An object detection method, system and vehicle based on millimeter-wave radar

Abstract

An embodiment of the present invention provides an object detection method, system and vehicle based on an in-vehicle millimeter-wave radar. Through the present invention, a plurality of original point information of a detection target is collected, and the original point information includes original three-dimensional coordinate information and original velocity information; the plurality of original points are divided into multiple groups according to the original three-dimensional coordinate information and the original velocity information, and each group is represented by a pixel point, and the pixel point includes pixel three-dimensional coordinate information and pixel velocity information; a detection target image is formed according to the pixel points. The resolution sizes of the radar in the horizontal and pitch directions are obtained, and the problem in the prior art that the collected original points cannot be accurately distinguished according to the height information, and the original point distribution cannot accurately represent the contour of the target object when processing the original point information, resulting in errors in target object recognition, is solved, and the accuracy of the millimeter-wave radar in recognizing the target object can be improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of on-vehicle millimeter-wave radar detection. Specifically, the present invention relates to an object detection method, system and vehicle based on millimeter-wave radar. Background Art

[0002] With the further popularization of autonomous driving, the accuracy of millimeter-wave radar in detecting three-dimensional target objects has received more and more attention.

[0003] Traditional radars are unable to accurately distinguish the collected original points according to height information during the recognition of target objects due to the lack of height information. Therefore, when processing the original point information, the distribution of the original points cannot accurately represent the contour of the target object, resulting in errors in the recognition of the target object.

[0004] In the prior art, there is also a method of using a weighted method of a camera and a millimeter-wave radar to identify target objects. However, due to the relatively low resolution of the millimeter-wave radar in the pitch angle, during data fusion, due to inaccurate height information, it is impossible to classify the pixel point blocks of the millimeter-wave radar. Image recognition mainly relies on the recognition ability of the camera itself, and the information recognized by the millimeter-wave radar is used as a supplement for weighted fusion, resulting in a high cost. Summary of the Invention

[0005] Embodiments of the present invention provide an object detection method and device based on millimeter-wave radar to at least solve the problem that the distribution of original points cannot accurately represent the contour of the target object when the millimeter-wave radar processes the original point information in the related art, resulting in errors in the recognition of the target object.

[0006] According to an embodiment of the present invention, there is provided an object detection method based on millimeter-wave radar, including: collecting a plurality of pieces of original point information of a detection target, where the original point information includes original three-dimensional coordinate information and original velocity information; dividing the plurality of original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information of the original point information, each group is represented by a pixel point, and the pixel point includes pixel three-dimensional coordinate information and pixel velocity information; forming a detection target image according to the pixel points.

[0007] Furthermore, a sparse antenna array arranged horizontally and vertically is used to collect the original point information of the detection target.

[0008] Further, dividing multiple original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information of the original point information further includes: when the difference in the original three-dimensional coordinate information of the multiple original points is less than a first threshold and the difference in the original velocity information of the multiple original points is less than a second threshold, representing the multiple original points by one pixel point; the pixel three-dimensional coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of the multiple original points; the pixel velocity information of the center point of the pixel point is the average value or weighted average value of the original velocity information of the multiple original points.

[0009] Further, within a preset time, collecting the information of the pixel points at a preset position at a preset period; preprocessing the pixel points to obtain stable pixel points, and after weighted superposition of the pixel three-dimensional information of the stable pixel points, forming the detection target image.

[0010] In an exemplary embodiment, when the appearance probability of the pixel point is greater than a preset stability threshold, taking the pixel point as a stable pixel point.

[0011] In an exemplary embodiment, when the pixel point is not collected more than a preset number of times, taking the pixel point as an unstable pixel point; after removing the unstable pixel points, obtaining stable pixel points.

[0012] Further, collecting multiple detection target images through multiple cycles; obtaining the feature information of the multiple detection target images through a deep learning algorithm; determining the type of the detection target according to the feature information.

[0013] Further, providing different warning information according to the type of the detection target.

[0014] According to another embodiment of the present invention, there is provided an object detection system based on a millimeter-wave radar, including: an information acquisition module for collecting multiple original point information of a detection target, the original point information including original three-dimensional coordinate information and original velocity information; a pixel generation module for dividing multiple original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information of the original point information, each group being represented by one pixel point, and the pixel point including pixel three-dimensional coordinate information and pixel velocity information; an image processing module for forming a detection target image according to the pixel points.

[0015] Further, the information acquisition module is a sparse antenna array arranged horizontally and vertically.

[0016] Further, when the difference in the original three-dimensional coordinate information of multiple original points is less than the first threshold and the difference in the original velocity information of multiple original points is less than the second threshold, the pixel generation module is further configured to represent the multiple original points by one pixel point; the pixel three-dimensional coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of the multiple original points; the pixel velocity information of the center point of the pixel point is the average value or weighted average value of the original velocity information of the multiple original points.

[0017] Further, the information acquisition module is further configured to acquire the information of the pixel points at the preset position at a preset period within a preset time; the system further includes a pixel point preprocessing module configured to preprocess the pixel points to obtain stable pixel points.

[0018] The image processing module is further configured to form the detection target image after weighted superposition of the pixel three-dimensional information of the stable pixel points.

[0019] In an exemplary embodiment, when the appearance probability of the pixel point is greater than a preset stable threshold, the pixel point preprocessing module is further configured to use the pixel point as a stable pixel point.

[0020] In an exemplary embodiment, when a pixel point is not acquired for more than a preset number of times, the pixel point preprocessing module is further configured to use the pixel point as an unstable pixel point, and after removing the unstable pixel points, stable pixel points are obtained.

[0021] Further, the information acquisition module is further configured to acquire multiple detection target images through multiple periods; the system further includes a deep learning module configured to obtain the feature information of the multiple detection target images through a deep learning algorithm; the system further includes a target type determination module configured to determine the type of the detection target according to the feature information.

[0022] Further, a warning module is further included, configured to provide different warning information according to the type of the detection target.

[0023] Through the present invention, since the resolution sizes of the radar in the horizontal and pitch directions are obtained, and a certain redundant size is added as the minimum pixel point for radar recognition, and this pixel point represents the radar raw data points with similar attributes integrated into one minimum pixel point. The original data is classified and represented in the form of pixel points one by one. Then the pixel points in multiple periods are superimposed to obtain a stable target object image. It solves the problem in the prior art that the collected original points cannot be accurately distinguished according to the height information, and the distribution of the original points cannot accurately represent the contour of the target object when processing the original point information, resulting in errors in the recognition of the target object, and can improve the accuracy of the millimeter-wave radar in recognizing the target object.

[0024] Using a sparse antenna design both horizontally and vertically can simulate the detection resolution of more antennas with a small number of antennas, thereby effectively reducing the design cost of the radar. At the same time, since more antennas are simulated, the detection performance that can only be achieved by more antennas is obtained, improving the resolution of detecting targets in the horizontal and pitch directions of the radar. The target object can be recognized through a deep learning algorithm, which is beneficial to improving the recognition of target objects such as pedestrians, motor vehicles, non-motor vehicles, and road environments by the millimeter-wave radar, effectively reducing the false alarms of the millimeter-wave radar, and improving the early warning ability of the millimeter-wave radar. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0026] Figure 1 It is a sparse antenna structure diagram of the target method acquisition module based on the millimeter-wave radar according to an embodiment of the present invention;

[0027] Figure 2 It is a flowchart of the target detection method based on the millimeter-wave radar according to an embodiment of the present invention;

[0028] Figure 3 It is a flowchart of the target detection method based on the millimeter-wave radar according to another embodiment of the present invention;

[0029] Figure 4 It is a schematic diagram of pixel point superposition on a two-dimensional plane at different times according to an embodiment of the present invention.

[0030] Figure 5 It is a schematic structural diagram of the target detection system based on the millimeter-wave radar according to an embodiment of the present invention;

[0031] Figure 6 It is a flowchart of the schematic structural diagram of the millimeter-wave radar target detection system according to another embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] The embodiments of the present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0033] In the description of the present application, it should be noted that the orientation or positional relationship indicated by terms such as "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present application. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0034] In the description of the present application, it should be noted that unless otherwise clearly specified and limited, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and can also be the communication inside two elements. It can be a wireless connection or a wired connection. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific situations.

[0035] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only for the convenience of the description of the present invention, and they have no specific meaning themselves. Therefore, "module" and "component" can be used interchangeably.

[0036] In addition, the technical features involved in different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0037] Embodiment 1

[0038] The method embodiment provided in the embodiment of the present application can be executed in an in-vehicle millimeter-wave radar target recognition system. In this embodiment, a target recognition method running on an in-vehicle millimeter-wave radar system is provided. Figure 1 It is a flowchart of the target recognition method according to the embodiment of the present invention, as Figure 1 shown, and this process includes the following steps:

[0039] Step S101, collect multiple pieces of original point information of the detection target, and the original point information includes original three-dimensional coordinate information and original velocity information;

[0040] Specifically, in the antenna design of this embodiment, a sparse antenna layout is used in both the horizontal and vertical directions of the antenna, as Figure 1 shown.

[0041] Among them, solid circles represent actual antennas, RX1, RX2, RX5, and RX7 are all receiving antennas, TX1, TX2, and TX3 are all transmitting antennas, and they are set at regular intervals in both the horizontal and vertical directions. Hollow circles represent simulated virtual antennas.

[0042] By using a sparse antenna design for the radar antenna in both the horizontal and vertical directions, the mutual spacing difference between the antennas in the horizontal and pitch directions of the radar can provide calculations of different angular resolutions and detection accuracies, thereby improving the detection resolution of the radar in the horizontal and pitch directions.

[0043] Specifically, the vehicle-mounted millimeter-wave radar uses sparse antennas set in both the horizontal and vertical directions to collect multiple pieces of original point information of the detection target, P1(x1,y1,z1,v1), P2(x2,y2,z2,v2), P3(x3,y3,z3,v3), P4(x4,y4,z4,v4), P5(x5,y5,z5,v5)……Pn(xn,yn,zn,vn)

[0044] Among them, xn represents the original horizontal direction coordinate information of the original point, yn represents the vertical direction coordinate information of the original point, zn represents the original height information of the original point, and vn represents the speed information of the original point.

[0045] Step S102, divide multiple original points into multiple groups according to the original three-dimensional coordinate information and original speed information of the original point information, and each group is represented by a pixel point, and the pixel point contains pixel three-dimensional coordinate information and pixel speed information;

[0046] According to the three-dimensional coordinate information of the original point, the distance information of the original point information can be obtained. The distance between P1 and the detection system is d1, the distance between P2 and the detection system is d2, the distance between P3 and the detection system is d3, the distance between P4 and the detection system is d4, the distance between P5 and the detection system is d5, …… the distance between Pn and the detection system is dn.

[0047] Among them, when the difference between the three-dimensional coordinate information of P1, P2, and P3 is less than the first threshold, and the difference between the speeds v1, v2, and v3 is less than the second threshold, P1, P2, and P3 are classified as the minimum pixel point M1. This minimum target pixel point M1 contains the three-dimensional coordinate information of the speed, and the detection target image is formed according to the pixel point.

[0048] When the difference between the three-dimensional coordinate information of P4 and P5 is less than the first threshold, and the differences between the speeds v1, v2, and v3 are all less than the second threshold, P1, P2, and P3 are classified as the minimum pixel point M1. The minimum target pixel point M1 includes pixel three-dimensional coordinate information and pixel speed information, and the center point coordinate of the pixel point M1 is the average value or weighted average value of the three-dimensional coordinates of P1, P2, and P3.

[0049] Step S105: Form a detection target image according to the pixel points.

[0050] Through the above steps, using a MIMO sparse antenna and leaving a certain distance in both the horizontal and vertical directions of the antenna. In this way, when receiving radar echoes, there will be two different phase differences in the horizontal and height directions, and phase differences in the horizontal and pitch directions. Therefore, high-resolution detection can be obtained in the horizontal and height directions. At the same time, a small number of antennas can be used in the horizontal and vertical directions to simulate the high-resolution detection of more antennas, which can effectively reduce the design cost of the antenna and the resource occupancy cost of signal detection and processing.

[0051] In this embodiment, representing the original point information with similar attributes in space as pixel points can effectively group the original data points with similar radar information onto the same pixel block for representation. This is beneficial to reducing the data operation of the radar, effectively reducing signal interference, and at the same time providing pixel information for image recognition running on the millimeter-wave radar.

[0052] Embodiment 2

[0053] In this embodiment, a target recognition method running on a vehicle-mounted millimeter-wave radar system is provided. Figure 3 It is a flowchart of the target recognition method according to the embodiment of the present invention, as Figure 3 shown. This process includes the following steps:

[0054] Step S201: Collect multiple pieces of original point information of the detection target.

[0055] Step S202: When the difference between the original three-dimensional coordinate information of multiple original points is less than the first threshold, and the difference between the original speed information of multiple original points is less than the second threshold, represent multiple original points with one pixel point.

[0056] Step S203: The pixel three-dimensional coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of multiple original points;

[0057] Step S204: The pixel speed information of the center point of the pixel point is the average value or weighted average value of the original speeds of multiple original points.

[0058] Step S205: Collect the information of the pixel points at the preset position within the preset time at the preset period;

[0059] Step S206: Preprocess the pixel points to obtain stable pixel points;

[0060] Step S207: After weighted superposition of the three-dimensional pixel information of the stable pixel points, form a detection target image.

[0061] Specifically, collect the information of the pixel points at the (X, Y, Z) position at the preset period, count the occurrence times of the pixel points. When the occurrence probability of the pixel points is less than 10%, or the pixel points at this position have not been collected for more than 10 times, then regard this pixel point as an unstable pixel point.

[0062] After eliminating all unstable pixel points, superimpose the remaining pixel points according to the three-dimensional information of the pixels to form a detection target image. The occurrence probability of the pixel points is used as a performance index for the reliability of the pixel points.

[0063] Or within ΔT, collect the information of the pixel points at the (X, Y, Z) position, count the occurrence times of the pixel points. When the occurrence probability of the pixel points is greater than 75%, regard this pixel point as a stable pixel point, perform weighted superposition on the three-dimensional information of the stable pixel points to form a detection target image. The occurrence probability of the pixel points is used as a performance index for the reliability of the pixel points.

[0064] The above is only an example that can be implemented. The judgment thresholds for stable pixel points and unstable pixel points can both be set according to different recognition requirements.

[0065] For the smallest target pixel points detected in different periods, according to the acceleration, time interval △t and speed information of the pixel points, the three-dimensional coordinates of different pixel points collected in the latest 100 cycle times can be calculated and superimposed on the latest cycle. Cumulatively weight the smallest pixel points at each position according to the detection times. The higher the frequency of the same pixel point appears, the higher the reliability of the information contained in this point. At the same time, some unreliable target points can be filtered out according to the frequency of the pixel points appearing. At the same time, if the pixel point update has not been received for more than n times, then this pixel point can be eliminated. Thus, a stable target object contour can be obtained. The above 100 cycles are only an example that can be implemented and can be sampled according to the debugging requirements. The schematic diagram of superposition on the two-dimensional plane is as Figure 3 shown.

[0066] By optimizing the pixel points, retaining the stable pixel points for superposition to form a target detection image, the problem of inaccurate detection imaging caused by unstable pixel points can be solved, the detection imaging accuracy can be improved, and the recognition accuracy can be improved.

[0067] This embodiment further includes:

[0068] Step S208: Collect multiple detection target images through multiple cycles;

[0069] Step S209: Obtain the feature information of the detection target image through a deep learning algorithm;

[0070] Step S210: Determine the type of the detection target according to the feature information.

[0071] In this embodiment, after collecting 100 cycles, multiple detection target images are formed. The feature information of the target image is obtained through a deep learning algorithm, and the type of the detection target is determined according to the feature information. A matching list of target types and feature information can be preset inside the processor. When the preset features meet the corresponding conditions, the category of the detected object is judged. Different targets such as pedestrians, motor vehicles, non-motor vehicles, and roadblocks are distinguished.

[0072] Through the superposition of stable pixel points over multiple cycles, the feature points of the target object can be recognized. The image superimposed by the radar is recognized through a deep learning algorithm. At the same time, in the deep learning algorithm, the input of the distance, height, and speed information of the pixel points is increased, which is beneficial to improving the recognition accuracy of the deep learning algorithm.

[0073] This embodiment further includes:

[0074] Step S211: Provide different warning information according to the type of the detection target.

[0075] According to the recognized three-dimensional information of the target object, different warnings are provided for the vehicle. The three-dimensional spatial positions and movement speeds of target objects such as pedestrians, motor vehicles, non-motor vehicles, and road environments can be recognized. It is beneficial to improve the recognition accuracy and recognition speed of the millimeter-wave radar for target objects such as pedestrians, motor vehicles, non-motor vehicles, and road environments, effectively reduce the false alarms of the millimeter-wave radar, and improve the warning ability of the millimeter-wave radar.

[0076] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented through hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes a contribution to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc), and includes several instructions for causing a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of the various embodiments of the present invention.

[0077] Embodiment III

[0078] In this embodiment, a vehicle-mounted millimeter-wave radar target object detection system is also provided. This system is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and / or hardware that realizes a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0079] Figure 2 is a structural block diagram of a vehicle-mounted millimeter-wave radar target object detection system according to an embodiment of the present invention. As Figure 2 shown, the device includes an information acquisition module, a pixel generation module, and an image processing module.

[0080] Among them, the information acquisition module can be configured as an antenna array arranged horizontally and vertically, and is used to acquire a plurality of original point information of the detection target. The original point information includes original three-dimensional coordinate information and original velocity information; the pixel generation module is used to divide a plurality of original points into multiple groups according to the original three-dimensional coordinate information and original velocity information of the original point information, and each group is represented by a pixel point. The pixel point includes pixel three-dimensional coordinate information and pixel velocity information; the image processing module forms a detection target image according to the pixel points.

[0081] The pixel generation module is further used to represent a plurality of original points by one pixel point when the difference in the original three-dimensional coordinate information of the plurality of original points is less than a first threshold and the difference in the original velocity information of the plurality of original points is less than a second threshold; the pixel three-dimensional coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of the plurality of original points; the pixel velocity information of the center point of the pixel point is the average value or weighted average value of the original velocity information of the plurality of original points;

[0082] Figure 3 is a structural block diagram of a vehicle-mounted millimeter-wave radar target object detection system according to an embodiment of the present invention. As Figure 3 shown, in addition to including Figure 2 all the modules shown, the device further includes a pixel point preprocessing module, a deep learning module, a target type judgment module, and an early warning module.

[0083] Furthermore, the information acquisition module is further used to acquire the information of the pixel points at a preset position within a preset time at a preset period; the system further includes a pixel point preprocessing module, which is used to preprocess the pixel points to obtain stable pixel points; the image processing module is further used to perform weighted superposition on the pixel three-dimensional information of the stable pixel points to form the detection target image.

[0084] In an exemplary embodiment, the pixel preprocessing module is further configured to use the pixel as a stable pixel when the occurrence frequency of the pixel is greater than a preset stable threshold;

[0085] In an exemplary embodiment, the pixel preprocessing module is further configured to use the pixel as an unstable pixel when the pixel has not been collected for more than a preset number of times. After removing the unstable pixels, stable pixels are obtained.

[0086] Furthermore, the information acquisition module is further configured to collect multiple detection target images through multiple cycles; the system further includes a deep learning module, which is configured to obtain the feature information of the multiple detection target images through a deep learning algorithm; the system further includes a target type determination module, which is configured to determine the type of the detection target according to the feature information.

[0087] Furthermore, a warning module is further included, which is configured to provide different warning information according to the type of the detection target.

[0088] The system is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated.

[0089] It should be noted that the above-mentioned respective modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited thereto: the above-mentioned modules are all located in the same processor; or, the above-mentioned respective modules are separately located in different processors in any combination form.

[0090] An embodiment of the present invention further provides a vehicle, including the above-mentioned in-vehicle millimeter-wave radar target object detection system.

[0091] An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, and wherein the computer program is configured to execute the steps in any one of the above method embodiments when running.

[0092] In an exemplary embodiment, the above-mentioned computer-readable storage medium may include, but is not limited to: USB flash drives, read-only memories (ROM for short), random access memories (RAM for short), mobile hard disks, magnetic disks or optical discs, etc., various media that can store computer programs.

[0093] An embodiment of the present invention further provides an electronic device, including a memory and a processor, 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.

[0094] In an exemplary embodiment, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0095] Specific examples in this embodiment may refer to the examples described in the above embodiments and exemplary embodiments, and will not be elaborated here.

[0096] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general-purpose 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 program codes executable by the computing device. Thus, 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 sequence different from that here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0097] The above is only the preferred embodiment of the present invention and is not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An object detection method based on millimeter-wave radar, characterized in that, Including: Collecting a plurality of original point information of a detection target, where the original point information includes original three-dimensional coordinate information and original velocity information; Dividing the plurality of original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information, with each group represented by a pixel point, and the pixel point includes pixel three-dimensional coordinate information and pixel velocity information; Forming the detection target image according to the pixel points; Collecting information of pixel points at a preset position within a preset time at a preset period; Preprocessing the pixel points to obtain stable pixel points; After weighted superposition of the pixel three-dimensional information of the stable pixel points, forming the detection target image.

2. The method according to claim 1, wherein Further including: Using sparse antenna arrays arranged horizontally and vertically to collect a plurality of original point information for detecting the target.

3. The method according to claim 1, wherein The step of dividing the plurality of original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information of the original point information further includes: When the difference in the original three-dimensional coordinate information of the plurality of original points is less than a first threshold and the difference in the original velocity information of the plurality of original points is less than a second threshold, representing the plurality of original points by a pixel point; The pixel three-dimensional coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of the plurality of original points; The pixel velocity information of the center point of the pixel point is the average value or weighted average value of the original velocity information of the plurality of original points.

4. The method according to claim 1, characterized in that, The step of preprocessing the pixel points according to the frequency of occurrence of the pixel points to obtain stable pixel points further includes: Calculating the frequency of occurrence of the pixel points; When the frequency of occurrence of the pixel point is greater than a preset stability threshold, taking the pixel point as a stable pixel point.

5. The method according to claim 1, wherein The step of preprocessing the pixel points according to the frequency of occurrence of the pixel points to obtain stable pixel points further includes: When the pixel point is not collected for more than a preset number of times, taking the pixel point as an unstable pixel point; After removing the unstable pixel points, obtaining stable pixel points.

6. The method according to claim 4 or 5, characterized in that Further including; Collecting multiple detection target images through multiple cycles; Obtaining the feature information of the detection target image through a deep learning algorithm; Determining the type of the detection target according to the feature information.

7. The method according to claim 6, wherein Further including: Providing different warning information according to the type of the detection target.

8. A target object detection system based on a millimeter-wave radar, characterized in that, Including: An information collection module for collecting a plurality of original point information of a detection target, where the original point information includes original three-dimensional coordinate information and original velocity information; A pixel generation module for dividing the plurality of original points into multiple groups according to the original three-dimensional coordinate information and the original velocity information of the original point information, with each group represented by a pixel point, and the pixel point includes pixel three-dimensional coordinate information and pixel velocity information; An image processing module for forming the detection target image according to the pixel points; The information collection module is further used for collecting information of pixel points at a preset position within a preset time at a preset period; Further including a pixel point preprocessing module for preprocessing the pixel points to obtain stable pixel points; The image processing module is further configured to perform weighted superposition on the three-dimensional pixel information of the stable pixel points to form the detected target image.

9. The system according to claim 8, wherein It further includes The information acquisition module is a sparse antenna array arranged horizontally and vertically.

10. The system according to claim 9, wherein It further includes: The pixel generation module is further configured to represent the multiple original points by one pixel point when the difference in the original three-dimensional coordinate information of the multiple original points is less than a first threshold and the difference in the original velocity information of the multiple original points is less than a second threshold. The three-dimensional pixel coordinate information of the center point of the pixel point is the average value or weighted average value of the original three-dimensional coordinate information of the multiple original points. The pixel velocity information of the center point of the pixel point is the average value or weighted average value of the original velocity information of the multiple original points.

11. The system according to claim 8, wherein It further includes: The pixel point preprocessing module is further configured to use the pixel point as a stable pixel point when the occurrence frequency of the pixel point is greater than a preset stability threshold.

12. The system according to claim 8, wherein It further includes: The pixel point preprocessing module is further configured to use the pixel point as an unstable pixel point when the pixel point has not been collected for more than a preset number of times, and obtain stable pixel points after removing the unstable pixel points.

13. The system according to claim 11 or 12, wherein It further includes: The information acquisition module is further configured to collect multiple detected target images through multiple cycles. The system further includes a deep learning module for obtaining the feature information of the multiple detected target images through a deep learning algorithm. The system further includes a target type determination module for determining the type of the detected target according to the feature information.

14. The system according to claim 13, wherein It further includes: An early warning module for providing different early warning information according to the type of the detected target.

15. A vehicle, characterized in that, It includes a target detection system based on a millimeter-wave radar according to any one of claims 8-14.

Images