Speed initialization method and apparatus

Abstract

The application discloses a speed initialization method and device, comprising: based on the motion parameters corresponding to the N feature point clouds of the first image frame of the target vehicle obtained, performing speed profile calculation on the N feature point clouds to obtain the first speed information of the target vehicle; based on the first speed information, calculating the first heading angle information of the target vehicle when the first image frame is shot; in the case where it is determined that the first heading angle information meets the first preset condition, determining the first speed information as the initialization speed after the trajectory speed of the target vehicle is initialized; in the case where it is determined that the first heading angle information does not meet the first preset condition, fitting the first position information of the target vehicle when each image frame in M image frames is shot to obtain the second heading angle information of the target vehicle in the first shooting period, and based on the second heading angle information, initializing the trajectory speed of the target vehicle to obtain the initialization speed, so as to accurately initialize the trajectory speed.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, specifically to a speed initialization method and apparatus. Background Technology

[0002] In recent years, autonomous driving technology has developed rapidly and attracted increasing attention. Environmental perception, path planning, and decision-making and control are the three main technological components of autonomous driving. Among them, environmental perception is the foundation of the other two; only with good environmental perception can better path planning be achieved and correct decisions be made.

[0003] Currently, millimeter-wave radar is commonly used in environmental perception technology to track targets. The target's trajectory velocity is then initialized based on the tracked trajectory. The quality of the target's trajectory velocity initialization directly affects the speed of subsequent target tracking convergence. If the velocity initialization differs significantly from the target's true velocity, the trajectory tracking will have difficulty converging to the true velocity, potentially leading to the loss of the target's trajectory.

[0004] Currently, when initializing the trajectory velocity of a target using millimeter-wave radar, the radial velocity of the target is usually projected laterally and longitudinally to initialize the trajectory velocity. However, this method is not precise enough for initializing the trajectory velocity. Summary of the Invention

[0005] The purpose of this application is to provide a speed initialization method and apparatus for accurately initializing track speed.

[0006] The technical solution of this application is as follows:

[0007] Firstly, a speed initialization method is provided, which includes:

[0008] Acquire a first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, where N is a positive integer;

[0009] Based on the motion parameters corresponding to N feature point clouds, velocity profiles are calculated for the N feature point clouds to obtain the first velocity information of the target vehicle.

[0010] Based on the first speed information, calculate the first heading angle information of the target vehicle when the first image frame is captured;

[0011] If the first heading angle information is determined to meet the first preset condition, the first speed information is determined to be the initial speed after initializing the trajectory speed of the target vehicle;

[0012] If the first heading angle information does not meet the first preset condition, the first position information of the target vehicle at each frame of the M-frame image is fitted to obtain the second heading angle information of the target vehicle during the first shooting period. Based on the second heading angle information, the trajectory speed of the target vehicle is initialized to obtain the initialized speed. The M-frame image includes the first image frame, the first shooting period is the shooting period corresponding to the M-frame image, and M is a positive integer.

[0013] Secondly, a speed initialization device is provided, the device comprising:

[0014] The acquisition module is used to acquire a first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, where N is a positive integer;

[0015] The first determining module is used to calculate the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds, so as to obtain the first velocity information of the target vehicle.

[0016] The first calculation module is used to calculate the first heading angle information of the target vehicle when the first image frame is captured, based on the first speed information;

[0017] The second determining module is used to determine the first speed information as the initial speed after initializing the trajectory speed of the target vehicle when the first heading angle information satisfies the first preset condition.

[0018] The third determining module is used to, when determining that the first heading angle information does not meet the first preset condition, fit the first position information of the target vehicle when each image frame in the M-frame image frames is captured, to obtain the second heading angle information of the target vehicle during the first shooting period, and initialize the trajectory speed of the target vehicle based on the second heading angle information to obtain the initialized speed, wherein the M-frame image frames include the first image frames, the first shooting period is the shooting period corresponding to the M-frame image frames, and M is a positive integer.

[0019] Thirdly, embodiments of this application provide an electronic device, which includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the steps of any of the speed initialization methods described in embodiments of this application.

[0020] Fourthly, embodiments of this application provide a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of any of the speed initialization methods described in embodiments of this application.

[0021] Fifthly, embodiments of this application provide a computer program product, wherein instructions in the computer program product, when executed by a processor of an electronic device, enable the electronic device to perform the steps of any of the speed initialization methods described in embodiments of this application.

[0022] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0023] In this embodiment, the first velocity information of the target vehicle is obtained by calculating the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds in the first image frame of the target vehicle. Then, based on the first velocity information, the first heading angle information of the target vehicle when the first image frame was captured is calculated. If the first heading angle information meets the first preset condition, the first velocity information is determined as the initial velocity after initializing the trajectory velocity of the target vehicle. If the first heading angle information does not meet the first preset condition, the first position information of the target vehicle when each image frame in the M image frames is captured is fitted to obtain the second heading angle information of the target vehicle in the first capturing period. Based on the second heading angle information, the trajectory velocity of the target vehicle is initialized to obtain the initial velocity. In this way, instead of mapping the radial velocity of the target vehicle in the horizontal and vertical directions, the velocity profile analysis of the N feature point clouds of the first image frame of the target vehicle is performed, or if the first heading angle information corresponding to multiple image frames is inaccurate, the trajectory velocity is initialized by fitting the first position information of the target vehicle when multiple image frames are captured. This improves the accuracy of trajectory velocity initialization.

[0024] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application, and do not constitute an undue limitation of this application.

[0026] Figure 1 This is a schematic diagram illustrating the principle of radial velocity provided in an embodiment of this application;

[0027] Figure 2 This is a flowchart illustrating a speed initialization method provided in an embodiment of this application;

[0028] Figure 3 This is a schematic diagram illustrating the fitting of the first position information of the target vehicle when capturing multiple image frames, as provided in an embodiment of this application.

[0029] Figure 4 This is a flowchart illustrating a speed initialization method provided in an embodiment of this application;

[0030] Figure 5 This is a schematic diagram of the structure of a speed initialization device provided in an embodiment of this application;

[0031] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0032] To enable those skilled in the art to better understand the technical solutions of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0033] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples consistent with some aspects of this application as detailed in the appended claims.

[0034] Before introducing the technical solutions of the embodiments of this application, let's first introduce the professional names involved in the embodiments of this application:

[0035] Radial velocity: Generally refers to the velocity component of an object's motion in the direction of the observer's line of sight, that is, the projection of the velocity vector in the direction of the line of sight, also known as the line-of-sight velocity.

[0036] In one example, reference Figure 1 , Figure 1 Vehicle 11 is the target vehicle, and vehicle 12 is a vehicle equipped with image acquisition equipment, such as a vehicle equipped with 3D or 4D millimeter-wave radar. Vehicle 12 and vehicle 11 meet at an intersection. At this time, vehicle 11 is directly in front of vehicle 12. If vehicle 11 wants to cross the intersection, its speed is 10 km / h. However, at this time, vehicle 11 is neither moving away from nor approaching vehicle 12. Therefore, from the perspective of vehicle 12, the radial speed of vehicle 11 is 0 km / h.

[0037] The background technology of the embodiments of this application is introduced below:

[0038] Currently, when initializing the trajectory velocity of a target using millimeter-wave radar, the radial velocity of the target is usually projected laterally and longitudinally to initialize the trajectory velocity. However, this method is not precise enough for initializing the trajectory velocity.

[0039] Continuing with the example above, vehicle 11 has a radial velocity of 0 km / h, and its velocity components in both the horizontal and longitudinal directions are 0 km / h. If the velocity of vehicle 11 is initialized based on the horizontal and longitudinal components of its radial velocity, then the initialized velocity of vehicle 11 will be 0 km / h. However, the actual velocity of vehicle 11 is 10 km / h. Therefore, the initialized velocity of vehicle 11 differs greatly from its actual velocity. Thus, initializing the track velocity using the above method is not accurate enough, which can easily lead to the loss of the track of vehicle 11 when it is being tracked.

[0040] To address the aforementioned problems, this application provides a velocity initialization method and apparatus. Based on motion parameters corresponding to N feature point clouds in a first image frame of the target vehicle, a velocity profile is calculated from the N feature point clouds to obtain first velocity information of the target vehicle. Then, based on the first velocity information, the first heading angle information of the target vehicle at the time the first image frame was captured is calculated. If the first heading angle information satisfies a first preset condition, the first velocity information is determined as the initial velocity after initializing the trajectory velocity of the target vehicle. If the first heading angle information does not satisfy the first preset condition, the velocity initialization is performed on M captured image frames. The first position information of the target vehicle in each image frame is fitted to obtain the second heading angle information of the target vehicle in the first shooting period. Based on the second heading angle information, the trajectory velocity of the target vehicle is initialized to obtain the initialized velocity. In this way, instead of mapping the radial velocity of the target vehicle in the lateral and longitudinal directions, the velocity profile analysis is performed on the N feature point clouds of the target vehicle in the first image frame. Alternatively, if it is determined that the first heading angle information corresponding to multiple image frames is inaccurate, the trajectory velocity is initialized by fitting the first position information of the target vehicle when shooting multiple image frames. This improves the accuracy of trajectory velocity initialization.

[0041] The speed initialization method provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0042] It should be noted that the example scenarios in the following embodiments of this application are all based on the above. Figure 1 Let's take a scenario from the example to illustrate.

[0043] Figure 2 This is a flowchart illustrating a speed initialization method provided in an embodiment of this application. The execution entity of this speed initialization method can be a processor, such as... Figure 2 As shown, the speed initialization method provided in this application embodiment may include steps 210-250.

[0044] Step 210: Obtain the first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, where N is a positive integer.

[0045] The target vehicle can be the vehicle whose trajectory and speed are to be initialized, as described above. Figure 1 Vehicle 11 in the middle.

[0046] The first image frame may be an image frame of the target vehicle taken at a certain moment. The first image frame may include N feature point clouds of the target vehicle. Here, the feature point clouds may be point cloud data used to characterize the feature points of the vehicle, such as point cloud data of feature points of the vehicle body, tires, and front of the vehicle.

[0047] In some embodiments of this application, the first image frame of the target vehicle may be acquired by an image acquisition device on the first vehicle. The image acquisition device may be, but is not limited to, a camera, millimeter-wave radar, or lidar. Since millimeter-wave radar has the characteristic of being less affected by all-weather environment, the image acquisition device here is described using 3D millimeter-wave radar as an example.

[0048] The first vehicle can be any vehicle in the same scene as the target vehicle, whose image acquisition device can clearly capture image frames of the target vehicle, such as the one described above. Figure 1 Vehicle 12.

[0049] Step 220: Based on the motion parameters corresponding to the N feature point clouds, calculate the velocity profile of the N feature point clouds to obtain the first velocity information of the target vehicle.

[0050] The motion parameters can be parameters used to describe the motion state of a certain feature point cloud. For each feature point cloud, the corresponding motion parameters may include: radial distance information between the feature point cloud and the first vehicle, azimuth information of the feature point cloud relative to the first vehicle, and radial velocity information of the feature point cloud relative to the first vehicle.

[0051] The aforementioned radial distance information could be the radial distance between the feature point cloud and the first vehicle.

[0052] The first velocity information can be the velocity information of the target vehicle when the first image frame is captured, obtained by calculating the velocity profile of N feature point clouds.

[0053] In some embodiments of this application, in order to accurately obtain the first velocity information, step 220 may specifically include:

[0054] For each feature point cloud, the second velocity information of the feature point cloud is calculated based on the azimuth information and radial velocity information;

[0055] A loss function is constructed based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud;

[0056] Find the optimal solution for the loss function to obtain the first speed information of the target vehicle.

[0057] The second velocity information can be the velocity information of the feature point cloud in the target vehicle when the first image frame is captured, calculated based on the azimuth information and radial velocity information for each feature point cloud.

[0058] In some embodiments of this application, for each feature point cloud, the second velocity information of the feature point cloud can be calculated according to the following formula (1) based on the radial distance information, azimuth information, and radial velocity information:

[0059]

[0060] Among them, V di θ represents the second velocity information of the i-th feature point cloud among N feature point clouds in the first image frame. i The azimuth information of the i-th feature point cloud among N feature point clouds in the first image frame. For the pitch angle information of the i-th feature point cloud among N feature point clouds in the first image frame, (V x V y V z ) represents the radial velocity information of the i-th feature point cloud among the N feature point clouds in the first image frame.

[0061] In some embodiments of this application, the radial velocity of each feature point cloud in the first image frame is actually consistent with the radial velocity of the target vehicle when the first image frame was captured. Therefore, the radial velocity information of each feature point cloud in the first image frame is the same. That is, for N feature point clouds in the first image frame, their radial velocity information is (V x V y V z ).

[0062] After obtaining the second velocity information of each feature point cloud, a loss function as shown in formula (2) can be constructed based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud. Then, the optimal solution of the loss function shown in formula (2) can be obtained. Specifically, it can be to find (V x Vy V z The minimum solution can be used to obtain the first speed information of the target vehicle.

[0063]

[0064] In the embodiments of this application, for each feature point cloud, the second velocity information of the feature point cloud is calculated based on the azimuth information and radial velocity information. Then, based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud, a loss function is constructed. By finding the optimal solution of the loss function, the first velocity information of the target vehicle can be accurately obtained.

[0065] In some embodiments of this application, if the image acquisition device on the first vehicle is a 3D millimeter-wave radar, then for each feature point cloud, its motion parameters include the azimuth information of the feature point cloud relative to the first vehicle and the radial velocity information of the feature point cloud relative to the first vehicle. However, if the image acquisition device on the first vehicle is a 4D millimeter-wave radar, then for each feature point cloud, its motion parameters may further include: the pitch angle information of the feature point cloud relative to the first vehicle. Therefore, calculating the second velocity information of each feature point cloud based on the azimuth information and the radial velocity information may specifically include:

[0066] For each feature point cloud, the second velocity information of the feature point cloud is calculated based on the azimuth angle information, pitch angle information, and radial velocity information.

[0067] In the embodiments of this application, when the image acquisition device on the first vehicle is a 4D millimeter-wave radar, the pitch angle information of the feature point cloud relative to the first vehicle can also be acquired. Thus, when calculating the second velocity information of the feature point cloud, in addition to referring to the azimuth information and radial velocity information, the pitch angle information of the feature point cloud relative to the first vehicle can also be referenced. The obtained second velocity information will be more accurate, and the trajectory velocity of the target vehicle can be initialized more accurately.

[0068] Step 230: Based on the first velocity information, calculate the first heading angle information of the target vehicle when the first image frame is captured.

[0069] The first heading angle information can be the heading angle information of the target vehicle relative to the first vehicle when the first image frame is captured, obtained by calculating the first velocity information.

[0070] In some embodiments of this application, based on the first velocity information, the first heading angle information of the target vehicle when the first image frame is captured can be obtained according to the following formula (3):

[0071] HA = tan -1 (Vx / V y (3)

[0072] Here, HA represents the first heading angle information.

[0073] Step 240: If the first heading angle information meets the first preset condition, the first speed information is determined as the initial speed after initializing the trajectory speed of the target vehicle.

[0074] The first preset condition can be a condition that the preset first heading angle information must satisfy, and the first preset condition can be that the first heading angle information satisfies any one of the following:

[0075] The absolute value of the first heading angle information is less than the preset heading angle threshold, i.e., |HA| < ε;

[0076] The absolute value of the difference between 180° and the first heading angle information is less than the preset heading angle threshold, that is, |180°-HA|<ε;

[0077] The absolute value of the difference between 90° and the third heading angle information is less than the preset heading angle threshold, i.e., |90°-|HA|<ε.

[0078] The aforementioned ε is the preset heading angle threshold. This preset heading angle threshold can be a pre-set heading angle threshold. Here, the preset heading angle threshold can be the heading angle between the first vehicle and the target vehicle obtained through experiments when the target vehicle is traveling straight, the target vehicle is traveling towards the first vehicle, or the target vehicle is crossing the first vehicle.

[0079] The aforementioned third heading angle information can be the absolute value of the first heading angle information.

[0080] If the first heading angle satisfies any of the above conditions, the first speed information is determined as the initial speed after initializing the trajectory speed of the target vehicle.

[0081] The initial speed can be the speed after initializing the trajectory speed of the target vehicle.

[0082] In some embodiments of this application, since the velocity profile calculation of the N feature point clouds in step 220 above assumes that the target vehicle's motion only involves translation and that the target vehicle does not rotate around the first vehicle, the calculated first heading angle information would be inaccurate if the target vehicle's motion involves rotation. Therefore, it is necessary to judge the first heading angle information to determine whether it meets the first preset condition. If it does, the first velocity information is determined as the initial velocity after initializing the trajectory velocity of the target vehicle. If it does not meet the condition, step 250 is executed.

[0083] Step 250: If the first heading angle information does not meet the first preset condition, fit the first position information of the target vehicle in each frame of the M-frame image to obtain the second heading angle information of the target vehicle in the first shooting period, and initialize the trajectory speed of the target vehicle based on the second heading angle information to obtain the initialized speed.

[0084] The first location information can be the location information of the target vehicle when each of the M image frames was captured. Here, the M image frames can include the first image frame, and M is a positive integer.

[0085] The first shooting period is the shooting period corresponding to the shooting of M image frames. That is, during the first shooting period, the millimeter-wave radar in the first vehicle shoots M image frames of the target vehicle, which include the first image frame.

[0086] It should be noted that each of the M image frames includes N feature point clouds of the target vehicle. The number and location of the feature point clouds are the same in each image frame. In other words, these M image frames are only image frames of the target vehicle taken at different times within the first shooting period, and the feature point clouds selected in each image frame are consistent to facilitate subsequent speed initialization.

[0087] The second heading angle information can be obtained by fitting the first position information of the target vehicle at each frame of the M-frame image to the heading angle information of the target vehicle during the first shooting period.

[0088] In some embodiments of this application, since the first heading angle information calculated in step 220 above does not meet the first preset condition, it is necessary to recalculate the accurate heading angle information. In this way, the first position information of the target vehicle at each frame of the M-frame image can be fitted, and then the accurate heading angle information of the target vehicle during the first shooting period can be obtained.

[0089] In some embodiments of this application, in order to accurately obtain the second heading angle information, before fitting the first position information of the target vehicle at each frame in the M-frame image frames to obtain the second heading angle information of the target vehicle during the first shooting period, the method described above may further include:

[0090] Record the first position information of the target vehicle and the radial velocity information of each feature point cloud in the first image frame relative to the first vehicle when the first image frame is captured, and add 1 to the frame number of the recorded image frames to obtain the target frame number M;

[0091] The process of fitting the first position information of the target vehicle at the time of capturing each of the M image frames to obtain the second heading angle information of the target vehicle during the first shooting period includes:

[0092] If the target number of frames M is greater than the preset frame number threshold, the first position information of the target vehicle at each frame in the M frames is fitted to obtain the second heading angle information of the target vehicle in the first shooting period.

[0093] The target frame number can be the number of recorded image frames.

[0094] In some embodiments of this application, if it is determined that the first heading angle information does not meet the first preset condition, the number of image frames is recorded once, that is, the number of recorded image frames is incremented by 1.

[0095] The preset frame rate threshold can be a pre-set threshold for the target frame rate. This threshold can be set by the user according to their needs, and is not limited in this embodiment.

[0096] If the number of target frames recorded is greater than a preset frame number threshold, the first position information of the target vehicle at each frame in the M-frame image is fitted to obtain the second heading angle information of the target vehicle during the first shooting period.

[0097] If the number of recorded target frames is less than or equal to the preset frame number threshold, then the second image frame of the target vehicle is obtained. Here, the second image frame is the next image frame after the first image frame. That is, when the number of recorded target frames is less than or equal to the preset frame number threshold, the next image frame of the target vehicle is obtained, and then the process returns to execute the steps 220-250 above.

[0098] In the embodiments of this application, by recording the first position information of the target vehicle and the radial velocity information of each feature point cloud relative to the first vehicle in the first image frame when the first image frame is captured, and adding 1 to the frame number of the recorded image frame to obtain the target frame number M, when it is determined that the target frame number M is greater than a preset frame number threshold, the first position information of the target vehicle when each image frame in the M image frames is captured can be fitted to obtain the second heading angle information of the target vehicle in the first shooting period. In this way, when the first position information reaches a certain number, the first position information is fitted again, so the fitting accuracy is higher, and the second heading angle information of the target vehicle in the first shooting period is also more accurate.

[0099] In some embodiments of this application, in order to obtain the second heading angle information more accurately, the first position information of the target vehicle at each frame of the M-frame image is fitted to obtain the second heading angle information of the target vehicle during the first shooting period, including:

[0100] Calculate the covariance matrix corresponding to the first position information of the target vehicle in each of the M image frames captured;

[0101] Calculate the eigenvalues ​​and eigenvectors of the covariance matrix, where the eigenvalues ​​include a first eigenvalue and a second eigenvalue, and the eigenvectors include a first eigenvector and a second eigenvector. The first eigenvalue corresponds to the first eigenvector, and the second eigenvalue corresponds to the second eigenvector.

[0102] The direction of movement of the target vehicle during the first shooting period is determined based on the third eigenvector corresponding to the third eigenvalue.

[0103] Based on the direction of movement of the target vehicle during the first shooting period, determine the second heading angle information of the target vehicle during the first shooting period.

[0104] The third eigenvalue is the larger of the first and second eigenvalues, and the third eigenvector is the eigenvector corresponding to the third eigenvalue among the first and second eigenvectors. That is, if the third eigenvalue is the first eigenvalue, then the third eigenvector is the first eigenvector; if the third eigenvalue is the second eigenvalue, then the third eigenvector is the second eigenvector.

[0105] In some embodiments of this application, the first eigenvalue and the first eigenvector are a pair of eigenvalues ​​and eigenvectors of the covariance matrix, and the second eigenvalue and the second eigenvector are another pair of eigenvalues ​​and eigenvectors of the covariance matrix.

[0106] In some embodiments of this application, based on the first position information of the target vehicle at the time of capturing each of the M image frames, the covariance matrix cov(x,y) shown in the following formula (4) can be calculated:

[0107]

[0108] In some embodiments of this application, after obtaining the covariance matrix shown in formula (4) above, the two eigenvalues ​​and two eigenvectors corresponding to the covariance matrix can be calculated. The eigenvector corresponding to the larger eigenvalue is parallel to the fitted straight line, such as... Figure 3 As shown, Figure 3 The straight line is obtained by fitting the first position information of the target vehicle when the M frames of images are captured. However, since the vector has direction, after obtaining the third feature vector, it is necessary to determine the motion direction of the target vehicle in the first shooting period based on the third feature vector corresponding to the third feature value. Only then can the second heading angle information of the target vehicle in the first shooting period be determined based on the motion direction of the target vehicle in the first shooting period.

[0109] In the embodiments of this application, the covariance matrix corresponding to the first position information of the target vehicle in each of the M image frames is calculated, and then the eigenvalues ​​and eigenvectors of the covariance matrix are calculated. Based on the third eigenvalue and the third eigenvector, the motion direction of the target vehicle in the first shooting period is determined. Based on the motion direction of the target vehicle in the first shooting period, the second heading angle information of the target vehicle in the first shooting period is determined. In this way, the second heading angle information of the target vehicle in the first shooting period can be accurately obtained.

[0110] In some embodiments of this application, in order to further accurately obtain the second heading angle information, determining the motion direction of the target vehicle during the first shooting period based on the third feature vector corresponding to the third feature value may specifically include:

[0111] Based on the first position information of the target vehicle when the i-th image frame is captured and the first position information of the target vehicle when the (i+1)-th image frame is captured, M-1 position vector formulas are obtained, where 1≤i≤M-1;

[0112] Calculate the first angle between each of the M-1 position vectors and the third eigenvector to obtain the set of first angles;

[0113] Calculate the second angle between each of the M-1 position vectors and the fourth eigenvector to obtain the set of second angles;

[0114] Obtain the first number of values ​​greater than 0 in the first set of included angles and the second number of values ​​greater than 0 in the second set of included angles;

[0115] The direction of the feature vector corresponding to the larger value between the first and second quantities is determined as the direction of movement of the target vehicle during the first shooting period.

[0116] The first included angle can be the angle between each of the M-1 position vector expressions and the third eigenvector. The set of first included angles can be the set of the first included angles between the M-1 position vector expressions and the third eigenvector.

[0117] The second included angle can be the angle between each of the M-1 position vector expressions and the fourth eigenvector. The set of first included angles can be the set of the second included angles between the M-1 position vector expressions and the fourth eigenvector.

[0118] The fourth eigenvector is the inverse of the third eigenvector.

[0119] The first quantity can be the number of first included angles greater than 0 in the first included angle set. The second quantity can be the number of second included angles greater than 0 in the second included angle set.

[0120] In some embodiments of this application, M-1 position vectors can be obtained based on the first position information of the target vehicle when the i-th image frame is captured and the first position information of the target vehicle when the (i+1)-th image frame is captured. Specifically, M-1 position vectors can be obtained based on the following formula (5):

[0121] n i =p i+1 -p i (5)

[0122] Where, p i To obtain the first position information of the target vehicle when the i-th image frame is captured, p i+1 This refers to the first position information of the target vehicle when the (i+1)th image frame is captured.

[0123] According to the above formula (5), a total of M-1 position vector formulas can be obtained.

[0124] Then determine the M-1 vectors n i Which of the first angles formed by the first eigenvector u and the third eigenvector u are acute angles, and which of the M-1 vectors n are determined? i Which of the second included angles formed by the inverse vector -u of the third eigenvector u and the second included vector n are acute angles? Specifically, M-1 vectors n can be calculated according to the following formula (6). i The inner product with the third eigenvector u, and the M-1 vectors n i The inner product between the first angle set b and the inverse vector -u of the third eigenvector u is considered. i ={b1,b2,…b M-1} and the set of second included angles Which numbers in the set are greater than 0, and what is the first included angle set b? i The set of the second included angle b i A number greater than 0 indicates that the angle between two vectors is less than 90°.

[0125]

[0126] Obtain the first included angle set b respectively i The first number of the first included angles greater than 0, and the set of the second included angles b. i The second number of the second included angles greater than 0 is used as the direction of the vector corresponding to the larger of the first and second included angles, which is taken as the direction of movement of the target vehicle during the first shooting period. For example, taking M=20 as an example, the first included angle set b i The first number of the first included angles greater than 0 is 15, and the set of the second included angles is b. i If the second number of the second included angles greater than 0 is 13, then the first included angle set b will be calculated. iThe direction of the vector u corresponding to the time is determined as the direction of movement of the target vehicle during the first shooting period.

[0127] In the embodiments of this application, based on the first position information of the target vehicle when the i-th image frame is captured and the first position information of the target vehicle when the (i+1)-th image frame is captured, M-1 position vectors are obtained. The first angle between each of the M-1 position vectors and the third feature vector is calculated to obtain a first angle set. The second angle between each of the M-1 position vectors and the fourth feature vector is calculated to obtain a second angle set. The first number of values ​​greater than 0 in the first angle set and the second number of values ​​greater than 0 in the second angle set are obtained. The direction of the feature vector corresponding to the larger value among the first and second numbers is determined as the movement direction of the target vehicle during the first shooting period. In this way, the movement direction of the target vehicle during the first shooting period can be accurately determined.

[0128] In some embodiments of this application, after obtaining the direction of motion of the target vehicle during the first shooting period, the direction of motion is projected in the horizontal and vertical directions respectively to obtain the second heading angle information of the target vehicle during the first shooting period.

[0129] In some embodiments of this application, after obtaining the second heading angle information, the trajectory speed of the target vehicle can be initialized based on the second heading angle information to obtain the initial speed.

[0130] In some embodiments of this application, in order to accurately obtain the initial velocity, the initialization of the target vehicle's trajectory velocity based on the second heading angle information to obtain the initial velocity includes:

[0131] Based on the second heading angle information, calculate the third speed information of the target vehicle relative to the first vehicle when each image frame in the M-frame image frame is captured;

[0132] The fourth speed information is obtained by averaging the values ​​of the third speed information.

[0133] Based on the fourth velocity information, the trajectory velocity of the target vehicle is initialized to obtain the initial velocity.

[0134] The third speed information can be the speed information of the target vehicle relative to the first vehicle at each frame of the M-frame image captured, calculated based on the second heading angle information.

[0135] The fourth velocity information can be obtained by averaging the values ​​of each third velocity information.

[0136] In some embodiments of this application, the third speed information of the target vehicle at the time of capturing each image frame in the M-frame image frames can be calculated based on the second heading angle information according to the following formula (7):

[0137]

[0138] Among them, V i To obtain the third speed information of the target vehicle when the i-th image frame is captured, V di V is the radial velocity of the target vehicle when the i-th image frame is captured. di Let x be the average radial velocity information of N feature point clouds in the i-th image frame, (x) i ,y i ) represents the first position information of the target vehicle when the i-th image frame is captured, and HA' represents the second heading angle information.

[0139] After obtaining the third speed information of the target vehicle relative to the first vehicle at each frame of the M-frame image, the fourth speed information can be obtained based on the average value of each third speed information. Then, based on the fourth speed information, the trajectory speed of the target vehicle is initialized to obtain the initial speed.

[0140] In some embodiments of this application, the trajectory velocity of the target vehicle is initialized based on the fourth velocity information according to the following formula (8) to obtain the initial velocity:

[0141]

[0142] Among them, (V) x V y V z ) represents the initial velocity, MV represents the fourth velocity information, and α represents the second heading angle information.

[0143] In the embodiments of this application, based on the second heading angle information, the third speed information of the target vehicle relative to the first vehicle is calculated when each image frame in the M-frame image frames is captured. Based on the average value of each third speed information, the fourth speed information is obtained. Based on the fourth speed information, the trajectory speed of the target vehicle is initialized, and the initial speed can be accurately obtained.

[0144] In some embodiments of this application, to more clearly understand the technical solutions of the embodiments of this application, the embodiments of this application use specific scenarios as examples to illustrate the speed initialization method provided by the embodiments of this application, such as... Figure 4 As shown, Figure 4 The flowchart is for a speed initialization method provided in an embodiment of this application, which includes steps 401-412.

[0145] Step 401: Initialize the number of image frames Q = 0.

[0146] Step 402: Obtain the first image frame of the target vehicle.

[0147] Step 403: Based on the motion parameters corresponding to the N feature point clouds in the first image frame, calculate the velocity profile of the N feature point clouds to obtain the first velocity information of the target vehicle.

[0148] Step 404: Based on the first velocity information, calculate the first heading angle information HA of the target vehicle when the first image frame is captured.

[0149] Steps 402-404 are the same as steps 210-230 above, and will not be repeated here.

[0150] Step 405: Determine whether the first heading angle information HA meets the first preset condition. If it does, proceed to step 406; if it does not, proceed to step 407.

[0151] Step 406: Initialize the trajectory speed of the target vehicle to obtain the initial speed.

[0152] Specifically, the first speed information can be determined as the initial speed after initializing the trajectory speed of the target vehicle.

[0153] Step 407 records the first position information of the target vehicle and the radial velocity information of each feature point cloud relative to the first vehicle when the first image frame is captured.

[0154] Step 408: Increment the number of recorded image frames Q by 1 to obtain the target number of frames Q = M.

[0155] Step 409: Determine whether M is greater than the preset frame number threshold. If yes, proceed to step 410. If no, return to step 402 to obtain the next image frame of the target vehicle.

[0156] Step 410: Fit the first position information of the target vehicle at each frame of the M-frame image to obtain the second heading angle information of the target vehicle during the first shooting period.

[0157] Step 411: Based on the second heading angle information, calculate the third speed information of the target vehicle relative to the first vehicle when each image frame in the M-frame image frame is captured.

[0158] Step 412: Obtain the fourth speed information based on the average value of each third speed information.

[0159] After obtaining the fourth speed information, the trajectory speed of the target vehicle can be initialized based on the fourth speed information to obtain the initial speed. For details on how to initialize the trajectory speed of the target vehicle based on the fourth speed information to obtain the initial speed, please refer to the calculation formula (8) in the above embodiment, which will not be repeated here.

[0160] In the embodiments of this application, the trajectory velocity of the target vehicle is initialized by combining velocity profile analysis with multi-frame image frame fitting, thereby improving the accuracy of the trajectory velocity initialization of the target vehicle.

[0161] It should be noted that the speed initialization method provided in this application embodiment can be executed by a speed initialization device or a control module in the speed initialization device for executing the speed initialization method.

[0162] Based on the same inventive concept as the speed initialization method described above, this application also provides a speed initialization device. The following is in conjunction with… Figure 5 The speed initialization device provided in the embodiments of this application will be described in detail.

[0163] Figure 5 This is a schematic diagram of a speed initialization device according to an exemplary embodiment.

[0164] like Figure 5 As shown, the speed initialization device 500 may include:

[0165] The acquisition module 510 is used to acquire a first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, where N is a positive integer;

[0166] The first determining module 520 is used to calculate the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds, so as to obtain the first velocity information of the target vehicle.

[0167] The first calculation module 530 is used to calculate the first heading angle information of the target vehicle when the first image frame is captured, based on the first speed information.

[0168] The second determining module 540 is used to determine the first speed information as the initial speed after initializing the trajectory speed of the target vehicle when it is determined that the first heading angle information meets the first preset condition.

[0169] The third determining module 550 is used to, when determining that the first heading angle information does not meet the first preset condition, fit the first position information of the target vehicle when each image frame in the M-frame image frames is captured, to obtain the second heading angle information of the target vehicle during the first shooting period, and initialize the trajectory speed of the target vehicle based on the second heading angle information to obtain the initialized speed, wherein the M-frame image frames include the first image frames, the first shooting period is the shooting period corresponding to the M-frame image frames, and M is a positive integer.

[0170] In the embodiments of this application, the first velocity information of the target vehicle is obtained by calculating the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds in the first image frame of the target vehicle. Then, based on the first velocity information, the first heading angle information of the target vehicle when the first image frame was captured is calculated. If the first heading angle information meets the first preset condition, the first velocity information is determined as the initial velocity after initializing the trajectory velocity of the target vehicle. If the first heading angle information does not meet the first preset condition, the first position information of the target vehicle when each image frame in the M image frames is captured is fitted to obtain the second heading angle information of the target vehicle in the first capturing period. Based on the second heading angle information, the trajectory velocity of the target vehicle is initialized to obtain the initial velocity. In this way, instead of mapping the radial velocity of the target vehicle in the horizontal and vertical directions, the velocity profile analysis of the N feature point clouds of the first image frame of the target vehicle is performed, or if the first heading angle information corresponding to multiple image frames is determined to be inaccurate, the trajectory velocity is initialized by fitting the first position information of the target vehicle when multiple image frames are captured. This improves the accuracy of trajectory velocity initialization.

[0171] In some embodiments of this application, the motion parameters of each feature point cloud include: the azimuth information of the feature point cloud relative to the first vehicle and the radial velocity information of the feature point cloud relative to the first vehicle, wherein the first image frame is acquired by an image acquisition device on the first vehicle.

[0172] The first determining module 520 can be specifically used for:

[0173] For each of the feature point clouds, the second velocity information of the feature point cloud is calculated based on the azimuth information and the radial velocity information;

[0174] A loss function is constructed based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud;

[0175] The optimal solution is obtained for the loss function to obtain the first speed information of the target vehicle.

[0176] In some embodiments of this application, the third determining module 550 may include:

[0177] The first determining unit is used to calculate the third speed information of the target vehicle relative to the first vehicle when each image frame in the M-frame image frames is captured, based on the second heading angle information.

[0178] The second determining unit is used to obtain the fourth speed information based on the average value of each of the third speed information;

[0179] The third determining unit is used to initialize the trajectory speed of the target vehicle based on the fourth speed information to obtain the initial speed.

[0180] In some embodiments of this application, the third determining unit is specifically used for:

[0181] Based on the fourth speed information, the trajectory speed of the target vehicle is initialized according to the following formula to obtain the initial speed:

[0182] V x =MV×sinα

[0183] V y =MV×cosα

[0184] V z =0

[0185] Among them, (V) x V y V z ) represents the initial velocity, MV represents the fourth velocity information, and α represents the second heading angle information.

[0186] In some embodiments of this application, the first determining unit is specifically used for:

[0187] Based on the second heading angle information, the third speed information of the target vehicle is calculated for each image frame in the M-frame image series, according to the following formula:

[0188] V i =V di / cos(α-β)

[0189] α=HA'

[0190] β = tan -1 (x i ,y i )

[0191] Among them, V i V is the third speed information of the target vehicle when the i-th image frame is captured. di V is the radial velocity of the target vehicle when the i-th image frame is captured. di The radial velocity information of the N feature point clouds in the i-th image frame is the average value, (x i ,y i ) represents the first position information of the target vehicle when the i-th image frame is captured, and HA' represents the second heading angle information.

[0192] In some embodiments of this application, the apparatus described above may further include:

[0193] The recording module is used to record the first position information of the target vehicle and the radial velocity information of each feature point cloud in the first image frame relative to the first vehicle when the first image frame is captured, and to add 1 to the frame number of the recorded image frame to obtain the target frame number M;

[0194] The third determining module 550 is specifically used for:

[0195] If the target number of frames M is determined to be greater than a preset frame number threshold, the first position information of the target vehicle when each of the M image frames is captured is fitted to obtain the second heading angle information of the target vehicle during the first shooting period.

[0196] In some embodiments of this application, the third determining module 550 may further include:

[0197] The fourth determining unit is used to calculate the covariance matrix corresponding to the first position information of the target vehicle when each of the M image frames is captured;

[0198] The fifth determining unit is used to calculate the eigenvalues ​​and eigenvectors of the covariance matrix, wherein the eigenvalues ​​include a first eigenvalue and a second eigenvalue, and the eigenvectors include a first eigenvector and a second eigenvector, wherein the first eigenvalue corresponds to the first eigenvector, and the second eigenvalue corresponds to the second eigenvector;

[0199] The sixth determining unit is used to determine the movement direction of the target vehicle during the first shooting period based on the third feature vector corresponding to the third feature value, wherein the third feature value is the larger value between the first feature value and the second feature value, and the third feature vector is the feature vector between the first feature vector and the second feature vector that corresponds to the third feature value;

[0200] The seventh determining unit is used to determine the second heading angle information of the target vehicle during the first shooting period based on the movement direction of the target vehicle during the first shooting period.

[0201] In some embodiments of this application, the sixth determining unit is specifically used for:

[0202] Based on the first position information of the target vehicle when the i-th image frame is captured and the first position information of the target vehicle when the (i+1)-th image frame is captured, M-1 position vector formulas are obtained, where 1≤i≤M-1;

[0203] Calculate the first angle between each of the M-1 position vectors and the third feature vector to obtain the first angle set;

[0204] Calculate the second angle between each of the M-1 position vectors and the fourth feature vector to obtain the set of second angles, wherein the fourth feature vector is the inverse vector of the third feature vector;

[0205] Obtain the first number of angles greater than 0 in the first set of included angles and the second number of angles greater than 0 in the second set of included angles, respectively;

[0206] The direction of the feature vector corresponding to the larger value between the first quantity and the second quantity is determined as the movement direction of the target vehicle during the first shooting period.

[0207] In some embodiments of this application, the acquisition module 510 may also be used for:

[0208] If the target frame number M is determined to be less than or equal to the preset frame number threshold, a second image frame of the target vehicle is obtained, wherein the second image frame is the next image frame after the first image frame;

[0209] The aforementioned apparatus may further include:

[0210] The execution module is returned to update the first image frame to the second image frame, and then returns to execute the steps of calculating the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds to obtain the first velocity information of the target vehicle; calculating the first heading angle information of the target vehicle when the first image frame was captured based on the first velocity information; determining the first velocity information as the initial velocity after initializing the trajectory velocity of the target vehicle if the first heading angle information does not meet the first preset condition; fitting the first position information of the target vehicle when each image frame in the M image frames was captured to obtain the second heading angle information of the target vehicle during the first shooting period, and initializing the trajectory velocity of the target vehicle based on the second heading angle information to obtain the initial velocity.

[0211] In some embodiments of this application, the motion parameters of each feature point cloud further include: pitch angle information of the feature point cloud relative to the first vehicle.

[0212] The first determining module 520 can also be specifically used for:

[0213] For each of the feature point clouds, the second velocity information of the feature point cloud is calculated based on the azimuth information, the pitch information, and the radial velocity information.

[0214] The speed initialization device provided in this application embodiment can be used to execute the speed initialization methods provided in the above method embodiments. Its implementation principle and technical effect are similar, and for the sake of brevity, it will not be described in detail here.

[0215] Based on the same inventive concept, embodiments of this application also provide an electronic device.

[0216] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. For example... Figure 6 As shown, the electronic device may include a processor 601 and a memory 602 storing computer programs or instructions.

[0217] Specifically, the processor 601 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of the present invention.

[0218] Memory 602 may include mass storage for data or instructions. For example, and not limitingly, memory 602 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 602 may include removable or non-removable (or fixed) media. Where appropriate, memory 602 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 602 is non-volatile solid-state memory. Memory may include read-only memory (ROM), random-access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, a memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described in the speed initialization method provided in the above embodiments.

[0219] The processor 601 implements any of the speed initialization methods described in the above embodiments by reading and executing computer program instructions stored in the memory 602.

[0220] In one example, the electronic device may also include a communication interface 603 and a bus 610. For example, Figure 6 As shown, the processor 601, memory 602, and communication interface 603 are connected through bus 610 and complete communication with each other.

[0221] The communication interface 603 is mainly used to realize communication between various modules, devices, units and / or devices in the embodiments of the present invention.

[0222] Bus 610 includes hardware, software, or both, that couples components of an electronic device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 610 may include one or more buses. Although specific buses are described and illustrated in embodiments of the invention, the invention contemplates any suitable bus or interconnect.

[0223] The electronic device can execute the speed initialization method in the embodiments of the present invention, thereby achieving... Figure 2 The speed initialization method is described.

[0224] Furthermore, in conjunction with the speed initialization methods described in the above embodiments, this invention can be implemented using a readable storage medium. This readable storage medium stores program instructions, which, when executed by a processor, implement any of the speed initialization methods described in the above embodiments.

[0225] In addition, in conjunction with the speed initialization methods in the above embodiments, the present invention can provide a computer program product in which the instructions are executed by the processor of an electronic device, enabling the electronic device to execute any of the speed initialization methods in the above embodiments.

[0226] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0227] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the required tasks. The programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0228] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0229] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, 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, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0230] The above description is merely a specific embodiment of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A speed initialization method, characterized by, The method includes: Acquire a first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, N≥2, and N is a positive integer; Based on the motion parameters corresponding to N feature point clouds, velocity profiles are calculated for the N feature point clouds to obtain the first velocity information of the target vehicle. Based on the first speed information, calculate the first heading angle information of the target vehicle when the first image frame is captured; If the first heading angle information satisfies the first preset condition, the first speed information is determined as the initial speed after initializing the trajectory speed of the target vehicle. The first preset condition includes any one of the following: the absolute value of the first heading angle information is less than a preset heading angle threshold; the absolute value of the difference between 180° and the first heading angle information is less than the preset heading angle threshold; the absolute value of the difference between 90° and the third heading angle information is less than the preset heading angle threshold, wherein the third heading angle information is the absolute value of the first heading angle information. If it is determined that the first heading angle information does not meet the first preset condition, the second heading angle information of the target vehicle during the first shooting period is obtained based on the first position information of the target vehicle when each image frame in the M-frame image frames is captured, and the trajectory speed of the target vehicle is initialized based on the second heading angle information to obtain the initialized speed. Here, the M-frame image frames include the first image frames, the first shooting period is the shooting period corresponding to the M-frame image frames, M≥2, and M is a positive integer. The motion parameters of each feature point cloud include: the azimuth information of the feature point cloud relative to the first vehicle and the radial velocity information of the feature point cloud relative to the first vehicle, wherein the first image frame is acquired by an image acquisition device on the first vehicle. The step of calculating the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds to obtain the first velocity information of the target vehicle includes: For each of the feature point clouds, the second velocity information of the feature point cloud is calculated based on the azimuth information and the radial velocity information; A loss function is constructed based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud; The optimal solution is obtained for the loss function to obtain the first speed information of the target vehicle.

2. The method of claim 1, wherein, The initialization of the trajectory velocity of the target vehicle based on the second heading angle information to obtain the initial velocity includes: Based on the second heading angle information, calculate the third speed information of the target vehicle relative to the first vehicle when each image frame in the M-frame image frames is captured; The fourth speed information is obtained based on the average value of each of the third speed information. Based on the fourth speed information, the trajectory speed of the target vehicle is initialized to obtain the initial speed.

3. The method according to claim 2, characterized in that, The initialization of the trajectory speed of the target vehicle based on the fourth speed information to obtain the initial speed includes: Based on the fourth speed information, the trajectory speed of the target vehicle is initialized according to the following formula to obtain the initial speed: in, For the initial speed, This refers to the fourth speed information. This is the second heading angle information.

4. The method according to claim 2, characterized in that, The step of calculating the third speed information of the target vehicle for each image frame in the M-frame image series based on the second heading angle information includes: Based on the second heading angle information, the third speed information of the target vehicle is calculated for each image frame in the M-frame image frames according to the following formula: in, The third speed information of the target vehicle at the time the i-th image frame was captured. The radial velocity of the target vehicle is given when the i-th image frame is captured. The radial velocity information of the N feature point clouds in the i-th image frame is the average value. This refers to the first position information of the target vehicle when the i-th image frame is captured. This is the second heading angle information.

5. The method according to claim 1, characterized in that, Before obtaining the second heading angle information of the target vehicle during the first shooting period based on the first position information of the target vehicle at each frame of the M-frame image, the method further includes: Record the first position information of the target vehicle and the radial velocity information of each feature point cloud in the first image frame relative to the first vehicle when the first image frame is captured, and add 1 to the frame number of the recorded image frame to obtain the target frame number M; The step of obtaining the second heading angle information of the target vehicle during the first shooting period based on the first position information of the target vehicle at each frame of the M-frame image capture includes: If the target number of frames M is determined to be greater than a preset frame number threshold, the second heading angle information of the target vehicle during the first shooting period is obtained based on the first position information of the target vehicle when each of the M image frames is captured.

6. The method according to claim 5, characterized in that, The method further includes: If the target frame number M is determined to be less than or equal to the preset frame number threshold, a second image frame of the target vehicle is obtained, wherein the second image frame is the next image frame after the first image frame; The process involves updating the first image frame to the second image frame, returning to the step of performing velocity profile calculations on the N feature point clouds based on motion parameters corresponding to the N feature point clouds to obtain the first velocity information of the target vehicle; calculating the first heading angle information of the target vehicle when the first image frame was captured based on the first velocity information; determining the first velocity information as the initial velocity after initializing the trajectory velocity of the target vehicle if the first heading angle information does not meet the first preset condition; and obtaining the second heading angle information of the target vehicle during the first shooting period based on the first position information of the target vehicle when each image frame in the M image frames was captured, and initializing the trajectory velocity of the target vehicle based on the second heading angle information to obtain the initial velocity.

7. The method according to claim 5, characterized in that, The step of obtaining the second heading angle information of the target vehicle during the first shooting period based on the first position information of the target vehicle at each frame of the M-frame image capture includes: Calculate the covariance matrix corresponding to the first position information of the target vehicle when each of the M image frames is captured; Calculate the eigenvalues ​​and eigenvectors of the covariance matrix, wherein the eigenvalues ​​include a first eigenvalue and a second eigenvalue, and the eigenvectors include a first eigenvector and a second eigenvector, wherein the first eigenvalue corresponds to the first eigenvector, and the second eigenvalue corresponds to the second eigenvector; The direction of motion of the target vehicle during the first shooting period is determined based on the third feature vector corresponding to the third feature value, wherein the third feature value is the larger value between the first feature value and the second feature value, and the third feature vector is the feature vector between the first feature vector and the second feature vector that corresponds to the third feature value; Based on the direction of movement of the target vehicle during the first shooting period, the second heading angle information of the target vehicle during the first shooting period is determined.

8. The method according to claim 7, characterized in that, Determining the direction of motion of the target vehicle during the first shooting period based on the third feature vector corresponding to the third feature value includes: Based on the first position information of the target vehicle when the i-th image frame is captured and the first position information of the target vehicle when the (i+1)-th image frame is captured, M-1 position vector formulas are obtained, where 1≤i≤M-1; Calculate the first angle between each of the M-1 position vectors and the third feature vector to obtain the first angle set; Calculate the second angle between each of the M-1 position vectors and the fourth eigenvector to obtain the set of second angles, wherein the fourth eigenvector is the inverse of the third eigenvector; Obtain the first number of angles greater than 0 in the first set of included angles and the second number of angles greater than 0 in the second set of included angles, respectively; The direction of the feature vector corresponding to the larger value between the first quantity and the second quantity is determined as the movement direction of the target vehicle during the first shooting period.

9. A speed initialization device, characterized in that, The device includes: The acquisition module is used to acquire a first image frame of the target vehicle, wherein the first image frame includes N feature point clouds of the target vehicle, N≥2, and N is a positive integer; The first determining module is used to calculate the velocity profile of the N feature point clouds based on the motion parameters corresponding to the N feature point clouds, so as to obtain the first velocity information of the target vehicle. The first calculation module is used to calculate the first heading angle information of the target vehicle when the first image frame is captured, based on the first speed information; The second determining module is used to determine the first speed information as the initial speed after initializing the trajectory speed of the target vehicle when the first heading angle information satisfies the first preset condition. The first preset condition includes any one of the following: the absolute value of the first heading angle information is less than a preset heading angle threshold; the absolute value of the difference between 180° and the first heading angle information is less than the preset heading angle threshold; the absolute value of the difference between 90° and the third heading angle information is less than the preset heading angle threshold, wherein the third heading angle information is the absolute value of the first heading angle information. The third determining module is used to, when it is determined that the first heading angle information does not meet the first preset condition, fit the first position information of the target vehicle when each image frame in the M-frame image frames is captured, to obtain the second heading angle information of the target vehicle during the first shooting period, and initialize the trajectory speed of the target vehicle based on the second heading angle information to obtain the initialized speed, wherein the M-frame image frames include the first image frames, the first shooting period is the shooting period corresponding to the M-frame image frames, M≥2, and M is a positive integer; The motion parameters of each feature point cloud include: the azimuth information of the feature point cloud relative to the first vehicle and the radial velocity information of the feature point cloud relative to the first vehicle, wherein the first image frame is acquired by an image acquisition device on the first vehicle. The first determining module is specifically used for: For each of the feature point clouds, the second velocity information of the feature point cloud is calculated based on the azimuth information and the radial velocity information; A loss function is constructed based on the motion parameters corresponding to each feature point cloud and the second velocity information corresponding to each feature point cloud; The optimal solution is obtained for the loss function to obtain the first speed information of the target vehicle.

Images