Tracking method and device of perceptual target, storage medium and electronic device

Abstract

The embodiment of the invention provides a sensing target tracking method and device, a storage medium and an electronic device, and the method comprises the steps: continuously obtaining the vehicle positioning information of a target vehicle, and determining an attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle; updating the vehicle positioning information of the target vehicle by using an attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain updated vehicle positioning information of the target vehicle; acquiring target positioning information of each sensing target sensed by a sensing system of the target vehicle; based on the vehicle positioning information corresponding to each sensing target, converting the target positioning information of each sensing target into local positioning information under a local coordinate system to obtain local positioning information of each sensing target; and based on the local positioning information of each sensing target, performing fusion tracking on the sensing target sensed by the sensing system of the target vehicle.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of radar, and more specifically, to a method and device for tracking a perceived target, a storage medium, and an electronic device. Background Art

[0002] In the field of target fusion perception technology for intelligent connected vehicles, being able to obtain the precise motion state of the vehicle within a certain period of time is a necessary condition for the fusion perception and tracking of surrounding targets. One vehicle positioning method is to use signals such as wheel speed meter, wheel angle, IMU (Inertial Measurement Unit), etc., and use integration and other means to establish a local coordinate system with a given point as the origin, and calculate the vehicle posture of the vehicle in the local coordinate system for target perception and tracking. The vehicle posture is usually described in three-dimensional space.

[0003] However, due to the estimated cumulative error of the vehicle body posture in the local coordinate system, and the positioning does not have satellite positioning correction in order to avoid jumps, the difference between the angle between the vehicle and the horizontal plane (i.e., the orientation angle of the vehicle body) and the actual vehicle angle (i.e., the actual direction of movement) will gradually increase. The distortion of the angle between the vehicle and the horizontal plane will cause the target position of the perceived target to shift, increase the perception error, and thus reduce the accuracy of target tracking. It can be seen that the tracking method of the perceived target in the related art has the problem of poor target tracking accuracy due to the large perception error. Summary of the invention

[0004] The embodiments of the present application provide a method and device for tracking a perceived target, a storage medium, and an electronic device, so as to at least solve the problem of poor target tracking accuracy caused by large perception errors in the method for tracking a perceived target in the related art.

[0005] According to one aspect of an embodiment of the present application, a tracking method for a perceived target is provided, comprising: continuously acquiring vehicle positioning information of a target vehicle, and determining a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in a local coordinate system, the local coordinate system is a fixed coordinate system with a specified position as the origin, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction in the local coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent; the vehicle positioning information of the target vehicle is updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, and the target vehicle is obtained. Updated vehicle positioning information; obtaining target positioning information of each perceived target perceived by the perception system of the target vehicle, wherein the target positioning information of each perceived target is the positioning information of each perceived target in the body coordinate system of the target vehicle, and the target positioning information of each perceived target is used to represent the target posture of each perceived target at the perception time of each perceived target; based on the vehicle positioning information corresponding to each perceived target, converting the target positioning information of each perceived target into local positioning information in the local coordinate system to obtain the local positioning information of each perceived target, wherein the vehicle positioning information corresponding to each perceived target is the vehicle positioning information whose positioning time is consistent with the perception time of each perceived target in the vehicle positioning information updated by the target vehicle; based on the local positioning information of each perceived target, performing fusion tracking on the perceived targets perceived by the perception system of the target vehicle.

[0006] According to another aspect of an embodiment of the present application, a tracking device for sensing a target is provided, comprising: a first execution unit, configured to continuously acquire vehicle positioning information of a target vehicle, and determine a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in a local coordinate system, and the local coordinate system is a fixed coordinate system with a specified position as the origin, and the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction in the local coordinate system while keeping the horizontal mileage of the target vehicle at the positioning time of the vehicle positioning information consistent with the spatial mileage; an updating unit, configured to update the vehicle positioning information of the target vehicle using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, and obtain an updated Vehicle positioning information; an acquisition unit, used to acquire target positioning information of each perception target perceived by the perception system of the target vehicle, wherein the target positioning information of each perception target is the positioning information of each perception target in the body coordinate system of the target vehicle, and the target positioning information of each perception target is used to represent the target posture of each perception target at the perception time of each perception target; a conversion unit, used to convert the target positioning information of each perception target into local positioning information in the local coordinate system based on the vehicle positioning information corresponding to each perception target, so as to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information of the target vehicle updated, and the positioning time is consistent with the perception time of each perception target; a second execution unit, used to perform fusion tracking on the perception target perceived by the perception system of the target vehicle based on the local positioning information of each perception target.

[0007] In an exemplary embodiment, the first execution unit includes: an execution module, which is used to continuously obtain vehicle positioning information of the target vehicle, and use the vehicle positioning information of the target vehicle obtained each time as the current vehicle positioning information to perform the following determination operations to obtain a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle obtained each time: obtaining an initial transformation matrix corresponding to the current vehicle positioning information, wherein, when the current vehicle positioning information is the vehicle positioning information of the target vehicle obtained for the first time, the initial transformation matrix corresponding to the current vehicle positioning information is a unit matrix, and when the current vehicle positioning information is the vehicle positioning information of the target vehicle obtained for the Nth time, the initial transformation matrix corresponding to the current vehicle positioning information is a unit matrix. The initial transformation matrix corresponding to the current vehicle positioning information is the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle obtained for the N-1th time, where N is a positive integer greater than or equal to 2; the current vehicle positioning information is updated using the initial transformation matrix corresponding to the current vehicle positioning information to obtain reference vehicle positioning information; the posture transformation matrix corresponding to the reference vehicle positioning information is determined, wherein the posture transformation matrix corresponding to the reference vehicle positioning information is a transformation matrix that uses the vehicle position indicated by the reference vehicle positioning information as an anchor point and is used to rotate the vehicle posture indicated by the reference vehicle positioning information to the horizontal direction; based on the posture transformation matrix corresponding to the reference vehicle positioning information, the posture transformation matrix corresponding to the current vehicle positioning information is determined.

[0008] In an exemplary embodiment, the execution module includes: a first determination submodule, used to determine the rotation matrix corresponding to the reference vehicle positioning information, wherein the rotation matrix corresponding to the reference vehicle positioning information is a rotation matrix used to rotate the vehicle posture indicated by the reference vehicle positioning information to the horizontal direction; a second determination submodule, used to determine the posture transformation matrix corresponding to the reference vehicle positioning information based on the rotation matrix corresponding to the reference vehicle positioning information.

[0009] In an exemplary embodiment, the first determination submodule includes: a first determination subunit, used to determine the horizontal plane normal vector of the vehicle posture of the target vehicle in the local coordinate system based on the rotation component corresponding to the reference vehicle positioning information, and obtain the target normal vector; a second determination subunit, used to determine the rotation axis direction vector of the target normal vector, and the rotation angle corresponding to the rotation axis direction vector of the target normal vector; a third determination subunit, used to determine the rotation matrix corresponding to the reference vehicle positioning information based on the rotation axis direction vector of the target normal vector and the rotation angle corresponding to the rotation axis direction vector of the target normal vector.

[0010] In an exemplary embodiment, the second determination submodule includes: a fourth determination subunit, which is used to determine the product of the sum of the unit matrix and the vehicle coordinates in the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information, and the difference between the unit matrix and the vehicle coordinates in the reference vehicle positioning information as a posture transformation matrix corresponding to the reference vehicle positioning information.

[0011] In an exemplary embodiment, the second determination module includes: a third determination submodule, which is used to determine the product of the initial transformation matrix corresponding to the current vehicle positioning information and the posture transformation matrix corresponding to the reference vehicle positioning information as the posture transformation matrix corresponding to the current vehicle positioning information.

[0012] In an exemplary embodiment, the conversion unit includes: a first conversion module, used to convert the vehicle positioning information corresponding to each perception target into two-dimensional positioning information in the local coordinate system, so as to obtain two-dimensional vehicle positioning information corresponding to each perception target; and a second conversion module, used to use the two-dimensional vehicle positioning information corresponding to each perception target to convert the target positioning information of each perception target into local positioning information in the local coordinate system, so as to obtain local positioning information of each perception target.

[0013] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when run.

[0014] According to another aspect of the embodiments of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0015] According to another aspect of the embodiments of the present application, a computer program product or a computer program is also provided, the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps in any one of the above method embodiments.

[0016] Through the embodiment of the present application, the vehicle positioning information of the vehicle in the local coordinate system is converted so that the vehicle posture is rotated to the horizontal direction in the local coordinate system while keeping the horizontal mileage of the vehicle consistent with the spatial mileage. The vehicle positioning information of the target vehicle is continuously acquired, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is determined. The vehicle positioning information of the target vehicle is updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain the updated vehicle positioning information of the target vehicle. Here, the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in the local coordinate system, and the local coordinate system is based on the specified position. The fixed coordinate system of the target vehicle, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction of the local coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent. The vehicle positioning information of the target vehicle is updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle. The vehicle positioning information collected each time can be converted to rotate the vehicle posture to the horizontal direction, thereby reducing the self-correction caused by the cumulative error. The angle between the vehicle and the horizontal plane is distorted; the target positioning information of each perceived target perceived by the perception system of the target vehicle is obtained, wherein the target positioning information of each perceived target is the positioning information of each perceived target in the body coordinate system of the target vehicle, and the target positioning information of each perceived target is used to represent the target position of each perceived target at the perception time of each perceived target; based on the vehicle positioning information corresponding to each perceived target, the target positioning information of each perceived target is converted into the local positioning information in the local coordinate system to obtain the local positioning information of each perceived target, wherein the vehicle positioning information corresponding to each perceived target is the vehicle positioning information updated by the target vehicle , the vehicle positioning information whose positioning time is consistent with the perception time of each perceived target. Here, since the target positioning information of the perceived target is converted into the local coordinate system using the vehicle positioning information updated by the target vehicle, the offset of the target position of the perceived target in the local coordinate system caused by the distortion of the angle between the vehicle and the horizontal plane can be reduced, and then based on the local positioning information of each perceived target, the perceived target perceived by the perception system of the target vehicle is fused and tracked, which can ensure the accuracy of fusion tracking and achieve the technical effect of improving the accuracy of target tracking, thereby solving the problem of poor target tracking accuracy caused by large perception error in the tracking method of the perceived target in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of an application scenario of a method for tracking a perceived target according to an embodiment of the present application;

[0018] Figure 2 is a flow chart of an optional method for tracking a perceived target according to an embodiment of the present application;

[0019] Figure 3 is a schematic diagram of an optional method for tracking a perceived target according to an embodiment of the present application;

[0020] Figure 4 is a schematic diagram of another optional method for tracking a perceived target according to an embodiment of the present application;

[0021] Figure 5 is a schematic diagram of another optional method for tracking a perceived target according to an embodiment of the present application;

[0022] Figure 6 is a structural block diagram of an optional tracking device for sensing a target according to an embodiment of the present application;

[0023] Figure 7 It is a block diagram of a computer system structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0025] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0026] The method embodiments provided in the embodiments of the present application can be executed in a vehicle-mounted terminal or a similar computing device. Taking running on a vehicle-mounted terminal as an example, Figure 1 FIG. 1 is a hardware structure diagram of an optional vehicle-mounted terminal according to this embodiment. Figure 1 As shown, the vehicle terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the above-mentioned vehicle terminal may also include a transmission device 106 for communication functions and an input and output device 108. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the vehicle-mounted terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0027] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the hard disk failure prediction method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the vehicle-mounted terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0028] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by the communication provider of the vehicle terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be an RF (Radio Frequency) module, which is used to communicate with the Internet wirelessly.

[0029] According to one aspect of an embodiment of the present application, a method for tracking a perceived target is provided. The method for tracking a perceived target in this embodiment can be executed by a vehicle-mounted terminal, or by a server, or by both the vehicle-mounted terminal and the server. Taking the vehicle-mounted terminal as an example, Figure 2 is a flow chart of an optional method for tracking a sensing target according to an embodiment of the present application, such as Figure 2 As shown, the process includes the following steps:

[0030] Step S202, continuously acquiring vehicle positioning information of the target vehicle, and determining a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle;

[0031] Step S204, updating the vehicle positioning information of the target vehicle using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain updated vehicle positioning information of the target vehicle;

[0032] Step S206, obtaining target positioning information of each perceived target perceived by the perception system of the target vehicle;

[0033] Step S208, based on the vehicle positioning information corresponding to each perception target, converting the target positioning information of each perception target into local positioning information in a local coordinate system, thereby obtaining the local positioning information of each perception target;

[0034] Step S210, based on the local positioning information of each perceived target, the perceived targets perceived by the perception system of the target vehicle are fused and tracked.

[0035] The tracking method of the perceived target in this embodiment can be applied to the field of target fusion perception technology, and applied to intelligent networked vehicles or similar vehicles to obtain the precise motion state (vehicle posture) of the vehicle within a certain period of time, so as to integrate the perception and tracking of surrounding targets. Taking the target fusion perception of intelligent networked vehicles as an example, intelligent networked vehicles use the target perception results output by their perception systems (such as various visual, millimeter wave, and lidar perception systems) to perform fusion tracking using advanced algorithms and technologies to determine the position, direction, speed, size and other information of targets around the vehicle body, so as to achieve accurate perception and understanding of the vehicle's surrounding environment, provide a perception basis for various intelligent driving solutions, and improve the vehicle's adaptability and safety to complex traffic environments.

[0036] Being able to obtain the precise motion state of the vehicle within a certain period of time is a necessary condition for the fusion perception and tracking of surrounding targets. Obtaining the motion state of the vehicle within a certain period of time can be achieved by positioning the vehicle. An optional positioning implementation method is: using satellite positioning, for example, GPS (Global Positioning System), Beidou satellite navigation system, etc. However, when the above positioning method encounters scenes where satellite signals are blocked and restored, positioning drift and jump will inevitably occur, thereby affecting the perception and tracking of surrounding targets. Another optional positioning implementation method is: using signals such as wheel speed meter, wheel angle, IMU, etc., using integration and other means, to establish a local coordinate system with a given point as the origin, and calculate the vehicle posture in the local coordinate system for target perception and tracking. This positioning scheme will not have positioning jumps and meet the target fusion perception and tracking requirements. In addition, the global coordinate projection of surrounding targets can also be performed through the conversion relationship between the local coordinate system and the satellite positioning coordinate system.

[0037] However, due to the estimated cumulative error of the vehicle body posture in the local coordinate system, and the positioning does not have satellite positioning correction to avoid jumps, the difference between the angle between the ego vehicle and the horizontal plane (i.e., the orientation angle of the vehicle body) and the actual vehicle angle (i.e., the actual direction of movement) will gradually increase. The distortion of the angle between the ego vehicle and the horizontal plane will cause the target position of the perceived target to shift, increase the perception error, and thus reduce the accuracy of target tracking.

[0038] In order to at least partially solve the above technical problems, in this embodiment, the vehicle positioning information of the vehicle in the local coordinate system is converted so that the vehicle posture is rotated to the horizontal direction in the local coordinate system while keeping the horizontal mileage of the vehicle consistent with the spatial mileage. For the vehicle positioning information of the target vehicle in the local coordinate system, the vehicle positioning information of the target vehicle can be updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle. The vehicle positioning information collected each time can be converted to rotate the vehicle posture to the horizontal direction, thereby reducing the distortion of the angle between the self-vehicle and the horizontal plane caused by the cumulative error; the target positioning information of the perceived target is converted to the local coordinate system using the vehicle positioning information updated by the target vehicle, which can reduce the offset of the target position of the perceived target in the local coordinate system caused by the distortion of the angle between the self-vehicle and the horizontal plane. Then, based on the local positioning information of each perceived target, the perceived target perceived by the perception system of the target vehicle is fused and tracked, thereby ensuring the accuracy of fusion tracking.

[0039] For the target vehicle, the vehicle positioning information of the target vehicle can be continuously obtained, wherein the target vehicle can be an intelligent connected vehicle or a similar vehicle. The vehicle positioning information of the target vehicle can be the positioning information of the target vehicle in a local coordinate system, and the local coordinate system is a fixed coordinate system with a specified position as the origin, wherein the specified position can be the vehicle start position of the target vehicle. The vehicle positioning information of the target vehicle can be used to indicate the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and can include the plane position (longitude, latitude), height, pitch angle, roll angle and heading angle, etc. During the driving process of the target vehicle, the vehicle positioning information of the target vehicle can be continuously obtained by integrating and other means based on the signals of the wheel speed meter, wheel angle, IMU, etc. The wheel speed meter can detect the rotation speed of the wheel and calculate the instantaneous speed of the vehicle based on the diameter and rotation speed of the wheel. After integrating the speed, the cumulative driving distance of the vehicle can be obtained; the wheel angle can detect the steering angle of the wheel, and combined with the speed information of the vehicle, the driving direction and trajectory of the vehicle can be calculated; the IMU can detect the acceleration, angular velocity and other parameters of the vehicle, and the position, speed and posture of the vehicle can be calculated in real time through algorithm processing.

[0040] The vehicle positioning information of the target vehicle can be obtained by a positioning module, which may include multiple sensors, including but not limited to at least one of the following: a wheel speed sensor for detecting the rotation speed of the wheel; a steering angle sensor for detecting the steering condition of the wheel; an accelerometer sensor for detecting the acceleration of the vehicle; a gyroscope sensor for detecting the angular velocity and posture (pitch angle, roll angle and yaw angle) of the vehicle.

[0041] The process of acquiring the vehicle positioning information of the target vehicle may be a continuous process. For example, the vehicle positioning information of the target vehicle may be acquired once at regular intervals. The time interval may be 10ms, 20ms, 50ms, 100ms or other values.

[0042] For each acquired vehicle positioning information of the target vehicle, the attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle can be determined. The attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle can be used to rotate the vehicle posture of the target vehicle to the horizontal direction transformation matrix in the local coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent. Among them, horizontal mileage refers to the distance between two points in the horizontal direction, excluding the influence of height difference; spatial mileage refers to the relative position or interval between two points or objects in three-dimensional space, which may include the distance in the horizontal direction and the vertical direction; horizontal direction refers to the direction perpendicular to the gravity direction (i.e., the vertical direction or the vertical direction). Each time the vehicle positioning information of the target vehicle is acquired, due to the different posture of the target vehicle, its transformation matrix rotated to the horizontal direction in the local coordinate system is also different.

[0043] After obtaining the attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle, the vehicle positioning information of the target vehicle can be updated using the attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain the updated vehicle positioning information of the target vehicle. The attitude transformation matrix matching the vehicle positioning information of the target vehicle is used to update the vehicle positioning information of the target vehicle each time. Through the above update, the orientation angle of the body of the target vehicle can be basically kept parallel to the horizontal plane. Here, the horizontal direction and the horizontal plane are relative to the local coordinate system, the horizontal plane refers to the xy plane of the local coordinate system, and the horizontal direction is the direction parallel to the horizontal plane. Here, it is set that during driving, the orientation angle of the body of the target vehicle is basically kept parallel to the xy plane of the local coordinate system.

[0044] The perception system on the vehicle can perceive the targets in the environment around the vehicle body to determine the positioning information of the perceived targets. For the perceived targets perceived by different perception systems at different times, the target fusion perception technology can be used to perform target fusion tracking to fully and accurately perceive the environment around the vehicle. Here, the perception system is a comprehensive system that can include a variety of sensors, such as millimeter wave radar, lidar, camera (including visible light camera and infrared camera, etc.), ultrasonic sensor, etc. Since each sensor has its own unique advantages and limitations, the sensors can work together to achieve complementarity and mutual verification, and jointly achieve target fusion perception. The perception timing of the perception system is a dynamic and continuous process, which can be dynamically adjusted with the changes in the vehicle's driving state and the surrounding environment, and can also continue to monitor and perceive the surrounding environment continuously.

[0045] The perception system outputs the perception results based on the coordinate system defined by the perception system (usually the vehicle body coordinate system, or the vehicle coordinate system). Taking the vehicle body coordinate system as an example, the target positioning information of each perceived target perceived by the perception system of the target vehicle can be obtained through the perception system of the target vehicle. Among them, the target positioning information of each perceived target is the positioning information of each perceived target in the vehicle body coordinate system of the target vehicle, and the target positioning information of each perceived target is used to represent the target position and posture of each perceived target at the perception time of each perceived target. Here, the target position and posture of each perceived target in the vehicle body coordinate system at the perception time of each perceived target.

[0046] In the perception system of the target vehicle, since different sensors (such as cameras, radars, etc.) are located in different positions and postures, corresponding to different coordinate systems (usually coordinate systems with the sensor itself as the coordinate origin), for a perception target perceived by a certain sensor, the positioning information of the perception target in the coordinate system corresponding to the sensor can be converted to the vehicle body coordinate system, thereby obtaining the target positioning information of the perception target, wherein the transformation matrix between the coordinate system corresponding to the sensor and the vehicle body coordinate system can be used to describe the relative position and posture relationship between the coordinate system corresponding to the sensor and the vehicle body coordinate system. Since the relative position between the sensor and the target vehicle is usually fixed, the transformation matrix between the coordinate system corresponding to the sensor and the vehicle body coordinate system is determinable (if the sensor posture remains unchanged, the transformation matrix is ​​fixed, if the sensor posture changes, the transformation matrix changes accordingly and can be determined), which can be obtained through a calibration process, wherein calibration is the process of measuring and calculating the relative position and posture between the sensor and the vehicle sound, and the calibration result (i.e., the transformation matrix) can include a rotation matrix and a translation vector. Different annotation methods may be required for different sensors and vehicles. At the same time, as the vehicle travels and the sensor is used for an increasing amount of time, the relative position and posture between the sensor and the vehicle body may change, and can be recalibrated regularly to ensure the accuracy and reliability of the transformation matrix.

[0047] The body coordinate system is a professional dynamic coordinate system used to accurately depict the vehicle's motion state. Its origin coincides with the projection point of the center point of the vehicle's rear axle on the ground. This feature enables the vehicle's state to be more accurately reflected when analyzing vehicle motion. When the vehicle is parked stably on a horizontal road, the x-axis of the body coordinate system points in the vehicle's forward direction and can be parallel to the ground; the z-axis is perpendicular to the ground and points upward; the y-axis points to the driver's left (in some definitions, the y-axis may point to the driver's right, but it is common to point to the left).

[0048] As a non-inertial system, the vehicle body coordinate system is not suitable for motion tracking in a non-inertial system because of the inapplicability of Newton's laws, the need for additional inertial forces, and the rapid changes in motion state. In order to perform fusion tracking on the perceived target, it is necessary to convert the perception results output by each perception system into a coordinate system with a fixed origin (i.e., a local coordinate system) for data association (fusion process), and track the motion trajectory of the perceived target based on the perception results at different times, and output more accurate and stable motion information of the perceived target (tracking process). The operation of fusion tracking on the perceived target can be performed by the fusion tracking module on the vehicle terminal. The pose conversion method of the target around the vehicle body is shown in formula (1):

[0049] P 原点固定坐标系 =H 自车 ·P 自车坐标系 (1)

[0050] Among them, P 原点固定坐标系 represents the converted position of the perceived target in the fixed origin coordinate system (i.e., local coordinate system), H 自车 represents the position of the vehicle in the fixed coordinate system at the origin, P 车身坐标系 Represents the position and posture of the perceived target output from each perception system in the vehicle body coordinate system. It can be seen from formula (1) that the accuracy of the vehicle's position and posture in the origin fixed coordinate system has a great influence on the target position and posture conversion.

[0051] In order to perform fusion tracking on the perceived targets perceived by the perception system of the target vehicle, the target positioning information of each perceived target can be converted into local positioning information in a local coordinate system based on the vehicle positioning information corresponding to each perceived target, so as to obtain the local positioning information of each perceived target. The above conversion process can be shown as formula (1). Here, the vehicle positioning information corresponding to each perceived target is the vehicle positioning information updated by the target vehicle, and the positioning time is consistent with the perception time of each perceived target.

[0052] The consistency of positioning time and perception time can minimize the time difference between positioning time and perception time. Optionally, in order to improve the accuracy of target fusion tracking, the vehicle positioning information of the target vehicle output by the positioning module can be parsed to supplement the vehicle positioning information of the target vehicle at different positioning times, and then obtain the vehicle positioning information of the target vehicle updated at different positioning times; or, the vehicle positioning information updated by the discrete target vehicle is supplemented to obtain the vehicle positioning information of the target vehicle updated at different positioning times. In this case, the consistency of positioning time and perception time can be that the positioning time is the same as the perception time.

[0053] Here, each sensor, each type of sensor, or each sensor combination can be used as a perception system. Since different perception systems may perceive the perception target at different perception moments, the above-mentioned conversion of the target positioning information of the perception target in the vehicle body coordinate system to the local positioning information in the local coordinate system can be performed for the perception target perceived by a perception system at a perception moment. The fusion tracking of the perception target is performed based on the local positioning information of the perception target after conversion to the local coordinate system.

[0054] like Figure 3As shown, the intelligent connected car is equipped with sensors S1, S2 and S3, and the position information of the perceived target sensed by each sensor is the position information in the corresponding sensor coordinate system, where sensor S1 corresponds to sensor coordinate system C1, sensor S2 corresponds to sensor coordinate system C2, and sensor S3 corresponds to sensor coordinate system C3. Based on the conversion matrix between different sensor coordinate systems and vehicle body coordinate systems, the position information of the perceived target in the sensor coordinate system can be converted to the vehicle body coordinate system, thereby obtaining the target position information of the perceived target sensed by different sensors in the vehicle body coordinate system.

[0055] For the perceived targets perceived by each sensor, after obtaining the target position information in the vehicle body coordinate system, it can be converted to the local coordinate system to obtain the position information of each perceived target in the same local coordinate system, that is, the local position information. Since the local position information of all perceived targets is the position information in the same local coordinate system, the perceived targets perceived by sensors S1, S2, and S3 can be fused and tracked based on the local position information.

[0056] Based on the local positioning information of each perceived target, the perceived targets perceived by the perception system of the target vehicle can be fused and tracked. When performing target fusion tracking, data association algorithms (such as joint probability data association JPDAF, multi-hypothesis tracking MHT, etc.) can be used to match and fuse the targets output by different perception systems; based on the fused target data, motion trajectory tracking is performed to output the target's relatively accurate and stable motion information (position, velocity, acceleration, etc.) in a fixed origin coordinate system.

[0057] Fusion tracking can include two parts: data fusion and target tracking. Data fusion is the fusion of data from different sensors (local positioning information of the perceived target) to obtain more comprehensive and accurate target information. Data fusion can be performed at different levels, including data-level fusion, feature-level fusion, and decision-level fusion. In the process of data fusion, the weights of each sensor can be adaptively adjusted according to factors such as the performance of different sensors and the motion state of the target to achieve the best fusion effect. For target tracking, after the perceived target is first identified, it can be initialized for tracking (which may include the initial position, velocity, acceleration and other state information of the perceived target), and the historical motion state information of the perceived target can be used in combination with the motion model (such as Kalman filter, particle filter, etc.) to predict the future state of the perceived target. At the same time, the sensor is used to observe and update the predicted state of the perceived target. If the observation result is consistent with the predicted result, the state information of the perceived target is updated; if not, the predicted state is adjusted according to the observation result. Repeating the process of state prediction and observation update can achieve continuous tracking of the perceived target.

[0058] In this embodiment, in order to improve the positioning accuracy, the posture of the vehicle can be described in three-dimensional space, and the posture estimation of the vehicle body in the local coordinate system uses three-dimensional coordinates. The tracking model used for the fusion tracking of the vehicle body surrounding targets can be a three-dimensional tracking model. Considering that the target tracking results displayed to the driver or other relevant personnel are usually two-dimensional, in order to reduce unnecessary calculation costs, the fusion tracking of the vehicle body surrounding targets can be reduced in dimension and adopt a two-dimensional plane tracking model. In the case of using a two-dimensional plane tracking model, the body posture in the local coordinate system is estimated due to the cumulative error, and the positioning does not have satellite positioning correction in order to avoid jumps. The orientation angle of the vehicle body will gradually increase with the actual movement direction. If the position of the surrounding targets in the local coordinate system is simply tracked by taking only two-dimensional information, the distortion of the angle between the vehicle and the horizontal plane will cause the converted target position in formula (1) to shift, resulting in distortion of target positioning and motion estimation, and bringing unpredictable perception errors. By adjusting the vehicle positioning information (the vehicle posture of the target vehicle in the local coordinate system) through the posture transformation matrix, the error of the angle between the vehicle and the horizontal plane can meet the error tolerance of the algorithm for perceiving target fusion tracking, thereby improving the accuracy of target fusion tracking.

[0059] Through the above steps, the vehicle positioning information of the target vehicle is continuously obtained, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is determined, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in the local coordinate system, the local coordinate system is a fixed coordinate system with the specified position as the origin, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction in the fixed coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent; the vehicle positioning information of the target vehicle is updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain the updated vehicle positioning information of the target vehicle; the target of each perception target perceived by the perception system of the target vehicle is obtained. Positioning information, wherein the target positioning information of each perception target is the positioning information of each perception target in the body coordinate system of the target vehicle, and the target positioning information of each perception target is used to represent the target posture of each perception target at the perception time of each perception target; based on the vehicle positioning information corresponding to each perception target, the target positioning information of each perception target is converted into local positioning information in the local coordinate system to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information updated by the target vehicle, and the positioning time is consistent with the perception time of each perception target; based on the local positioning information of each perception target, the perception target perceived by the perception system of the target vehicle is fused and tracked, which solves the problem of poor target tracking accuracy caused by large perception error in the tracking method of the perception target in the related art, and improves the accuracy of target tracking.

[0060] In an exemplary embodiment, continuously acquiring vehicle positioning information of a target vehicle and determining a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle includes:

[0061] S11, continuously acquiring the vehicle positioning information of the target vehicle, and using the vehicle positioning information of the target vehicle acquired each time as the current vehicle positioning information to perform the following determination operation, to obtain a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle acquired each time:

[0062] Obtaining an initial transformation matrix corresponding to the current vehicle positioning information;

[0063] The current vehicle positioning information is updated using an initial transformation matrix corresponding to the current vehicle positioning information to obtain reference vehicle positioning information;

[0064] Determine a posture transformation matrix corresponding to the reference vehicle positioning information;

[0065] Based on the attitude transformation matrix corresponding to the reference vehicle positioning information, a attitude transformation matrix corresponding to the current vehicle positioning information is determined.

[0066] The vehicle positioning information of the target vehicle is a process of continuous acquisition. For example, the vehicle positioning information of the target vehicle can be acquired periodically. For each acquired vehicle positioning information, the corresponding posture transformation matrix can be acquired respectively. Acquiring different times of vehicle positioning information can be a separate process. In this case, for each acquired vehicle positioning information, a transformation matrix can be initialized first as an initial transformation matrix. After the initial transformation matrix is ​​used to try to perform posture transformation on the corresponding vehicle positioning information, the initial transformation matrix is ​​adjusted based on the transformation matrix that is still needed to rotate the transformed vehicle positioning information to the horizontal direction, thereby obtaining the desired posture transformation matrix. However, the difference between the initial transformation matrix and the final posture transformation matrix may be large. Therefore, determining the transformation matrix that still needs to be used is not only difficult, but also has low accuracy.

[0067] To this end, in this embodiment, considering that the estimated cumulative error of the vehicle body posture is a continuous process, the difference between the posture transformation matrices corresponding to the vehicle positioning information obtained twice adjacently is usually not too large. Therefore, the posture transformation matrix corresponding to the vehicle positioning information obtained last time can be used as the initial transformation matrix corresponding to the vehicle positioning information obtained currently, thereby reducing the difficulty of determining the transformation matrix that still needs to be used, and thus improving the accuracy of the posture change matrix determination.

[0068] For each acquired vehicle positioning information of the target vehicle, the determination operation of the posture transformation matrix can be performed as the current vehicle positioning information, so as to determine the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle acquired each time. When determining the posture change matrix corresponding to the current vehicle positioning information, the initial transformation matrix corresponding to the current vehicle positioning information can be first acquired, and the current vehicle positioning information can be updated using the initial transformation matrix corresponding to the current vehicle positioning information to obtain the reference vehicle positioning information. Here, if the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the first time, the initial transformation matrix corresponding to it is the unit matrix; if the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the Nth time (N is a positive integer greater than or equal to 2), the initial transformation matrix corresponding to it is the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle acquired for the N-1th time.

[0069] For example, the local positioning information generated by the positioning system (i.e., the vehicle positioning information of the target vehicle) can be represented by a Cartesian coordinate system, where the xy axis is in the horizontal plane and the z axis is in the vertical direction. The attitude transformation matrix H can be calculated by an iterative algorithm, and the iterative steps are as follows: Step 1, define H as the initial transformation, and the initial value of H is the E matrix (i.e., the unit matrix). For non-first positioning information, H is the H finally determined in the previous iteration; Step 2, take the coordinates H of the vehicle in the local coordinate system at the current moment 自车 , perform the transformation shown in formula (2):

[0070]

[0071] here, is the coordinate of the vehicle in the local coordinate system after the preliminary transformation, that is, the aforementioned reference vehicle positioning information.

[0072] For the reference vehicle positioning information obtained, the attitude transformation matrix corresponding to the reference vehicle positioning information can be determined, that is, the vehicle posture of the target vehicle (here, the vehicle posture corresponding to the reference positioning information) is rotated to the horizontal direction of the local coordinate system while keeping the horizontal mileage of the target vehicle at the positioning time of the vehicle positioning information consistent with the spatial mileage. Based on the attitude transformation matrix corresponding to the reference vehicle positioning information, the attitude transformation matrix corresponding to the current vehicle positioning information can be determined. The attitude transformation matrix corresponding to the current vehicle positioning information can be the superposition of the initial transformation matrix corresponding to the current vehicle positioning information and the attitude transformation matrix corresponding to the reference vehicle positioning information. The superposition method can be the product of the two matrices, or other iterative methods, which are not limited in this embodiment.

[0073] For example, the iterative steps for calculating the posture transformation matrix H also include: Step 3, calculating The transformation matrix F that rotates to the horizontal direction with its own coordinates as the anchor point + ; Step 4, based on F + The transformation matrix H after iterative optimization is calculated. The updated transformation matrix H is the posture transformation matrix used this time.

[0074] It should be noted that after obtaining the attitude transformation matrix corresponding to the reference vehicle positioning information, the attitude transformation matrix corresponding to the reference vehicle positioning information can also be directly used to update the reference vehicle positioning information to the updated vehicle positioning information of the target vehicle (the vehicle positioning information to which the current vehicle positioning information is updated). By determining the attitude transformation matrix corresponding to the vehicle positioning information of the target vehicle each time, it is convenient to perform iterative processing (i.e., to facilitate subsequent calculations), thereby improving the convenience of information processing.

[0075] Through this embodiment, the posture transformation matrix corresponding to the vehicle positioning information acquired last time is used as the initial transformation matrix corresponding to the vehicle positioning information acquired currently, so that the efficiency and accuracy of determining the posture transformation matrix can be improved.

[0076] In an exemplary embodiment, determining a posture transformation matrix corresponding to reference vehicle positioning information includes:

[0077] S21 determines a rotation matrix corresponding to the reference vehicle positioning information, wherein the rotation matrix corresponding to the reference vehicle positioning information is a rotation matrix used to rotate the vehicle posture indicated by the reference vehicle positioning information to a horizontal direction;

[0078] S22, determining a posture transformation matrix corresponding to the reference vehicle positioning information according to the rotation matrix corresponding to the reference vehicle positioning information.

[0079] In this embodiment, in order to determine the posture transformation matrix corresponding to the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information can be first determined. Here, the rotation matrix corresponding to the reference vehicle positioning information is a rotation matrix used to rotate the vehicle posture indicated by the reference vehicle positioning information to the horizontal direction. The rotation matrix determined at this time is a rotation matrix that does not take into account the vehicle coordinates (without considering keeping the horizontal mileage consistent with the spatial mileage).

[0080] For the rotation matrix corresponding to the reference vehicle positioning information, taking into account the vehicle position coordinates represented by the reference vehicle positioning information (and the vehicle position coordinates represented by the current vehicle positioning information, where the vehicle position coordinates are the position coordinates in the local coordinate system), the posture transformation matrix corresponding to the reference vehicle positioning information can be determined based on the rotation matrix corresponding to the reference vehicle positioning information.

[0081] Optionally, the method for determining the posture transformation matrix corresponding to the reference vehicle positioning information can be: based on the rotation matrix corresponding to the reference vehicle positioning information and the reference vehicle positioning information (the vehicle position information in the reference vehicle positioning information), determine the posture transformation matrix corresponding to the reference vehicle positioning information. Here, the posture transformation matrix corresponding to the reference vehicle positioning information is a transformation matrix that rotates to the horizontal direction with the reference vehicle positioning information as the anchor point with its own coordinates (the vehicle body coordinates indicated by the reference vehicle positioning information).

[0082] Through this embodiment, by first determining the rotation matrix corresponding to the vehicle positioning information, and then determining the corresponding posture transformation matrix based on the determined rotation matrix combined with the vehicle positioning information, the convenience and accuracy of determining the posture transformation matrix can be improved.

[0083] In an exemplary embodiment, determining a rotation matrix corresponding to reference vehicle positioning information includes:

[0084] S31, determining a horizontal plane normal vector of a vehicle posture of a target vehicle in a local coordinate system based on a rotation component corresponding to the reference vehicle positioning information, and obtaining a target normal vector;

[0085] S32, determining a rotation axis direction vector of the target normal vector and a rotation angle corresponding to the rotation axis direction vector of the target normal vector;

[0086] S33, determining a rotation matrix corresponding to the reference vehicle positioning information according to the rotation axis direction vector of the target normal vector and the rotation angle corresponding to the rotation axis direction vector of the target normal vector.

[0087] In this embodiment, the rotation matrix corresponding to the reference vehicle positioning information can be constructed by rotation components. Here, the rotation components can be used to describe the rotation characteristics of an object in space (which may include a rotation axis and a rotation angle). In the vehicle positioning information, the rotation components may refer to a set of parameters (which may include Euler angles, rotation matrices, quaternions, and rotation vectors) of the rotation state of the vehicle in three-dimensional space relative to a reference coordinate system (e.g., a local coordinate system), which can accurately represent the orientation or posture of the vehicle in space.

[0088] The rotation matrix corresponding to the reference vehicle positioning information can be determined using a Rodrigues rotation formula or a similar rotation formula. In order to determine the rotation matrix, the rotation axis direction vector and the rotation angle corresponding to the reference vehicle positioning information can be determined first, and then the rotation matrix corresponding to the reference vehicle positioning information can be determined based on the determined rotation axis direction vector and the rotation angle. Here, the method for determining the aforementioned rotation axis direction vector and the rotation angle can be: based on the rotation component corresponding to the reference vehicle positioning information, determine the horizontal plane normal vector of the vehicle posture of the target vehicle in the local coordinate system to obtain the target normal vector; determine the rotation axis direction vector of the target normal vector, and the rotation angle corresponding to the rotation axis direction vector of the target normal vector.

[0089] For example, in calculating When the rotation matrix F to the horizontal plane is obtained, the normal vector of the plane where the vehicle is located can be calculated first, as shown in formula (3):

[0090]

[0091] Among them, Z 自车 The xy plane normal vector representing the current position of the vehicle in the local coordinate system, SO3 自车 express The rotation component of .

[0092] Calculate the rotation axis direction vector D, as shown in formula (4):

[0093]

[0094] Calculate the rotation angle θ as shown in formula (5):

[0095]

[0096] According to the Rodrigues rotation formula, combined with formula (4) and formula (5), we can calculate The rotation matrix F for the horizontal rotation.

[0097] Through this embodiment, by first determining the rotation axis direction vector and rotation angle corresponding to the vehicle positioning information, and then determining the rotation matrix for rotating the vehicle positioning information to a horizontal level based on the rotation formula used in combination with the determined rotation axis direction vector and rotation angle, the accuracy of the rotation matrix determination can be improved.

[0098] In an exemplary embodiment, determining a posture transformation matrix corresponding to the reference vehicle positioning information according to a rotation matrix corresponding to the reference vehicle positioning information includes:

[0099] S41, the sum of the unit matrix and the vehicle coordinates in the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information, and the product of the difference between the unit matrix and the vehicle coordinates in the reference vehicle positioning information are determined as the posture transformation matrix corresponding to the reference vehicle positioning information.

[0100] Compared with the attitude transformation matrix corresponding to the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information does not consider the translation component. Therefore, after determining the rotation matrix corresponding to the reference vehicle positioning information, the attitude transformation matrix corresponding to the reference vehicle positioning information can be calculated in combination with the vehicle coordinates in the reference vehicle positioning information. The calculation method can be: the sum of the unit matrix and the vehicle coordinates in the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information, and the product of the difference between the unit matrix and the vehicle coordinates in the reference vehicle positioning information are determined as the attitude transformation matrix corresponding to the reference vehicle positioning information.

[0101] For example, calculate After the rotation matrix F to the horizontal plane, we can calculate The transformation matrix F that rotates to the horizontal direction with its own coordinates as the anchor point + , as shown in formula (6):

[0102] F + =(E+t)·F·(Et) (6)

[0103] Where t represents coordinate.

[0104] Through this embodiment, the posture transformation matrix corresponding to the reference vehicle positioning information is determined according to the rotation matrix corresponding to the reference vehicle positioning information, so that the object posture can be updated and the accuracy of the posture transformation matrix determination can be improved.

[0105] In an exemplary embodiment, determining a posture transformation matrix corresponding to current vehicle positioning information based on a posture transformation matrix corresponding to reference vehicle positioning information includes:

[0106] S51, determining the product of the initial transformation matrix corresponding to the current vehicle positioning information and the posture transformation matrix corresponding to the reference vehicle positioning information as the posture transformation matrix corresponding to the current vehicle positioning information.

[0107] In this embodiment, for the current vehicle positioning information, the product of the initial transformation matrix corresponding to it and the posture transformation matrix corresponding to it can be directly determined as the posture transformation matrix corresponding to it, without adding additional processing operations, thereby improving the convenience of determining the posture transformation matrix.

[0108] For example, based on F + The method of calculating the transformation matrix H after iterative optimization can be shown as formula (7):

[0109] H=F + ·H (7)

[0110] After obtaining the latest vehicle positioning each time, the aforementioned steps 1 to 4 can be repeated according to the latest positioning information to calculate the updated H transformation matrix for updating H 自车 , to avoid target offset, that is, to remove the perceptual fusion target offset caused by the accumulated error of local positioning.

[0111] refer to Figure 4 ,The unprocessed local positioning information is iteratively processed to obtain the available local positioning information, and the available local positioning information can be used in the process of determining the local positioning information of the perceived target.

[0112] Through this embodiment, the product of the initial transformation matrix corresponding to the vehicle positioning information and the posture transformation matrix corresponding to the vehicle positioning information is directly determined as the posture transformation matrix corresponding to the vehicle positioning information, which can improve the convenience of determining the posture transformation matrix.

[0113] In an exemplary embodiment, based on the vehicle positioning information corresponding to each perception target, the target positioning information of each perception target is converted into local positioning information in a local coordinate system to obtain the local positioning information of each perception target, including:

[0114] S61, converting the vehicle positioning information corresponding to each perception target into two-dimensional positioning information in a local coordinate system to obtain two-dimensional vehicle positioning information corresponding to each perception target;

[0115] S62, using the two-dimensional vehicle positioning information corresponding to each perception target to convert the target positioning information of each perception target into local positioning information in a local coordinate system, to obtain the local positioning information of each perception target.

[0116] In order to improve the positioning accuracy, the vehicle body posture estimation in the local coordinate system uses three-dimensional coordinates. When the vehicle body surrounding target fusion tracking is performed, the local positioning information can choose to use three-dimensional coordinates. The computational complexity of surrounding target fusion tracking based on three-dimensional coordinates is high, and some perception target display scenes do not need to display the height information of the surrounding targets. Even if the height information needs to be displayed, considering that the surrounding targets are usually vehicles, the height can directly use the vehicle height of the target vehicle. In order to reduce unnecessary computational costs, the surrounding target fusion tracking can be reduced in dimension to use a two-dimensional plane tracking model. Due to the cumulative error in the estimation of the vehicle body posture in the local coordinate system, the orientation angle of the vehicle body will gradually increase with the actual direction of movement. If the position of the surrounding targets in the local coordinate system is simply taken as two-dimensional information for fusion tracking, it will cause distortion of target positioning and motion estimation, resulting in unpredictable perception errors.

[0117] For example, referring to formula (1), the ego vehicle posture in formula (1) is described in three-dimensional space. Due to the calculation of cumulative errors and the fact that there is no satellite positioning correction in order to avoid jumps, the error between the angle between the ego vehicle and the horizontal plane and the actual vehicle angle will gradually increase. When the tracking model uses simplified two-dimensional coordinates, the distortion of the angle between the ego vehicle and the horizontal plane will cause the converted target position in formula (1) to shift, increasing the perception error.

[0118] In this embodiment, after determining the vehicle positioning information corresponding to each perception target by adopting the vehicle positioning information processing method in any of the aforementioned embodiments, the vehicle positioning information corresponding to each perception target is converted into two-dimensional positioning information in a local coordinate system to obtain two-dimensional vehicle positioning information corresponding to each perception target; and the target positioning information of each perception target is converted into local positioning information in a local coordinate system using the two-dimensional vehicle positioning information corresponding to each perception target to obtain local positioning information of each perception target. By taking the converted self-vehicle positioning information (i.e., updated vehicle positioning information) as two-dimensional information for perception target fusion tracking, it can be well compatible with the two-dimensional plane tracking model, and can ensure that the positioning information of the perception target used meets the algorithm error tolerance, so that the error of perception target fusion tracking is reduced to an acceptable range.

[0119] Here, if Figure 5As shown, the process of converting three-dimensional positioning information into two-dimensional positioning information may include the following steps:

[0120] Step S502, inputting three-dimensional positioning information;

[0121] Step S504, select a projection plane. Common options include projection onto the xy plane (horizontal plane, ignoring height z), projection onto the xz plane (side view, ignoring width y), and projection onto the yz plane (top view, ignoring length x). In this embodiment, projection onto the xy plane is selected;

[0122] Step S506, performing projection calculation, and performing coordinate conversion according to the selected projection plane. In the case of projection onto the xy plane, the z coordinate can be ignored, and only the x and y coordinates are retained as two-dimensional positioning information;

[0123] Step S508: output the converted two-dimensional positioning information.

[0124] Through this embodiment, by using a two-dimensional plane tracking model to perform fusion tracking of targets around the vehicle, unnecessary calculation costs can be reduced, thereby reducing the occupation of computing resources and improving the efficiency of determining positioning information.

[0125] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0126] According to another aspect of the embodiments of the present application, a tracking device for sensing a target is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the term "module" may implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and contemplated.

[0127] Figure 6 is a structural block diagram of an optional tracking device for sensing a target according to an embodiment of the present application, such as Figure 6 As shown, including:

[0128] The first execution unit 602 is used to continuously obtain the vehicle positioning information of the target vehicle and determine the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in the local coordinate system, the local coordinate system is a fixed coordinate system with the specified position as the origin, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction of the fixed coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent;

[0129] An updating unit 604 is used to update the vehicle positioning information of the target vehicle using a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain updated vehicle positioning information of the target vehicle;

[0130] An acquisition unit 606 is used to acquire target positioning information of each perceived target perceived by the perception system of the target vehicle, wherein the target positioning information of each perceived target is positioning information of each perceived target in the body coordinate system of the target vehicle, and the target positioning information of each perceived target is used to represent the target position of each perceived target at the perception time of each perceived target;

[0131] The conversion unit 608 is used to convert the target positioning information of each perception target into local positioning information in a local coordinate system based on the vehicle positioning information corresponding to each perception target, so as to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information updated by the target vehicle, and the positioning time is consistent with the perception time of each perception target;

[0132] The second execution unit 610 is used to perform fusion tracking on the perception targets perceived by the perception system of the target vehicle based on the local positioning information of each perception target.

[0133] It should be noted that the first execution unit 602 in this embodiment can be used to execute the above step S202, the update unit 604 in this embodiment can be used to execute the above step S204, the acquisition unit 606 in this embodiment can be used to execute the above step S206, the conversion unit 608 in this embodiment can be used to execute the above step S208, and the second execution unit 610 in this embodiment can be used to execute the above step S210.

[0134] Through the embodiments of the present application, the vehicle positioning information of the target vehicle is continuously acquired, and a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is determined, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in a local coordinate system, the local coordinate system is a fixed coordinate system with a specified position as the origin, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to the horizontal direction in the fixed coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent; the vehicle positioning information of the target vehicle is updated using the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to obtain the updated vehicle positioning information of the target vehicle; the target of each perception target perceived by the perception system of the target vehicle is obtained; Positioning information, wherein the target positioning information of each perception target is the positioning information of each perception target in the body coordinate system of the target vehicle, and the target positioning information of each perception target is used to represent the target posture of each perception target at the perception time of each perception target; based on the vehicle positioning information corresponding to each perception target, the target positioning information of each perception target is converted into local positioning information in the local coordinate system to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information updated by the target vehicle, and the positioning time is consistent with the perception time of each perception target; based on the local positioning information of each perception target, the perception target perceived by the perception system of the target vehicle is fused and tracked, which solves the problem of poor target tracking accuracy caused by large perception error in the tracking method of the perception target in the related art, and improves the accuracy of target tracking.

[0135] In an exemplary embodiment, the first execution unit includes:

[0136] The execution module is used to continuously obtain the vehicle positioning information of the target vehicle, and use the vehicle positioning information of the target vehicle obtained each time as the current vehicle positioning information to perform the following determination operations to obtain the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle obtained each time:

[0137] Acquire an initial transformation matrix corresponding to the current vehicle positioning information, wherein, when the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the first time, the initial transformation matrix corresponding to the current vehicle positioning information is a unit matrix, and when the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the Nth time, the initial transformation matrix corresponding to the current vehicle positioning information is a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle acquired for the N-1th time, and N is a positive integer greater than or equal to 2;

[0138] The current vehicle positioning information is updated using an initial transformation matrix corresponding to the current vehicle positioning information to obtain reference vehicle positioning information;

[0139] Determine a posture transformation matrix corresponding to the reference vehicle positioning information, wherein the posture transformation matrix corresponding to the reference vehicle positioning information is a transformation matrix that uses the vehicle position indicated by the reference vehicle positioning information as an anchor point and is used to rotate the vehicle posture indicated by the reference vehicle positioning information to a horizontal direction;

[0140] Based on the attitude transformation matrix corresponding to the reference vehicle positioning information, a attitude transformation matrix corresponding to the current vehicle positioning information is determined.

[0141] In an exemplary embodiment, the first determining module includes:

[0142] A first determination submodule is used to determine a rotation matrix corresponding to the reference vehicle positioning information, wherein the rotation matrix corresponding to the reference vehicle positioning information is a rotation matrix used to rotate the vehicle posture indicated by the reference vehicle positioning information to a horizontal direction;

[0143] The second determination submodule is used to determine the posture transformation matrix corresponding to the reference vehicle positioning information according to the rotation matrix corresponding to the reference vehicle positioning information.

[0144] In an exemplary embodiment, the first determining submodule includes:

[0145] A first determination subunit is used to determine a horizontal plane normal vector of a vehicle posture of a target vehicle in a local coordinate system based on a rotation component corresponding to the reference vehicle positioning information, and obtain a target normal vector;

[0146] A second determining subunit, used to determine a rotation axis direction vector of a target normal vector and a rotation angle corresponding to the rotation axis direction vector of the target normal vector;

[0147] The third determination subunit is used to determine the rotation matrix corresponding to the reference vehicle positioning information according to the rotation axis direction vector of the target normal vector and the rotation angle corresponding to the rotation axis direction vector of the target normal vector.

[0148] In an exemplary embodiment, the second determining submodule includes:

[0149] The fourth determination subunit is used to determine the sum of the unit matrix and the vehicle coordinates in the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information, and the product of the difference between the unit matrix and the vehicle coordinates in the reference vehicle positioning information as the posture transformation matrix corresponding to the reference vehicle positioning information.

[0150] In an exemplary embodiment, the second determining module includes:

[0151] The third determination submodule is used to determine the product of the initial transformation matrix corresponding to the current vehicle positioning information and the posture transformation matrix corresponding to the reference vehicle positioning information as the posture transformation matrix corresponding to the current vehicle positioning information.

[0152] In an exemplary embodiment, the conversion unit includes:

[0153] A first conversion module is used to convert the vehicle positioning information corresponding to each perception target into two-dimensional positioning information in a local coordinate system to obtain two-dimensional vehicle positioning information corresponding to each perception target;

[0154] The second conversion module is used to use the two-dimensional vehicle positioning information corresponding to each perception target to convert the target positioning information of each perception target into local positioning information in a local coordinate system to obtain the local positioning information of each perception target.

[0155] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0156] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when executed.

[0157] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0158] According to another aspect of the embodiments of the present application, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0159] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0160] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0161] According to another aspect of the embodiment of the present application, a computer program product is also provided, which includes a computer program / instruction, and the computer program / instruction contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the central processing unit 701, various functions provided by the embodiment of the present application are executed. The above-mentioned serial numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0162] Figure 7 The computer system structure block diagram of the electronic device used to implement the embodiment of the present application is schematically shown. Figure 7 As shown, the computer system 700 includes a central processing unit 701 (CPU), which can perform various appropriate actions and processes according to the program stored in the read-only memory 702 (ROM) or the program loaded from the storage part 708 to the random access memory 703 (RAM). Various programs and data required for system operation are also stored in the random access memory 703. The central processing unit 701, the read-only memory 702 and the random access memory 703 are connected to each other through a bus 704. An input / output interface 705 (Input / Output interface, i.e., I / O interface) is also connected to the bus 704.

[0163] The following components are connected to the input / output interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed so that a computer program read therefrom is installed into the storage section 708 as needed.

[0164] In particular, according to an embodiment of the present application, the process described in each method flow chart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer readable medium, and the computer program contains a program code for executing the method shown in the flow chart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the central processor 701, various functions defined in the system of the present application are executed.

[0165] It should be noted that Figure 7 The computer system 700 of the electronic device (which may be the aforementioned electronic device) shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0166] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0167] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for tracking a perceived target, characterized in that: include: Continuously acquiring vehicle positioning information of the target vehicle, and determining a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in a local coordinate system, the local coordinate system is a fixed coordinate system with a specified position as the origin, the vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to a horizontal direction transformation matrix in the local coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent; Using a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle to update the vehicle positioning information of the target vehicle, to obtain updated vehicle positioning information of the target vehicle; Acquire target positioning information of each perceived target perceived by the perception system of the target vehicle, wherein the target positioning information of each perceived target is positioning information of each perceived target in the body coordinate system of the target vehicle, and the target positioning information of each perceived target is used to represent the target position of each perceived target at the perception time of each perceived target; Based on the vehicle positioning information corresponding to each of the perception targets, the target positioning information of each perception target is converted into local positioning information in the local coordinate system to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information updated by the target vehicle, and the positioning time is consistent with the perception time of each perception target; Based on the local positioning information of each perceived target, the perceived targets perceived by the perception system of the target vehicle are fused and tracked.

2. The method according to claim 1, characterized in that The continuously acquiring the vehicle positioning information of the target vehicle and determining the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle includes: Continuously obtain the vehicle positioning information of the target vehicle, and use the vehicle positioning information of the target vehicle obtained each time as the current vehicle positioning information to perform the following determination operation to obtain the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle obtained each time: Acquire an initial transformation matrix corresponding to the current vehicle positioning information, wherein, when the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the first time, the initial transformation matrix corresponding to the current vehicle positioning information is a unit matrix, and when the current vehicle positioning information is the vehicle positioning information of the target vehicle acquired for the Nth time, the initial transformation matrix corresponding to the current vehicle positioning information is a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle acquired for the N-1th time, and N is a positive integer greater than or equal to 2; The current vehicle positioning information is updated using an initial transformation matrix corresponding to the current vehicle positioning information to obtain reference vehicle positioning information; Determine a posture transformation matrix corresponding to the reference vehicle positioning information, wherein the posture transformation matrix corresponding to the reference vehicle positioning information is a transformation matrix that uses the vehicle position indicated by the reference vehicle positioning information as an anchor point and is used to rotate the vehicle posture indicated by the reference vehicle positioning information to the horizontal direction; Based on the attitude transformation matrix corresponding to the reference vehicle positioning information, a attitude transformation matrix corresponding to the current vehicle positioning information is determined.

3. The method according to claim 2, characterized in that The determining of the attitude transformation matrix corresponding to the reference vehicle positioning information comprises: Determine a rotation matrix corresponding to the reference vehicle positioning information, wherein the rotation matrix corresponding to the reference vehicle positioning information is a rotation matrix used to rotate the vehicle posture indicated by the reference vehicle positioning information to the horizontal direction; According to the rotation matrix corresponding to the reference vehicle positioning information, a posture transformation matrix corresponding to the reference vehicle positioning information is determined.

4. The method according to claim 3, characterized in that The determining of a rotation matrix corresponding to the reference vehicle positioning information comprises: Determine a horizontal plane normal vector of a vehicle posture of the target vehicle in the local coordinate system based on a rotation component corresponding to the reference vehicle positioning information, and obtain a target normal vector; Determine a rotation axis direction vector of the target normal vector and a rotation angle corresponding to the rotation axis direction vector of the target normal vector; A rotation matrix corresponding to the reference vehicle positioning information is determined according to the rotation axis direction vector of the target normal vector and the rotation angle corresponding to the rotation axis direction vector of the target normal vector.

5. The method according to claim 3, characterized in that: The step of determining a posture transformation matrix corresponding to the reference vehicle positioning information according to a rotation matrix corresponding to the reference vehicle positioning information comprises: The sum of the unit matrix and the vehicle coordinates in the reference vehicle positioning information, the rotation matrix corresponding to the reference vehicle positioning information, and the product of the difference between the unit matrix and the vehicle coordinates in the reference vehicle positioning information are determined as the posture transformation matrix corresponding to the reference vehicle positioning information.

6. The method according to claim 2, characterized in that The step of determining the posture transformation matrix corresponding to the current vehicle positioning information based on the posture transformation matrix corresponding to the reference vehicle positioning information comprises: The product of the initial transformation matrix corresponding to the current vehicle positioning information and the posture transformation matrix corresponding to the reference vehicle positioning information is determined as the posture transformation matrix corresponding to the current vehicle positioning information.

7. The method according to any one of claims 1 to 6, characterized in that The converting the target positioning information of each perception target into the local positioning information in the local coordinate system based on the vehicle positioning information corresponding to each perception target to obtain the local positioning information of each perception target includes: Converting the vehicle positioning information corresponding to each of the sensed targets into two-dimensional positioning information in the local coordinate system to obtain two-dimensional vehicle positioning information corresponding to each of the sensed targets; The target positioning information of each perception target is converted into local positioning information in the local coordinate system using the two-dimensional vehicle positioning information corresponding to each perception target, so as to obtain the local positioning information of each perception target.

8. A tracking device for sensing a target, characterized in that: include: A first execution unit is used to continuously acquire vehicle positioning information of a target vehicle, and determine a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, wherein the vehicle positioning information of the target vehicle is the positioning information of the target vehicle in a local coordinate system, and the local coordinate system is a fixed coordinate system with a specified position as an origin. The vehicle positioning information of the target vehicle is used to represent the vehicle posture of the target vehicle at the positioning time of the vehicle positioning information, and the posture transformation matrix corresponding to the vehicle positioning information of the target vehicle is used to rotate the vehicle posture of the target vehicle to a horizontal direction transformation matrix in the local coordinate system while keeping the horizontal mileage and spatial mileage of the target vehicle at the positioning time of the vehicle positioning information consistent; An updating unit, configured to update the vehicle positioning information of the target vehicle using a posture transformation matrix corresponding to the vehicle positioning information of the target vehicle, to obtain updated vehicle positioning information of the target vehicle; an acquisition unit, configured to acquire target positioning information of each perception target perceived by the perception system of the target vehicle, wherein the target positioning information of each perception target is positioning information of each perception target in a body coordinate system of the target vehicle, and the target positioning information of each perception target is used to represent a target position of each perception target at a perception time of each perception target; A conversion unit, configured to convert the target positioning information of each perception target into local positioning information in the local coordinate system based on the vehicle positioning information corresponding to each perception target, so as to obtain the local positioning information of each perception target, wherein the vehicle positioning information corresponding to each perception target is the vehicle positioning information updated by the target vehicle, and the positioning time of the vehicle positioning information is consistent with the perception time of each perception target; The second execution unit is used to perform fusion tracking on the perception targets perceived by the perception system of the target vehicle based on the local positioning information of each perception target.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the method described in any one of claims 1 to 7 when executed by a processor.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method described in any one of claims 1 to 7 are implemented.

Images