coordinate information of the positioning feature part in the top view of the image data;

[0070] S123 . Calculate world coordinate information corresponding to the coordinate information in the top view according to the conversion relationship, and use the world coordinate information as first positioning information of the positioning feature part at the acquisition moment.

[0071] For example, in step S121, the yolov3 model may be pre-trained with training data marked with positioning feature parts to obtain a corresponding recognition model, so that the image data containing the vehicle is input into the recognition model to extract the coordinate information of the vehicle's positioning feature parts in the image data.

[0072] For example, taking the roadside acquisition equipment as a wide-angle camera set at both ends of the road, the intrinsic and extrinsic parameters of the camera are calibrated, and the vehicle image captured by the camera can be mapped into a bird's-eye view. Through scale calibration, the conversion relationship between the world coordinate system and the bird's-eye view coordinate system can be obtained. Specifically, the pixel coordinates and world coordinates of points within a certain range of the bird's-eye view can be measured using a checkerboard, and a linear difference calculation is performed for the points in the middle of the checkerboard to solve the conversion relationship. Based on this conversion relationship, the world coordinate information corresponding to the coordinate information in the bird's-eye view can be calculated. As a specific implementation method, the conversion relationship is: kRP c +T=P w , where k is the scaling factor, the unit of k is plxel / m, R is the two-dimensional rotation matrix, P c is the coordinate information in the top view, T is the translation vector (x t ,y t ), P w It is the world coordinate information.

[0073] As an optional embodiment, the method further includes:

[0074] S15. When it is determined that any positioning release condition is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

[0075] Through this embodiment, the vehicle's positioning information can be released in a timely manner when any positioning release condition is met, thereby meeting the vehicle's high-precision positioning needs.

[0076] As an optional implementation, the location release condition includes: receiving a location acquisition request for the vehicle from any terminal; wherein the location acquisition request includes the time information corresponding to the desired location information. Accordingly, the location release strategy corresponding to this location release condition is: sending the vehicle's location information corresponding to the time information to the terminal. It should be noted that any terminal can be the vehicle itself, or another terminal that needs to obtain the vehicle's location, such as a mobile terminal held by a user.

[0077] As another optional implementation, the location release condition includes the expiration of a set release period. Accordingly, the location release strategy corresponding to this location release condition is to send the vehicle's current location information to a designated terminal. It should be noted that the designated terminal can be the vehicle itself, or another terminal that requires vehicle location information, such as a mobile terminal held by a user.

[0078] Furthermore, the positioning characteristic portion is any wheel hub;

[0079] The determining whether the first positioning information can match the corresponding vehicle being positioned and tracked includes:

[0080] S1311, obtaining the positions of all vehicles being positioned and tracked at the acquisition time;

[0081] S1312: Calculate, based on the position and posture of each vehicle being tracked and preset vehicle parameters, third positioning information of the positioning characteristic portion of each vehicle being tracked at the acquisition time; wherein the preset vehicle parameters include a wheelbase and a track width;

[0082] S133: Perform an association match between the third positioning information of each vehicle being tracked and the first positioning information to determine whether the first positioning information can match the corresponding vehicle being tracked.

[0083] For example, see Figure 3 , let the origin of the vehicle coordinate system be the center of the vehicle's rear axle, X vehicle Refers to the X coordinate axis of the vehicle coordinate system, Y vehicle Refers to the Y coordinate axis of the vehicle coordinate system, Z vehicle Refers to the Z coordinate axis of the vehicle coordinate system. Assume that the world coordinate system is right front up (x, y, z). The vehicle center coordinates and vehicle heading angle x are known. w 、yw and θ, front and rear wheel track l1, wheelbase l2, then the corresponding relationship between the preset vehicle parameters and the predicted positioning of the four wheel hubs specifically includes:

[0084] Left front wheel (x1, y1): x1 = x w +l1 / 2cosθ+l2 / 2sinθ,y1=y w +l1 / 2sinθ-l2 / 2cOsθ;

[0085] Left rear wheel (x2, y2): x2 = x w -l1 / 2cosθ+l2 / 2sinθ,y2=y w -l1 / 2sinθ-l2 / 2cosθ;

[0086] Right front wheel (x3, y3): x3 = x w +l1 / 2cosθ-l2 / 2sinθ,y3=y w +l1 / 2sinθ+l2 / 2cosθ;

[0087] Right rear wheel (x4, y4): x4 = x w -l1 / 2cosθ-l2 / 2sinθ, y1=y w -l1 / 2sinθ+l2 / 2cosθ.

[0088] In this embodiment, since the difference between the wheel hub of the vehicle and the background part is more prominent, by setting the positioning feature part to any wheel hub, the accuracy of the first positioning information extraction can be guaranteed, thereby ensuring the accuracy of vehicle positioning. In addition, since the vehicle's posture is usually the center point position of the vehicle, there will be a certain deviation between the vehicle's posture and the vehicle's wheel hub positioning. If the posture of the vehicle in each positioning tracking is directly matched with the first positioning information, it is easy to cause an erroneous matching result due to the above-mentioned deviation. In this embodiment, the posture of the vehicle in each positioning tracking is first converted into the wheel hub positioning of each vehicle according to the preset vehicle parameters such as wheelbase and wheelbase, and then the association matching with the first positioning information is performed. Thus, an accurate vehicle association result can be obtained, ensuring the accuracy of vehicle positioning.

[0089] Furthermore, the first positioning information includes coordinate information of a plurality of points on the wheel hub;

[0090] The correlating and matching the third positioning information of each vehicle being tracked with the first positioning information to determine whether the first positioning information can match the corresponding vehicle being tracked includes:

[0091] Constructing a cost matrix based on the third positioning information of each tracking vehicle and the first positioning information; wherein the element in the i-th row and j-th column of the cost matrix is the Euclidean distance between the third positioning information of the i-th vehicle being tracked and the coordinate information of the j-th point in the first positioning information;

[0092] Modifying the values of the elements in the cost matrix that exceed the first preset value to a second preset value to obtain a modified cost matrix;

[0093] Data association is performed based on a global nearest neighbor algorithm and the modified cost matrix to determine whether third positioning information that matches the first positioning information can be associated; if so, it is determined that the first positioning information can be matched to the corresponding vehicle in positioning tracking; if not, it is determined that the first positioning information cannot be matched to the corresponding vehicle in positioning tracking.

[0094] Optionally, when performing data association, association results whose corresponding Euclidean distance exceeds a preset threshold may be filtered out, thereby further improving the accuracy of matching association.

[0095] It should be noted that both i and j are integers, i is not greater than the total number of vehicles in positioning and tracking, and j is not greater than the number of coordinate information included in the first positioning information.

[0096] In specific implementation, the first preset value and the second preset value may be set according to the correlation accuracy, which is not limited here. Optionally, in this embodiment, the second preset value is 100.

[0097] In this embodiment, by constructing and correcting a cost matrix, and performing data association according to a global nearest neighbor algorithm and the corrected cost matrix, an accurate association matching result can be obtained.

[0098] Specifically, updating the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle includes:

[0099] S1321. Obtain the position and prediction error covariance matrix of the corresponding vehicle at the acquisition time;

[0100] S1322. Calculate the filter gain at the acquisition time based on a preset positioning conversion matrix between the vehicle and the positioning feature part, a preset field-side observation error covariance matrix, and the prediction error covariance matrix;

[0101] S1323. Calculate the difference according to the first positioning information, the positioning conversion matrix, and the posture at the acquisition time;

[0102] S1324: updating the pose at the acquisition moment according to the difference and the filter gain, and predicting the pose at the next observation moment after the acquisition moment according to the updated pose at the acquisition moment;

[0103] S1325. Update the prediction error covariance matrix at the acquisition moment according to the positioning conversion matrix and the filter gain, and predict the prediction error covariance matrix at the next observation moment after the acquisition moment according to the updated prediction error covariance matrix at the acquisition moment.

[0104] In this embodiment, the posture of the corresponding vehicle at the time of collection is filtered and corrected by using the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle, which can effectively reduce the deviation of the vehicle posture prediction, make the posture of the corresponding vehicle closer to the actual driving trajectory, and improve the accuracy of vehicle positioning.

[0105] In one embodiment, the specific process of updating the position and posture of the corresponding vehicle is as follows:

[0106] (1) Obtaining the position x' of the corresponding vehicle at the acquisition time n and the prediction error covariance matrix P' n ;

[0107] (2) Based on the preset positioning conversion matrix H between the vehicle and the positioning feature parts, the preset field end observation error covariance matrix R, and the prediction error covariance matrix P' at the acquisition time n The filter gain K is calculated using the first preset formula, wherein the first preset formula is K=P' n H(HP' n H T +R) -1 ;

[0108] (3) According to the filtering gain K, the first positioning information Z k , the positioning transformation matrix H, the pose x' at the acquisition time n The pose at the acquisition time is updated using the second preset formula to obtain the updated pose x at the acquisition time. n , where the second preset formula is x n =x' n +K(Z k -H×x' n );

[0109] (4) According to the updated pose x at the acquisition time n , the preset state transfer matrix A and the third preset formula, predict the pose x' at the next observation moment of the acquisition momentn+1 , where the third preset formula is x' n+1 =Ax n ;

[0110] (5) Update the prediction error covariance matrix P' at the acquisition time according to the positioning conversion matrix H, the filter gain K and the fourth preset formula n , get the updated prediction error covariance matrix P at the acquisition moment n , where the fourth preset formula is P n =(I-KH)P' n , I is the identity matrix;

[0111] (6) According to the updated prediction error covariance matrix P at the acquisition time n , Jacobian matrix F, preset process covariance matrix Q and the fifth preset formula, predict the prediction error covariance matrix P' of the next observation moment of the acquisition moment n+1 , where the fifth preset formula is P' n+1 =FP n F T +Q, F refers to the first-order derivative of the preset state transfer matrix to the state quantity.

[0112] As an optional implementation, the method for vehicle positioning using a roadside acquisition device provided in this embodiment can be applied to, but not limited to, Figure 2 In the environment shown, the roadside acquisition device 11 may be a wide-angle camera, and multiple roadside acquisition devices 11 are respectively set on both sides of the road, and the height of the roadside acquisition devices 11 may be about 0.5 meters.

[0113] As a specific implementation, the method provided in this embodiment can be executed by a cloud computing device, so that all data processing calculations are placed in the cloud. Thus, cloud computing is used to speed up the calculation speed, reduce the time required for calculation, and thus improve the efficiency of continuous positioning of multiple vehicles. Taking the cloud computing device executing the method for vehicle positioning using roadside data collection equipment provided in this embodiment as an example, the overall process of the method for vehicle positioning using roadside data collection equipment according to the embodiment of the present invention is described as follows:

[0114] Step 1: The roadside acquisition device 11 collects image data and uploads the image data to the cloud computing device 12 via the network 13;

[0115] Step 2: The cloud computing device 12 identifies the coordinate information of the vehicle's wheel hub in the image data from the image data, and performs perspective transformation and world coordinate system transformation on the coordinate information of the wheel hub in the image data to obtain first positioning information of the vehicle's wheel hub at the time of acquisition;

[0116] Step 3: The cloud computing device 12 compares the first positioning information with the third positioning information of the wheels of all the vehicles being tracked at the time of collection to determine whether the first positioning information matches the corresponding vehicle being tracked. If so, an error is constructed based on the first positioning information and the second positioning information of the corresponding vehicle to update the position of the corresponding vehicle. If not, the vehicle is identified as the vehicle being tracked based on the first positioning information to track the vehicle.

[0117] Step 4: The cloud computing device 12 uses the position of the corresponding vehicle at a certain moment as the positioning information of the vehicle at that moment, and sends the positioning information of the vehicle at the current moment to the designated terminal when the set publishing cycle arrives.

[0118] As another specific implementation method, the method provided in this embodiment can be executed by a system composed of a cloud computing device and a road-side computing device. Since the amount of image data is usually large, in this embodiment, the road-side computing device first processes the image data to obtain first positioning information with a smaller amount of data. The road-side computing device then uploads the first positioning information to the cloud computing device, rather than uploading all image data to the cloud computing device. This can reduce network transmission pressure, reduce network congestion, speed up network transmission rate, and thus improve the efficiency of continuous positioning of multiple vehicles. Taking the method of using road-side acquisition equipment to perform vehicle positioning provided by this embodiment as an example, the overall process of the method of using road-side acquisition equipment to perform vehicle positioning according to the embodiment of the present invention is described as follows:

[0119] Step 1: The roadside acquisition device collects image data and uploads the image data to the roadside computing device via the network;

[0120] Step 2: The road-side computing device identifies the coordinate information of the vehicle's wheel hub in the image data from the image data, performs perspective transformation and world coordinate system transformation on the coordinate information of the wheel hub in the image data to obtain first positioning information of the vehicle's wheel hub at the time of acquisition, and sends the first positioning information to the cloud computing device;

[0121] Step 3: The cloud computing device compares the first positioning information with the third positioning information of the wheels of all vehicles being tracked at the time of collection to determine whether the first positioning information matches the corresponding vehicle being tracked. If so, an error is constructed based on the first positioning information and the second positioning information of the corresponding vehicle to update the position of the corresponding vehicle. If not, the vehicle is identified as the vehicle being tracked based on the first positioning information to track the vehicle.

[0122] Step 4: The cloud computing device uses the position of the corresponding vehicle at a certain moment as the positioning information of the vehicle at that moment, and sends the positioning information of the vehicle at the current moment to the designated terminal when the set release cycle arrives.

[0123] Optionally, in this embodiment, the aforementioned network may include, but is not limited to, wired networks and wireless networks. Wired networks include local area networks, metropolitan area networks, and wide area networks, and wireless networks include Bluetooth, Wi-Fi, and other networks that enable wireless communication. The aforementioned cloud computing device may be a single server, a server cluster consisting of multiple servers, or a cloud server. The above is merely an example and is not intended to be limiting in this embodiment.

[0124] Compared with the existing technology, the beneficial effect of the embodiment of the present invention is that: this embodiment can realize the positioning and tracking of multiple vehicles, and does not rely on the high-precision positioning equipment and high-computing power computing platform of the vehicle itself, and can obtain the high-precision positioning of the vehicle required for autonomous driving.

[0125] See also Figure 4 , is a structural diagram of a device for vehicle positioning using a roadside acquisition device provided by an embodiment of the present invention.

[0126] An embodiment of the present invention provides a device 30 for vehicle positioning using a roadside acquisition device, comprising a processor 31, a memory 32, and a computer program stored in the memory 32 and configured to be executed by the processor 31. When the processor 31 executes the computer program, it implements the method for vehicle positioning using a roadside acquisition device as described in any of the above embodiments.

[0127] When the processor 31 executes the computer program, the steps in the above-mentioned method embodiment for vehicle positioning using a roadside acquisition device are implemented, for example: Figure 1 All steps of the method for vehicle positioning using a roadside data acquisition device are shown. For example, the computer program can be divided into one or more modules, which are stored in the memory 32 and executed by the processor 31 to implement the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution process of the computer program in the device 30 for vehicle positioning using a roadside data acquisition device.

[0128] The device 30 for locating a vehicle using a roadside data acquisition device can be a roadside computing device, a cloud computing device, or a combination of both a roadside computing device and a cloud computing device. The device 30 for locating a vehicle using a roadside data acquisition device can include, but is not limited to, a processor 31 and a memory 32. It will be understood by those skilled in the art that the schematic diagram is merely an example of the device 30 for locating a vehicle using a roadside data acquisition device and does not constitute a limitation on the device 30 for locating a vehicle using a roadside data acquisition device. The device 30 can include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the device 30 for locating a vehicle using a roadside data acquisition device can also include input and output devices, network access devices, buses, etc.

[0129] The processor 31 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor. The processor 31 is the control center of the device 30 for vehicle positioning using a roadside data acquisition device, and connects various parts of the device 30 for vehicle positioning using a roadside data acquisition device using various interfaces and lines.

[0130] The memory 32 can be used to store the computer programs and / or modules. The processor 31 implements the various functions of the device 30 for vehicle positioning using roadside data acquisition equipment by running or executing the computer programs and / or modules stored in the memory 32 and accessing the data stored in the memory 32. The memory 32 may primarily include a program storage area and a data storage area. The program storage area may store an operating system, at least one application required for a function, and the data storage area may store data generated based on the use of the device 30 for vehicle positioning using roadside data acquisition equipment. Furthermore, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0131] Wherein, if the module / unit integrated in the device 30 for vehicle positioning using roadside acquisition equipment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of each of the above-mentioned method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0132] See also Figure 5 , is a structural diagram of a vehicle positioning system provided by one embodiment of the present invention.

[0133] The vehicle positioning system provided in this embodiment includes at least one roadside acquisition device 40 and a vehicle tracking device 50; wherein,

[0134] The roadside acquisition device 40 is used to collect image data of the vehicle, and the image data includes the acquisition time;

[0135] The vehicle tracking device 50 includes a first processing module 51 and a second processing module 52;

[0136] The first processing module 51 is specifically configured to:

[0137] Acquiring image data of the vehicle collected by the roadside collection device 40;

[0138] Identifying a characteristic location of the vehicle from the image data to determine first positioning information of the characteristic location of the vehicle at the acquisition time;

[0139] The second processing module 52 is specifically configured to:

[0140] determining whether the first positioning information matches a corresponding vehicle being tracked; if so, updating the position of the corresponding vehicle using a difference between the first positioning information and the second positioning information of the positioning feature of the corresponding vehicle; if not, treating the vehicle as a tracked vehicle based on the first positioning information;

[0141] The position of the corresponding vehicle at a certain moment is used as the positioning information of the vehicle at that moment.

[0142] As an optional embodiment, the second processing module 52 is further configured to:

[0143] When it is determined that any one of the positioning release conditions is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

[0144] As one of the optional embodiments, the first processing module 51 is set at the road side, and the second processing module 52 is set at the cloud side; or,

[0145] The first processing module 51 and the second processing module 52 are both set in the cloud.

[0146] As a specific embodiment, see Figure 6 Since the amount of image data is usually large, this embodiment sets the first processing module 51 at the road end very close to the image data source and the second processing module 52 at the cloud end, so that the image data can be first processed by the first processing module 51 at the road end to obtain first positioning information with a smaller amount of data. The first processing module 51 at the road end then uploads the first positioning information to the second processing module 52 set at the cloud end, instead of uploading all the image data to the cloud end. Therefore, in the process of continuous positioning of multiple vehicles, the network transmission pressure can be reduced, network congestion can be reduced, and the network transmission rate can be accelerated, thereby improving the efficiency of continuous positioning of multiple vehicles.

[0147] As another specific embodiment, see Figure 7 In this embodiment, by setting both the first processing module 51 and the second processing module 52 in the cloud, all data processing calculations can be placed in the cloud, thereby utilizing cloud computing to speed up the computing speed, reducing the time required for computing, and thereby improving the efficiency of continuous positioning of multiple vehicles.

[0148] As one of the optional embodiments, there are multiple road-side collection devices 40, and the multiple road-side collection devices 40 are respectively arranged on both sides of the target road, and the collection area of each road-side collection device 40 corresponds to a different area of the target road.

[0149] In this embodiment, multiple road-end acquisition devices 40 are set on both sides of the target road. When multiple vehicles pass by the target road respectively, the multiple road-end acquisition devices 40 can continuously capture images of the driving process of the multiple vehicles, so that the vehicle tracking device 50 can continue to stably and continuously locate and track the multiple vehicles based on the multiple image data during the driving process of the multiple vehicles.

[0150] As one of the optional embodiments, identifying the characteristic location of the vehicle from the image data to determine the first positioning information of the characteristic location of the vehicle at the acquisition time includes:

[0151] identifying a characteristic location of the vehicle from the image data, and obtaining coordinate information of the characteristic location of the vehicle in the image data;

[0152] Performing perspective transformation on the coordinate information of the positioning feature portion in the image data to obtain the coordinate information of the positioning feature portion in the top view of the image data;

[0153] According to the conversion relationship between the world coordinate system and the coordinate system of the top view of the road-side acquisition device 40, the world coordinate information corresponding to the coordinate information in the top view is calculated, and the world coordinate information is used as the first positioning information of the characteristic part for positioning the vehicle at the acquisition time.

[0154] As one of the optional embodiments, the positioning feature portion is any wheel hub;

[0155] The determining whether the first positioning information can match the corresponding vehicle being positioned and tracked includes:

[0156] Obtaining the positions of all vehicles being tracked at the time of acquisition;

[0157] Calculating third positioning information of a characteristic location of each vehicle being tracked at the acquisition time based on the position and posture of each vehicle being tracked and preset vehicle parameters; wherein the preset vehicle parameters include a wheelbase and a track width;

[0158] The third positioning information of each vehicle in the positioning tracking is associated and matched with the first positioning information to determine whether the first positioning information can match the corresponding vehicle in the positioning tracking.

[0159] Furthermore, the first positioning information includes coordinate information of a plurality of points on the wheel hub;

[0160] The correlating and matching the third positioning information of each vehicle being tracked with the first positioning information to determine whether the first positioning information can match the corresponding vehicle being tracked includes:

[0161] Constructing a cost matrix based on the third positioning information of each tracking vehicle and the first positioning information; wherein the element in the i-th row and j-th column of the cost matrix is the Euclidean distance between the third positioning information of the i-th vehicle being tracked and the coordinate information of the j-th point in the first positioning information;

[0162] Modifying the values of the elements in the cost matrix that exceed the first preset value to a second preset value to obtain a modified cost matrix;

[0163] Data association is performed based on the global nearest neighbor algorithm and the corrected cost matrix to determine whether third positioning information that matches the first positioning information can be associated; if so, it is determined that the first positioning information can be matched to the corresponding vehicle in positioning tracking; if not, it is determined that the first positioning information cannot be matched to the corresponding vehicle in positioning tracking.

[0164] As one of the optional embodiments, updating the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle includes:

[0165] Obtaining the position and prediction error covariance matrix of the corresponding vehicle at the acquisition time;

[0166] Calculating the filter gain at the acquisition moment according to a preset positioning conversion matrix between the vehicle and the positioning feature part, a preset field-side observation error covariance matrix, and the prediction error covariance matrix;

[0167] Calculating the difference according to the first positioning information, the positioning conversion matrix, and the posture at the acquisition moment;

[0168] updating the pose at the acquisition moment according to the difference and the filter gain, and predicting the pose at the next observation moment after the acquisition moment according to the updated pose at the acquisition moment;

[0169] The prediction error covariance matrix of the acquisition moment is updated according to the positioning conversion matrix and the filter gain, and the prediction error covariance matrix of the next observation moment of the acquisition moment is predicted according to the updated prediction error covariance matrix of the acquisition moment.

[0170] It should be noted that those skilled in the art can clearly understand that, for the convenience and conciseness of description, the beneficial effects of the system described in the above embodiment and the specific working process of each module can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0171] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0172] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for vehicle positioning using roadside acquisition equipment, characterized in that: include: Obtaining image data of the vehicle collected by a roadside collection device, the image data includes a collection time; Identifying a characteristic location of the vehicle from the image data, wherein the characteristic location is any wheel hub; and determining first positioning information of the characteristic location of the vehicle at the acquisition time; determining whether the first positioning information matches a corresponding vehicle being tracked; if so, updating the position of the corresponding vehicle using a difference between the first positioning information and the second positioning information of the positioning feature of the corresponding vehicle; if not, treating the vehicle as a tracked vehicle based on the first positioning information; The determining whether the first positioning information can match the corresponding vehicle being positioned and tracked includes: Obtaining the position and posture of all vehicles being tracked at the time of collection; calculating third positioning information of a characteristic location of each vehicle being tracked at the time of collection based on the position and posture of each vehicle being tracked and preset vehicle parameters; wherein the preset vehicle parameters include wheelbase and track width; correlating and matching the third positioning information of each vehicle being tracked with the first positioning information to determine whether the first positioning information matches the corresponding vehicle being tracked; The position of the corresponding vehicle at a certain moment is used as the positioning information of the vehicle at that moment.

2. The method for vehicle positioning using roadside acquisition equipment according to claim 1, characterized in that: The identifying the characteristic part for positioning of the vehicle from the image data to determine the first positioning information of the characteristic part for positioning of the vehicle at the acquisition time includes: identifying a characteristic location of the vehicle from the image data, and obtaining coordinate information of the characteristic location of the vehicle in the image data; Performing perspective transformation on the coordinate information of the positioning feature portion in the image data to obtain the coordinate information of the positioning feature portion in the top view of the image data; According to the conversion relationship between the world coordinate system and the coordinate system of the overhead view of the road-side acquisition device, the world coordinate information corresponding to the coordinate information in the overhead view is calculated, and the world coordinate information is used as the first positioning information of the characteristic part for positioning the vehicle at the time of acquisition.

3. The method for vehicle positioning using roadside acquisition equipment according to claim 1, characterized in that: The method further comprises: When it is determined that any one of the positioning release conditions is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

4. The method for vehicle positioning using roadside acquisition equipment according to claim 1, characterized in that: The first positioning information includes coordinate information of multiple points on the wheel hub; The correlating and matching the third positioning information of each vehicle being tracked with the first positioning information to determine whether the first positioning information can match the corresponding vehicle being tracked includes: Constructing a cost matrix based on the third positioning information of each tracking vehicle and the first positioning information; wherein the element in the i-th row and j-th column of the cost matrix is the Euclidean distance between the third positioning information of the i-th vehicle being tracked and the coordinate information of the j-th point in the first positioning information; Modifying the values of the elements in the cost matrix that exceed the first preset value to a second preset value to obtain a modified cost matrix; Data association is performed based on the global nearest neighbor algorithm and the corrected cost matrix to determine whether third positioning information that matches the first positioning information can be associated; if so, it is determined that the first positioning information can be matched to the corresponding vehicle in positioning tracking; if not, it is determined that the first positioning information cannot be matched to the corresponding vehicle in positioning tracking.

5. The method for vehicle positioning using roadside acquisition equipment according to claim 1, characterized in that: The updating of the posture of the corresponding vehicle based on the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle includes: Obtaining the position and prediction error covariance matrix of the corresponding vehicle at the acquisition time; Calculating the filter gain at the acquisition moment according to a preset positioning conversion matrix between the vehicle and the positioning feature part, a preset field-side observation error covariance matrix, and the prediction error covariance matrix; Calculating the difference according to the first positioning information, the positioning conversion matrix, and the posture at the acquisition moment; updating the pose at the acquisition moment according to the difference and the filter gain, and predicting the pose at the next observation moment after the acquisition moment according to the updated pose at the acquisition moment; The prediction error covariance matrix of the acquisition moment is updated according to the positioning conversion matrix and the filter gain, and the prediction error covariance matrix of the next observation moment of the acquisition moment is predicted according to the updated prediction error covariance matrix of the acquisition moment.

6. A device for vehicle positioning using roadside acquisition equipment, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for vehicle positioning using a roadside acquisition device as described in any one of claims 1 to 5 is implemented.

7. A vehicle positioning system, characterized in that: It includes at least one roadside collection device and vehicle tracking device; wherein, The roadside acquisition device is used to collect image data of the vehicle, and the image data has the acquisition time; The vehicle tracking device includes a first processing module and a second processing module; The first processing module is specifically configured to: Acquiring image data of the vehicle collected by the roadside collection device; Identifying a characteristic location of the vehicle from the image data, wherein the characteristic location is any wheel hub; and determining first positioning information of the characteristic location of the vehicle at the acquisition time; The second processing module is specifically configured to: Determine whether the first positioning information can match the corresponding vehicle in positioning and tracking. If so, update the posture of the corresponding vehicle with the difference between the first positioning information and the second positioning information of the positioning feature part of the corresponding vehicle; if not, use the vehicle as the positioning and tracking vehicle based on the first positioning information; the determination of whether the first positioning information can match the corresponding vehicle in positioning and tracking includes: obtaining the posture of all vehicles in positioning and tracking at the acquisition time; calculating the third positioning information of the positioning feature part of each vehicle in positioning and tracking at the acquisition time based on the posture of each vehicle in positioning and tracking and preset vehicle parameters; wherein the preset vehicle parameters include wheelbase and wheelbase; perform correlation matching on the third positioning information of each vehicle in positioning and tracking and the first positioning information to determine whether the first positioning information can match the corresponding vehicle in positioning and tracking; The position of the corresponding vehicle at a certain moment is used as the positioning information of the vehicle at that moment.

8. The vehicle positioning system according to claim 7, wherein: The second processing module is further configured to: When it is determined that any one of the positioning release conditions is satisfied, the corresponding positioning information of the vehicle is released according to a positioning release strategy corresponding to the satisfied positioning release condition.

9. The vehicle positioning system according to claim 8, wherein: The first processing module is set at the road side, and the second processing module is set at the cloud side; or, The first processing module and the second processing module are both set in the cloud.

10. The vehicle positioning system according to claim 8, wherein: There are multiple roadside collection devices, and the multiple roadside collection devices are respectively arranged on both sides of the target road. The collection area of each roadside collection device corresponds to a different area of the target road.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method for vehicle positioning using a roadside acquisition device as described in any one of claims 1 to 5.

Images