Track planning method and device based on vehicle-mounted millimeter wave radar, computer readable storage medium and computer program product

Abstract

Embodiments of the present application provide a kind of based on vehicle-mounted millimeter wave radar's track planning method, device, computer readable storage medium and computer program product, the method includes: obtaining echo signal, pre-processing echo signal to generate three-dimensional radar data cube;Target detection is carried out to filter out all target point track, corresponding save each target point track corresponding point track information;The angle spectrum of each target point track is extracted from three-dimensional radar data cube, the target parameter corresponding to each target point track is calculated, target parameter includes: peak bottom noise ratio, main side lobe ratio, main peak three four peak ratio and target height;The actual confidence of each target point track corresponding is calculated;And the reliable state of each target point track is determined, based on the reliable state of each target point track corresponding track initiation clustering and track association.This embodiment can prevent false moving target to generate false track and be stronger.

Description

Technical Field

[0001] This invention relates to the field of vehicle-mounted millimeter-wave radar detection technology, and in particular to a trajectory planning method, apparatus, computer-readable storage medium, and computer program product based on vehicle-mounted millimeter-wave radar. Background Technology

[0002] Millimeter-wave radar has become an indispensable in-vehicle sensor for intelligent driving assistance systems. When detecting targets, millimeter-wave radar often detects false moving targets generated by the reflection of detection echoes from environmental interference objects (such as the ground, walls, or leaves blown by airflow). To prevent false moving targets from forming false tracks during trajectory planning, it is usually necessary to filter out false targets.

[0003] One existing method for filtering false moving targets in the trajectory planning process mainly relies on the target's distance and Doppler velocity information. However, this method is difficult to effectively distinguish between false moving targets and real targets. In addition, another existing method uses the target's angle information to filter false moving targets. However, this method associates the angle information as an independent attribute, and the analysis of the angle spectrum characteristics of false moving targets is still relatively limited. As a result, the filtering effect of false moving targets is still relatively poor in complex environments. Summary of the Invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a trajectory planning method based on vehicle-mounted millimeter-wave radar, which can prevent false moving targets from generating false trajectories and has stronger stability.

[0005] A further technical problem to be solved by the embodiments of the present invention is to provide a trajectory planning device based on vehicle-mounted millimeter-wave radar, which can prevent false moving targets from generating false trajectories and has stronger stability.

[0006] A further technical problem to be solved by the embodiments of the present invention is to provide a computer-readable storage medium for storing a computer program that can prevent false moving targets from generating false tracks and has greater stability.

[0007] A further technical problem to be solved by the embodiments of the present invention is to provide a computer program product that can prevent false moving targets from generating false tracks and has stronger stability.

[0008] To address the aforementioned technical problems, this invention first provides the following technical solution: a trajectory planning method based on vehicle-mounted millimeter-wave radar, comprising the following steps: The echo signal received by the vehicle-mounted millimeter-wave radar is acquired, and the echo signal is preprocessed to generate a three-dimensional radar data cube that reflects the three dimensions of range, Doppler, and angle. In the three-dimensional radar data cube, target detection is performed from two dimensions: range and Doppler to filter out all target points. The point information corresponding to each target point is saved, and the point information includes range, Doppler velocity information and signal-to-noise ratio. The angle spectrum of each target point is extracted from the three-dimensional radar data cube. Based on the angle spectrum and the point information, the target parameters corresponding to each target point are calculated. The target parameters include: peak noise floor ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height. The actual confidence level corresponding to each target point is calculated using a preset confidence level calculation model based on the point information and the target parameters; and The credibility status of each target point is marked based on the actual confidence level corresponding to each target point, and track initial clustering and track association are performed based on the credibility status corresponding to each target point. The credibility status includes credibility points and untrustworthy points.

[0009] Furthermore, the step of calculating the actual confidence level corresponding to each target point based on the point information and the target parameters using a preset confidence level calculation model includes: Based on the Doppler velocity information corresponding to each target point, moving targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height corresponding to the moving target are input into the confidence calculation model to calculate the corresponding actual confidence level.

[0010] Furthermore, the step of calculating the actual confidence level corresponding to each target point based on the point information and the target parameters using a preset confidence level calculation model further includes: Based on the Doppler velocity information corresponding to each target point, static targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, and target height corresponding to the static target are input into the confidence calculation model to calculate the corresponding actual confidence level.

[0011] Furthermore, the initial clustering of the trajectory based on the trustworthy state corresponding to each target point specifically refers to: filtering out moving targets that meet the first preset condition and are marked as untrustworthy points, and performing initial clustering of the trajectory on the remaining moving targets. The first preset condition is: the absolute value of the corresponding speed is greater than a preset speed threshold and the corresponding distance is less than a preset distance threshold.

[0012] Furthermore, the trajectory association based on the credible state corresponding to each target point specifically refers to: screening and analyzing each moving target and the target trajectory to be associated, associating all credible points with credible points, and associating uncredible points that meet the second preset condition with the target trajectory that meets the third preset condition; the second preset condition includes that the absolute value of the corresponding speed is greater than a preset speed threshold and the corresponding distance is less than a preset distance threshold; the third preset condition includes that the number of consecutive frames of the trajectory is greater than a preset frame threshold and that it is a moving trajectory.

[0013] Furthermore, the confidence calculation model is constructed based on a comprehensive discriminant function or a logical decision tree.

[0014] Furthermore, the preprocessing includes distance-dimensional FFT, Doppler-dimensional FFT, and angular-dimensional FFT; or, the preprocessing is beamforming.

[0015] On the other hand, in order to solve the above-mentioned further technical problems, the present invention provides the following technical solution: a trajectory planning device based on vehicle-mounted millimeter-wave radar, connected to vehicle-mounted millimeter-wave radar, the device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the trajectory planning method of vehicle-mounted millimeter-wave radar as described in any of the above claims.

[0016] Furthermore, in order to solve the aforementioned technical problems, the present invention provides the following technical solution: a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the trajectory planning method based on vehicle-mounted millimeter-wave radar as described above.

[0017] Furthermore, in order to solve the aforementioned technical problems, the present invention provides the following technical solution: a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, it implements the steps of the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any of the above claims.

[0018] After adopting the above technical solution, the embodiments of the present invention have at least the following beneficial effects: By preprocessing the echo signal of the vehicle-mounted millimeter-wave radar, the embodiments of the present invention generate a three-dimensional radar data cube reflecting the three dimensions of range, Doppler, and angle, which can comprehensively capture the spatial position, motion state, and angular characteristics of the target, avoiding the loss of target information due to insufficient data dimensions; furthermore, target detection and point information storage are performed from the range-Doppler two dimensions, improving the effectiveness and accuracy of target point screening; then, by using angle spectrum extraction and multi-dimensional target parameter calculation, combined with the stored point information, target parameters such as peak noise floor ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height are calculated, wherein the angle domain related parameters can accurately characterize the signal quality, and the target height... The parameters are specifically adapted for vehicle-mounted scenarios, effectively distinguishing between valid road targets and high-altitude interference signals, thus improving the accuracy of target recognition. Furthermore, a pre-defined confidence calculation model is employed, integrating point information with the aforementioned multi-dimensional target parameters to quantify the actual confidence level of each target point. Compared to existing methods that determine point validity based on a single feature, this approach provides a more comprehensive and objective assessment of point credibility, accurately classifying credible and uncredible points, providing a basis for subsequent trajectory processing, and reducing interference from false points. Finally, after determining the credibility status of the target points based on the actual confidence level, trajectory initiation clustering and trajectory association operations are performed. This effectively prevents false points from participating in trajectory construction, significantly reducing the probability of initiation misjudgment and association errors, resulting in more stable and accurate trajectory tracking. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the steps of an optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention.

[0020] Figure 2 This is an experimental scenario diagram of an optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention.

[0021] Figure 3 An optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention is shown, which is the angle spectrum of a false target.

[0022] Figure 4 The angle spectrum of a real target is shown in an optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention.

[0023] Figure 5 This is a trajectory distribution map before false target filtering in an optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention.

[0024] Figure 6 This is an optional embodiment of the trajectory planning method based on vehicle-mounted millimeter-wave radar of the present invention, showing the trajectory distribution map after false target filtering.

[0025] Figure 7 This is a schematic diagram of an optional embodiment of the trajectory planning device based on vehicle-mounted millimeter-wave radar of the present invention.

[0026] Figure 8 This is a functional block diagram of an optional embodiment of the trajectory planning device based on vehicle-mounted millimeter-wave radar of the present invention. Detailed Implementation

[0027] The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the following illustrative embodiments and descriptions are only used to explain the present invention and are not intended to limit the present invention. Moreover, the embodiments and features in the embodiments of the present application can be combined with each other unless otherwise specified.

[0028] like Figure 1 As shown, an optional embodiment of the present invention provides a trajectory planning method based on vehicle-mounted millimeter-wave radar, comprising the following steps: S1: Acquire the echo signal received by the vehicle-mounted millimeter-wave radar 1 through wave detection, and preprocess the echo signal to generate a three-dimensional radar data cube reflecting the three dimensions of distance, Doppler, and angle. S2: Target detection is performed in the three-dimensional radar data cube from two dimensions: range and Doppler to filter out all target traces. The trace information includes range, Doppler velocity (DV), and signal-to-noise ratio (SNR). S3: Extract the angle spectrum of each target point from the three-dimensional radar data cube, and calculate the target parameters corresponding to each target point based on the angle spectrum and the point information. The target parameters include: peak noise floor ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio and target height. S4: Calculate the actual confidence level for each target point based on the point information and the target parameters using a preset confidence level calculation model; and S5: Mark the credibility state of each target point based on the actual confidence level corresponding to each target point, and perform initial clustering and trajectory association based on the credibility state corresponding to each target point. The credibility state includes credibility points and untrustworthy points.

[0029] This invention preprocesses the echo signal from vehicle-mounted millimeter-wave radar to generate a three-dimensional radar data cube reflecting three dimensions: range, Doppler, and angle. This comprehensively captures the target's spatial position, motion state, and angular characteristics, avoiding the loss of target information due to insufficient data dimensions. Furthermore, it performs target detection and point information storage from the range-Doppler dimensions, improving the effectiveness and accuracy of target point selection. Then, using angle spectrum extraction and multi-dimensional target parameter calculation, combined with the stored point information, it calculates target parameters such as peak-to-noise ratio, main-to-sidelobe ratio, main-peak-to-third-to-four-peak ratio, and target height. The angle-domain parameters accurately characterize signal quality, while the target height parameter is specifically adapted to the vehicle-mounted scenario. It can effectively distinguish between valid targets on the road surface and high-altitude interference signals, improving the accuracy of target recognition. Furthermore, by employing a pre-set confidence calculation model, it integrates point information with the aforementioned multi-dimensional target parameters to quantify the actual confidence of each target point. Compared to existing methods that determine the validity of points based on a single feature, it can more comprehensively and objectively assess the credibility of points, accurately distinguish between credible and uncredible points, and provide a basis for judgment in subsequent track processing, reducing interference from false points. Finally, after determining the credibility status of target points based on the actual confidence, track initiation clustering and track association operations are performed, which can effectively prevent false points from participating in track construction, effectively reduce the probability of track initiation misjudgment and association errors, and make track tracking more stable and accurate.

[0030] In practical implementation, it can be understood that the Peak-to-Noise Ratio (MNR) refers to the ratio of peak energy to angular spectrum noise floor; the Main-to-Side Lobe Ratio (MSR) refers to the ratio of peak energy to second-highest peak energy; the Main-to-Third / Fourth-Peak Ratio (MTFR) refers to the ratio of peak energy to the average of the third and fourth side lobes; and the target height can be obtained by first obtaining the corresponding pitch angle elev from the total angular spectrum, and then calculating the target height z based on the pitch angle elev and the distance range, with the specific calculation formula as follows: z=range*sin(elev) (Formula 1).

[0031] In an optional embodiment of the present invention, the step of calculating the actual confidence level corresponding to each target point based on the point information and the target parameters using a preset confidence level calculation model includes: Based on the Doppler velocity information corresponding to each target point, moving targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height corresponding to the moving target are input into the confidence calculation model to calculate the corresponding actual confidence level.

[0032] In this embodiment, moving targets are first screened out using the Doppler velocity information corresponding to the target points. For moving targets (such as vehicles and pedestrians), the moving targets have obvious Doppler velocities, and their radar echo signals are affected by the motion state. Therefore, the angle spectrum is prone to main lobe shift and side lobe interference enhancement. The main lobe ratio and the ratio of the main peak to the third and fourth peaks can effectively characterize the purity of the moving target's angle domain signal (distinguishing between real moving targets and motion clutter, such as fluttering leaves and splashing rain). Therefore, the above five parameters need to be included in the confidence score to ensure accurate determination of moving target points.

[0033] In an optional embodiment of the present invention, the step of calculating the actual confidence level corresponding to each target point based on the point information and the target parameters using a preset confidence level calculation model further includes: Based on the Doppler velocity information corresponding to each target point, static targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, and target height corresponding to the static target are input into the confidence calculation model to calculate the corresponding actual confidence level.

[0034] In this embodiment, stationary targets are first screened out using the Doppler velocity information corresponding to the target point. For stationary targets (such as stationary vehicles and roadside obstacles), there is no Doppler velocity, and their radar echo signals are stable. The main lobe and side lobe of the angle spectrum are regularly distributed, and the ratio of the main lobe to the side lobe and the ratio of the main peak to the third and fourth peaks are basically within a fixed and reasonable range. Usually, it is impossible to further distinguish stationary targets from static clutter (such as road bumps and guardrails) using these two parameters. Therefore, only the above three parameters are used for confidence scoring to reduce the amount of calculation.

[0035] In practical implementation, since subsequent trajectory planning is usually based on moving targets, this embodiment of the invention also delineates credible and uncredible points by scoring the confidence of static targets. This filters static clutter during trajectory planning, ensures the validity of the points, prevents false static targets from entering the trajectory processing flow, and reduces trajectory misjudgment. At the same time, it improves environmental perception and supports safety decisions. Finally, since some static targets may become moving targets later (such as a stationary vehicle starting), locking credible static targets in advance by scoring confidence and recording their position, altitude, and other information enables smooth switching between moving and static targets and trajectory association, avoiding trajectory breaks due to changes in target state and improving the continuity and stability of the overall trajectory planning.

[0036] In an optional embodiment of the present invention, the initial clustering of the trajectory based on the credible state corresponding to each target point specifically refers to: filtering out moving targets that meet the first preset condition and are marked as untrusted points, and performing initial clustering of the trajectory on the remaining moving targets. The first preset condition is that the absolute value of the corresponding velocity is greater than a preset velocity threshold (e.g., 1.5 m / s) and the corresponding distance is less than a preset distance threshold (e.g., 50 m). In this embodiment, for untrusted points that meet the first preset condition, their speed is too fast and their distance is relatively close to the vehicle. Targets that meet this condition are usually highly suspected clutter, outliers, multipath reflections, or interference points, and are not real moving targets. For other untrusted points that do not meet this condition, their speed is slow or their distance is far. Although they are untrusted points, they may still be real moving targets, so initial clustering of the trajectory can be performed to participate in the generation of the trajectory header.

[0037] In an optional embodiment of the present invention, trajectory association based on the trusted state corresponding to each target point specifically refers to: screening and analyzing each moving target and the target trajectory to be associated, associating all trusted points with credible points, and associating untrusted points that meet the second preset condition with the target trajectory that meets the third preset condition; the second preset condition includes that the absolute value of the corresponding speed is greater than a preset speed threshold (e.g., 1.5 m / s) and the corresponding distance is less than a preset distance threshold (e.g., 50 m); the third preset condition includes that the number of consecutive frames of the trajectory (i.e., the life value of the trajectory) is greater than a preset frame number threshold (e.g., 15 frames) and that it is a moving trajectory (e.g., the speed of the trajectory is greater than 1.5 m / s). In this embodiment, trusted points participate normally in track clustering, while untrusted points that meet the second preset condition are suspicious because they are too fast and relatively close to the vehicle. To ensure the utilization rate of the track, an additional third preset condition is determined. If the target track to be associated is a track with a number of consecutive frames greater than a preset frame threshold and is a moving track, that is, the target track is a mature moving track. A mature moving track will not cause obvious track deviation or splitting due to a suspicious track. Therefore, the suspicious track can be associated with such a target track.

[0038] In an optional embodiment of the present invention, the confidence calculation model is constructed based on a comprehensive discriminant function or a logical decision tree. In this embodiment, the confidence calculation model constructed using a comprehensive discriminant function or a logical decision tree can comprehensively analyze multi-dimensional point features to determine confidence, improving the accuracy and robustness of confidence assessment; it can flexibly adapt to target characteristics and environmental interference in different scenarios, enhancing the ability to distinguish between false points and clutter interference; wherein, the confidence level discrimination is achieved through a logical decision tree, with clear calculation logic, high computational efficiency, and ease of engineering implementation and real-time processing, meeting the real-time requirements of radar or target tracking systems.

[0039] In an optional embodiment of the present invention, the preprocessing includes range-dimensional FFT, Doppler-dimensional FFT, and angle-dimensional FFT; or, the preprocessing is beamforming. In this embodiment, FFT (Fast Fourier Transform) processing can be performed separately in the range dimension, Doppler dimension, and angle dimension, or beamforming can be used directly to form a 3D radar data cube. In specific implementations, different preprocessing methods can be flexibly selected to reduce computational power consumption.

[0040] To verify the effectiveness of the embodiments of the present invention, verification was conducted by collecting data on spurious moving target points in actual road test scenarios. Specific experimental scenarios are as follows: Figure 2 As shown, the vehicle-mounted millimeter-wave radar 1 is installed at the right front corner of the car. The installation parameters are: longitudinal distance from the center of the rear axle 3.75m, lateral distance from the center of the rear axle 0.8m, height from the ground 0.55m, and installation angle -45°.

[0041] Additionally, around the vehicle: strong reflective targets are the sheet metal wall and scaffolding on the right. The radar emits electromagnetic waves and generates multipath reflections from strong reflective targets such as the sheet metal wall, creating false detection points in the open area directly in front. These points are continuously detected for multiple frames, thus generating false flight tracks.

[0042] Corresponding step S1: Generate a 3D radar data cube; Corresponding to step S2: Detect target points and calculate the point information for each target point, including the distance range, Doppler velocity information radvel, and signal-to-noise ratio snr; Corresponding to step S3: Extract the angular spectrum of each target point, and calculate the target parameters corresponding to each target point. The target parameters include: Peak Noise Floor (MNR), Main Lobe / Side Lobe Ratio (MSR), Main Peak / Triple Peak / Fourth Peak Ratio (MTFR), and target height z. For false targets, the angular spectrum distribution is as follows: Figure 3 As shown, its angle spectrum has poor stability and large overall fluctuations; the actual target angle spectrum distribution is as follows. Figure 4 The comparison of angular spectrum parameters is shown in the table below. The comparison results show that the real target is better than the false target in terms of peak-to-noise ratio, main-to-side lobe ratio, and the ratio of the average of the main peak to the third and fourth peaks.

[0043] Corresponding to step S4: Calculate the actual confidence level corresponding to each target point, wherein, firstly, the Doppler velocity information corresponding to each target point is used to filter out moving targets and stationary targets. In this embodiment of the invention, different upper and lower thresholds are set for signal-to-noise ratio, peak-to-noise ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height. Signal-to-noise ratio of target point trace (lower limit 10dB, upper limit 30dB); Peak noise floor value of target point (lower limit 9dB, upper limit 15dB); The ratio of main lobe to side lobe of the target point (lower limit 2dB, upper limit 6dB). The ratio of the main peak to the third and fourth peaks of the target point (i.e., the ratio of the main peak to the average of the third and fourth peaks) (lower limit 3dB, upper limit 8dB); The height value of the target point; the greater the height, the lower the confidence level (0~5m). For each moving target, the actual confidence level is calculated using five parameters: signal-to-noise ratio (SNR), peak-to-noise ratio (PNR), main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height. For stationary targets, the actual confidence level is calculated using three parameters: SNR, PNR, and target height. The specific calculation method is as follows: The corresponding parameters are mapped to a range of 0 to 1, with 0 below the lower limit and 1 above the upper limit. Then, the average value of each parameter is calculated to obtain a coefficient between 0 and 1. This coefficient is then multiplied by 100 to obtain the actual confidence level. Corresponding to step S5: Based on the actual confidence level of each target point, distinguish its corresponding confidence state. If the actual confidence level is greater than the preset confidence threshold (e.g., 45), then determine that the corresponding target point is a reliable point (confidence flag is 1), otherwise it is an unreliable point (confidence flag is 0). Then, at the start of the track, filter moving points: if the absolute value of the point speed is greater than 1.5m / s and the distance is less than 50m, and the confidence flag is 0, then the point cannot be clustered at the start of the track. When associating tracks, for points with an absolute value of the point speed greater than 1.5m / s and a distance less than 50m, and a confidence flag of 0, determine that the number of consecutive frames of the associated track is greater than 15 frames, and the track is a moving track. If so, association is allowed; otherwise, association is not allowed.

[0044] Finally, the trajectory planning before false target filtering is as follows: Figure 5 As shown, a false track with track number 26 was generated. The track planning after false target filtering using the embodiment of the present invention is as follows: Figure 6 As shown, the scene is empty and no false moving targets are generated.

[0045] On the other hand, such as Figure 7 As shown, this embodiment of the invention further provides a trajectory planning device 3 based on vehicle-mounted millimeter-wave radar, which is connected to vehicle-mounted millimeter-wave radar 1. It includes a processor 30, a memory 32, and a computer program stored in the memory 32 and configured to be executed by the processor 30. When the processor 30 executes the computer program, it implements the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any of the above embodiments.

[0046] For example, the computer program can be divided into one or more modules / units, which are stored in the memory 32 and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the vehicle-mounted millimeter-wave radar-based trajectory planning device 3. For example, the computer program can be divided into... Figure 8 The functional modules in the vehicle-mounted millimeter-wave radar-based trajectory planning device 3 include the signal acquisition and preprocessing module 41, the target detection module 42, the angle spectrum parameter calculation module 43, the confidence calculation module 44, and the point marking and trajectory planning module 45, which respectively perform the above steps S1-S5.

[0047] The trajectory planning device 3 based on vehicle-mounted millimeter-wave radar can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. The trajectory planning device 3 may include, but is not limited to, a processor 30 and a memory 32. Those skilled in the art will understand that the schematic diagram is merely an example of the trajectory planning device 3 based on vehicle-mounted millimeter-wave radar and does not constitute a limitation on the device. It may include more or fewer components than shown, or combine certain components, or use different components. For example, the trajectory planning device 3 based on vehicle-mounted millimeter-wave radar may also include input / output devices, network access devices, buses, etc.

[0048] The processor 30 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 30 is the control center of the vehicle-mounted millimeter-wave radar trajectory planning device 3, connecting all parts of the vehicle-mounted millimeter-wave radar-based trajectory planning device 3 via various interfaces and lines.

[0049] The memory 32 can be used to store the computer program and / or modules. The processor 30 implements various functions of the vehicle-mounted millimeter-wave radar trajectory planning device 3 by running or executing the computer program and / or modules stored in the memory 32 and calling the data stored in the memory 32. The memory 32 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as image recognition function, image overlay function, etc.), etc.; the data storage area may store data created according to the use of the device (such as image data, etc.). In addition, the memory 32 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0050] If the functions described in the embodiments of the present invention are implemented in the form of software functional modules or units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the embodiments of the present invention can implement all or part of the processes in the methods described above, or they can be accomplished by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by the processor 30, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0051] In another aspect, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any of the above embodiments.

[0052] In another aspect, embodiments of the present invention provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any of the above embodiments.

[0053] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0054] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the scope of protection of the present invention.

Claims

1. A trajectory planning method based on vehicle-mounted millimeter-wave radar, characterized in that, The method includes the following steps: The echo signal received by the vehicle-mounted millimeter-wave radar is acquired, and the echo signal is preprocessed to generate a three-dimensional radar data cube that reflects the three dimensions of distance, Doppler, and angle. In the three-dimensional radar data cube, target detection is performed from two dimensions: range and Doppler to filter out all target points. The point information corresponding to each target point is saved, and the point information includes range, Doppler velocity information and signal-to-noise ratio. The angle spectrum of each target point is extracted from the three-dimensional radar data cube. Based on the angle spectrum and the point information, the target parameters corresponding to each target point are calculated. The target parameters include: peak noise floor ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height. The actual confidence level corresponding to each target point is calculated using a preset confidence level calculation model based on the point information and the target parameters; and The credibility status of each target point is marked based on the actual confidence level corresponding to each target point, and track initial clustering and track association are performed based on the credibility status corresponding to each target point. The credibility status includes credibility points and untrustworthy points.

2. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 1, characterized in that, The step of calculating the actual confidence level of each target point based on the point information and the target parameters using a preset confidence level calculation model includes: Based on the Doppler velocity information corresponding to each target point, moving targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, main lobe-to-side lobe ratio, main peak-to-third-to-fourth-peak ratio, and target height corresponding to the moving target are input into the confidence calculation model to calculate the corresponding actual confidence level.

3. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 1 or 2, characterized in that, The step of calculating the actual confidence level of each target point based on the point information and the target parameters using a preset confidence level calculation model further includes: Based on the Doppler velocity information corresponding to each target point, static targets are selected from all target points; and The signal-to-noise ratio, peak-to-noise ratio, and target height corresponding to the static target are input into the confidence calculation model to calculate the corresponding actual confidence level.

4. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 2, characterized in that, The initial clustering of the trajectory based on the trustworthy state corresponding to each target point specifically refers to: filtering out moving targets that meet the first preset condition and are marked as untrustworthy points, and performing initial clustering of the trajectory on the remaining moving targets. The first preset condition is: the absolute value of the corresponding speed is greater than a preset speed threshold and the corresponding distance is less than a preset distance threshold.

5. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 2, characterized in that, The specific meaning of trajectory association based on the credible state corresponding to each target point is: to screen and analyze each moving target and the target trajectory to be associated, to associate all credible points with the trajectory, and to associate untrusted points that meet the second preset condition with the target trajectory that meets the third preset condition. The second preset condition includes that the absolute value of the corresponding speed is greater than a preset speed threshold and the corresponding distance is less than a preset distance threshold; the third preset condition includes that the number of consecutive frames of the track is greater than a preset frame number threshold and that it is a moving track.

6. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 1, characterized in that, The confidence calculation model is constructed based on a comprehensive discriminant function or a logical decision tree.

7. The trajectory planning method based on vehicle-mounted millimeter-wave radar as described in claim 1, characterized in that, The preprocessing includes distance-dimensional FFT, Doppler-dimensional FFT, and angle-dimensional FFT; or, the preprocessing is beamforming.

8. A trajectory planning device for a vehicle-mounted millimeter-wave radar, connected to the vehicle-mounted millimeter-wave radar, characterized in that, The device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the trajectory planning method based on vehicle-mounted millimeter-wave radar as described in any one of claims 1-7.

Images