Radar-based kick signal recognition method, radar, storage medium and automobile

Abstract

The present invention provides a radar-based kick signal recognition method, radar, storage medium and automobile, the method comprising: obtaining point cloud information of echo signals; for each frame of point cloud, calculating the feature mean of the points in the frame of point cloud whose distance dimension features meet preset conditions, and storing it in a feature mean matrix; dividing the detection area of ​​the radar in the angle dimension into grids, and determining the centroid position according to the angle range of each grid and the angle dimension feature mean of each frame of point cloud; if the centroid position is located in the detection area and the speed dimension feature mean of each frame of point cloud is located in the preset speed range, extracting the distance dimension feature mean of each frame of point cloud in the feature mean matrix to obtain a distance dimension array; judging whether a kick target is detected according to the distance dimension array. Through the above scheme, the present application can comprehensively identify kick signals according to the angle dimension features, speed dimension features and distance dimension features of the point cloud information, avoid false alarms, and improve the recognition accuracy of kick signals.

Description

Technical Field

[0001] The present invention relates to the field of radar technology, and in particular to a radar-based kick signal recognition method, a radar, a storage medium and a car. Background Art

[0002] The method of opening the car door by kicking it can open the car door conveniently and quickly, which meets people's pursuit of vehicle intelligence and convenience and is favored by people.

[0003] The existing technology mostly uses the method of extracting voltage difference to identify the correct kicking action. This method cannot deal with false alarms caused by car washing tools or wiping the car, and is also prone to missed kicks and multiple kicks. Summary of the invention

[0004] In view of this, the present invention provides a radar-based kick signal recognition method, radar, storage medium and automobile, which can solve the problem of poor accuracy in kick signal recognition.

[0005] In a first aspect, an embodiment of the present invention provides a radar-based kick signal recognition method, comprising:

[0006] Acquire point cloud information of the echo signal detected by the radar; the point cloud information includes distance dimension features, angle dimension features and speed dimension features;

[0007] For each frame of point cloud, calculate the feature mean of the points in the frame of point cloud whose distance dimension features meet the preset conditions, and store the feature mean corresponding to the frame of point cloud into the feature mean matrix;

[0008] Divide the detection area of ​​the radar in the angle dimension into grids, determine the angle range of each grid, and determine the centroid position of the point cloud trajectory according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix;

[0009] If the centroid position is located within the detection area and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located within a preset speed range, then extracting the distance dimension feature mean of each frame point cloud in the feature mean matrix to obtain a distance dimension array;

[0010] It is determined whether a kicking target is detected according to the distance dimension array.

[0011] In a second aspect, an embodiment of the present invention provides a radar, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method described in any possible implementation manner of the first aspect are implemented.

[0012] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in any possible implementation of the first aspect above are implemented.

[0013] In a fourth aspect, an embodiment of the present invention provides a car, comprising the radar as described in the third aspect above.

[0014] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0015] The embodiment of the present invention first obtains the point cloud information of the echo signal detected by the radar; for each frame of the point cloud, the feature mean of the point whose distance dimension feature meets the preset conditions in the frame of the point cloud is calculated, and the feature mean corresponding to the frame of the point cloud is stored in the feature mean matrix; then the detection area of ​​the radar in the angle dimension is gridded, the angle range of each grid is determined, and the centroid position of the point cloud trajectory is determined according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix; if the centroid position is located in the detection area and the speed dimension feature mean of each frame of the point cloud in the feature mean matrix is ​​located in the preset speed range, the distance dimension feature mean of each frame of the point cloud in the feature mean matrix is ​​extracted to obtain the distance dimension array; finally, according to the distance dimension array, it is determined whether the kicking target is detected. Through the above scheme, the present embodiment can comprehensively identify the kicking signal according to the angle dimension feature, speed dimension feature and distance dimension feature of the point cloud information detected by the radar, avoid false alarms, and improve the recognition accuracy of the kicking signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0017] Figure 1 is a flowchart of an implementation method of a radar-based kick signal recognition method provided by an embodiment of the present invention;

[0018] Figure 2 is a complete and specific flow chart of a radar-based kick signal recognition method provided by an embodiment of the present invention;

[0019] Figure 3 is a schematic structural diagram of a radar-based kick signal recognition device provided by an embodiment of the present invention;

[0020] Figure 4 Schematic diagram of a radar provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0022] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0023] See also Figure 1 , which shows a flowchart of the implementation of the radar-based kick signal recognition method provided by an embodiment of the present invention, which is described in detail as follows:

[0024] S101: Acquire point cloud information of an echo signal detected by a radar; the point cloud information includes distance dimension features, angle dimension features, and speed dimension features.

[0025] In an embodiment of the present invention, the radar includes a transmitting antenna and a receiving antenna. After the vehicle door is unlocked, the radar is powered on and sends electromagnetic waves of a specific frequency to a fixed kicking area through the transmitting antenna. The electromagnetic waves are reflected when encountering an object, and the receiving antenna receives the echo signal. After ADC sampling and bandpass filtering, the out-of-band interference signal is filtered out to a certain extent.

[0026] After obtaining the echo signal with the out-of-band interference signal filtered out, the radar performs FFT (fast Fourier transform) processing on the echo signal to obtain a range-Doppler spectrum diagram, and performs constant false alarm detection on the range-Doppler spectrum diagram to obtain multiple target points. The distance, angle, and speed of each target point relative to the radar form point cloud information.

[0027] Specifically, the point cloud information includes velocity dimension features, distance dimension features, angle dimension features and signal-to-noise ratio features, wherein the angle dimension features include azimuth angle features and elevation angle features.

[0028] S102: For each frame of point cloud, calculate the feature mean of the points in the frame of point cloud whose distance dimension features meet preset conditions, and store the feature mean corresponding to the frame of point cloud into a feature mean matrix.

[0029] In this embodiment, the kick signal usually needs to limit the identification distance range, such as the distance range corresponding to the entire bumper of the vehicle, and the signal beyond this distance range can be considered as an interference signal. Therefore, after acquiring the point cloud information, this embodiment extracts the features of the points in each frame of the point cloud whose distance dimension features meet the preset conditions to construct a feature mean matrix.

[0030] Specifically, the feature means include a velocity dimension feature mean, a distance dimension feature mean and an angle dimension feature mean, and the angle dimension feature mean includes an azimuth angle feature mean and / or an elevation angle feature mean.

[0031] S103: Divide the detection area of ​​the radar in the angle dimension into grids, determine the angle range of each grid, and determine the centroid position of the point cloud trajectory according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix.

[0032] In order to find the centroid position of the point cloud trajectory, this embodiment first divides the detection area of ​​the radar into grids, and each grid corresponds to a certain angle range. According to the angle dimension feature mean of each frame of point cloud and the angle range of each grid, the position of each frame of point cloud in the grid is determined, and the centroid position of the point cloud trajectory corresponding to the feature mean matrix KnickFeatureALL is determined according to the position of the grid where each frame of point cloud is located, and then the kick target and the false alarm target are accurately distinguished through the relative relationship between the centroid position and the detection area.

[0033] S104: If the centroid position is located within the detection area and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located within a preset speed range, the distance dimension feature mean of each frame point cloud in the feature mean matrix is ​​extracted to obtain a distance dimension array.

[0034] In this embodiment, the kick signal usually meets certain rules in terms of speed compared to other false alarm signals. Therefore, this embodiment can also detect whether there is a kick signal in the point cloud based on the speed dimension feature.

[0035] S105: Determine whether a kick target is detected according to the distance dimension array.

[0036] In this embodiment, if the centroid position is located in the detection area, and the speed dimension feature and the distance dimension feature of the point cloud both meet corresponding conditions, it can be determined that the point cloud information includes a kicking target.

[0037] After acquiring the kicking target, the radar sends CAN message data to the door controller, and the door controller controls the door to open according to the received CAN message data.

[0038] Through the above scheme, this embodiment can comprehensively identify the kick signal based on the angle dimension features, speed dimension features and distance dimension features of the point cloud information detected by the radar, avoid false alarms, and improve the recognition accuracy of the kick signal.

[0039] In one embodiment, the specific implementation process of S102 includes:

[0040] Extracting points in the frame point cloud whose distance dimension features are less than or equal to a first preset distance threshold as target points;

[0041] The feature means corresponding to various features of all target points in the frame point cloud are calculated, and the feature means corresponding to various features of the frame point cloud are stored in the feature mean matrix.

[0042] Specifically, the points in each frame of point cloud whose distance dimension features are less than or equal to the first preset distance threshold are selected as target points, and other points are removed as interference points to determine the target point. After obtaining the target point, for any type of feature, the features of this type in the same frame of point cloud are averaged, and the feature means corresponding to each type of feature of the frame of point cloud are calculated and stored in the feature mean matrix. The feature mean matrix stores the feature means of each frame of point cloud in chronological order.

[0043] For example, the feature mean matrix KnickFeatureALL may be KnickFeatureALL{{cfarAzi1,cfarPzi1,cfarV1,cfarR1,cfarSnr1},{cfarAzi2,cfarPzi2,cfarV2,cfarR2,cfarSnr2},…,{cfarAzi n ,cfarPzi n ,cfarV n ,cfarR n ,cfarSnr n}}.

[0044] Among them, cfarAzi n Represents the mean azimuth feature of the point cloud information of the nth frame, cfarPzi n Represents the mean value of the pitch angle feature of the point cloud information of the nth frame, cfarV n Represents the mean value of the velocity dimension feature of the n-th frame point cloud information, cfarR n Represents the distance dimension feature mean of the nth frame point cloud information, cfarSnr n Represents the mean signal-to-noise ratio feature of the n-th frame point cloud information.

[0045] In one embodiment, before S103, the method provided in this embodiment further includes:

[0046] If the frame point cloud is an empty frame or the distance dimension feature of the frame point cloud is greater than a first preset distance threshold, the first count value is added to the first value;

[0047] If the first count value reaches the first count threshold, stop updating the feature mean matrix, and continue to execute the step of gridding the detection area of ​​the radar in the angle dimension and determining the angle range of each grid; otherwise, return to the step of obtaining the point cloud information of the echo signal detected by the radar.

[0048] In this embodiment, the first value may be 1. When the radar is powered on, the first count value is initialized to zero. When a frame of point cloud information is detected to be an empty frame or the distance dimension feature of the frame of point cloud is greater than the first preset distance threshold, the first count value is added by 1. If the first count value reaches the first count threshold, the feature mean matrix is ​​stopped from being updated, and step S103 is executed to perform a kick target judgment for the currently acquired feature mean matrix. If the first count value is less than the first count threshold, the process returns to S101 to continue executing and update the feature mean matrix.

[0049] In one embodiment, the specific implementation process of S103 includes:

[0050] S201: Divide the detection area of ​​the radar in the angle dimension into a plurality of sub-areas, and divide each sub-area into grids to determine the angle range of each grid.

[0051] Specifically, the angle dimension features include azimuth features and elevation features. This embodiment divides the detection area into multiple sub-areas according to the azimuth range angleAzi[M, N] and the elevation range anglePit[A, B] of the radar detection beam. Each sub-area is divided into multiple rectangles to form a grid, and each grid corresponds to a different azimuth and elevation range.

[0052] Specifically, the angle range of each grid division is the same, and each grid can be a small square of 2*2°.

[0053] S202: Determine the number of point clouds located in each grid according to the angle range of each grid and the angle dimension feature mean of each frame of point cloud in the feature mean matrix.

[0054] S203: For each grid, if the number of point clouds in the grid is greater than a preset threshold corresponding to the sub-region where the grid is located, the value corresponding to the grid is set to the second value, otherwise it is set to the third value, and a grid matrix is ​​constructed, which includes the values ​​corresponding to all grids.

[0055] In this embodiment, since the center of gravity of the kick signal is different at different positions in the detection area, while the false alarm point does not have this feature, this embodiment divides the detection area into multiple sub-areas, and by setting different preset thresholds in different sub-areas, the false alarm points can be screened out to improve the recognition efficiency of the kick signal.

[0056] Specifically, the closer the sub-region is to the center, the larger the corresponding preset threshold is. For example, if the detection area is divided into three sub-regions: left, middle, and right, the preset threshold of the middle sub-region is greater than the preset thresholds of the left and right sub-regions.

[0057] Specifically, the second value is not equal to the third value, the second value may be 1, and the third value may be 0. Each grid in the grid matrix is ​​arranged according to the size of the angle range.

[0058] Exemplarily, after dividing the grid, construct the grid position matrix Among them, x mn In the azimuth dimension, y represents a grid with an azimuth angle ranging from p to q. mn Represents a grid with pitch angles ranging from m to n in the pitch angle dimension.

[0059] After calculating the values ​​corresponding to each grid, the values ​​corresponding to each grid are stored in the corresponding position of the position matrix to form a grid matrix.

[0060] S204: For each grid in the grid matrix, if the values ​​of the grid and its adjacent grids are both the second value, updating the value of the grid to a fourth value.

[0061] Specifically, if the values ​​of the current grid and the grids before, after, left and right of the current grid are all the second value, the value of the grid is updated to the fourth value.

[0062] Specifically, the fourth value is not equal to the second value and the third value. For example, the fourth value may be 2.

[0063] S205: extracting the angle range of all grids whose values ​​are the fourth value in the updated grid matrix, and calculating the centroid position of the point cloud trajectory according to the angle range of all grids whose values ​​are the fourth value.

[0064] This embodiment counts the number of point clouds in each grid, and assigns values ​​to the grids layer by layer according to the number of point clouds in each grid to achieve point cloud clustering and filter out stray points. Through the above clustering method, the grid assigned with the fourth value can be used as the cluster center point of the point cloud, so as to calculate the centroid position according to the angle range of the clustered grid, thereby improving the accuracy of the centroid position.

[0065] In one embodiment, the angle dimension feature includes an azimuth feature and an elevation feature; the angle range includes an azimuth range and an elevation range; the centroid position includes an average azimuth range and an average elevation range;

[0066] The specific implementation process of S205 includes:

[0067] Averaging the azimuth ranges of all grids whose values ​​are the fourth value to obtain an average azimuth range;

[0068] The pitch angle ranges of all grids whose values ​​are the fourth value are averaged to obtain an average pitch angle range.

[0069] In one embodiment, the detection area includes multiple sub-areas; the specific implementation process of S104 includes:

[0070] If the centroid position is located in one of the sub-areas and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located in a preset speed range, the distance dimension feature mean of each frame point cloud in the feature mean matrix is ​​extracted to obtain a distance dimension array.

[0071] Specifically, each distance dimension feature mean in the distance dimension array is sorted according to the numerical value. If the centroid position is not located in any sub-region, or the speed dimension feature mean of the point cloud in the feature mean matrix is ​​not located in the preset speed range, it is determined that there is no kick target.

[0072] In one embodiment, the specific implementation process of S105 includes:

[0073] For each distance dimension feature mean in the distance dimension array, the difference between the distance dimension feature mean and the next distance dimension feature mean is calculated; if the difference between the distance dimension feature mean and the next distance dimension feature mean is less than a first preset difference, the frame corresponding to the distance dimension feature mean is used as the envelope start frame; if the difference between the distance dimension feature mean and the next distance dimension feature mean is greater than a second preset difference, the frame corresponding to the next distance dimension feature mean is used as the envelope end frame.

[0074] The distance dimension feature mean between adjacent envelope start frames and envelope end frames is taken as an envelope, and the number of envelopes in the distance dimension array is counted.

[0075] If the number of envelopes is less than the preset number of envelopes, it is determined that a kick target is detected.

[0076] If the number of envelopes is greater than the preset number of envelopes, the distance dimension feature means less than a second preset distance threshold in the feature mean matrix are extracted to construct a distance dimension feature matrix.

[0077] The number of peak values ​​in the distance dimension feature matrix is ​​counted, and if the number of peak values ​​is less than a first preset number, it is determined that a kick target is detected.

[0078] The applicant has found through a large number of experiments that the difference between a kicking target and other false alarm targets is that the number of envelopes in the kicking target is small, while the number of envelopes in the false alarm target is large. Based on the above findings, the present embodiment traverses the distance dimension features in the distance dimension array, and each time a pair of adjacent envelope start frames and envelope end frames are detected, the frame between the envelope start frame and the envelope end frame is taken as an envelope, and the number of envelopes in the distance dimension array is counted. If the number of envelopes is less than the preset number of envelopes, it is determined that a kicking target is detected; if the number of envelopes is greater than or equal to the preset number of envelopes, it is determined whether there is a kicking target in the point cloud based on the number of peaks in the distance dimension feature matrix.

[0079] As a specific embodiment, if the number of envelopes is less than or equal to the preset number of envelopes, it is determined that a kick to the target is detected.

[0080] Specifically, the elements in the distance dimension feature matrix are sorted according to the frame order.

[0081] On the basis of judging the kicking target based on the number of envelopes, in order to improve the accuracy of judging the kicking target, the present embodiment can also identify the kicking target based on the number of peaks in the distance dimension feature matrix. Specifically, the applicant has found through a large number of experiments that when the point cloud includes a kicking target, the number of peaks in the corresponding distance dimension feature matrix is ​​small, and when the target in the point cloud is a false alarm target, the number of peaks in the corresponding distance dimension feature matrix is ​​large. Based on this finding, the present embodiment determines that the kicking target is detected when the number of peaks is less than the first preset number, otherwise it determines that there is no kicking target.

[0082] After the kick target determination of the current feature mean matrix is ​​completed, the current feature mean matrix is ​​deleted, and the first count value is reset to zero, and the subsequent point cloud data is screened again for kick target recognition.

[0083] The following takes a specific application scenario as an example to describe in detail the radar-based kick signal recognition method provided by an embodiment of the present invention, as follows;

[0084] Step A: After the car door is unlocked, the radar is powered on and starts working;

[0085] Step B: Obtain the point cloud information of the echo signal detected by the radar; select the points that meet the preset conditions in each frame of the point cloud through the distance dimension feature, and construct the feature mean matrix;

[0086] Step C: If the frame point cloud is an empty frame or the distance dimension feature of the frame point cloud is greater than the first preset distance threshold, the first count value is accumulated and added by 1;

[0087] Step D: Determine whether the first count value is greater than or equal to the first count threshold, if so, proceed to step E, otherwise return to step B;

[0088] Step E: Divide the two-dimensional area formed by the azimuth and elevation angles of the radar beam into three sub-areas, and set the same grid interval in each sub-area;

[0089] Step F: If the number of point clouds in the adjacent multiple grids is greater than the preset threshold corresponding to the area where the grid is located, the central grid among the adjacent multiple grids is extracted, and the azimuth angle range and pitch angle range of the central grid are used to calculate the centroid;

[0090] Step G: Determine whether the centroid is located in any sub-area. If so, proceed to step H; otherwise, proceed to step N.

[0091] Step H: Determine whether the velocity dimension mean of each frame point cloud in the feature mean matrix is ​​within the preset velocity range. If so, proceed to step I; otherwise, proceed to step N.

[0092] Step I: extract the distance dimension feature mean in the feature mean matrix to obtain a distance dimension array;

[0093] Step J: traverse the number of envelopes in the distance dimension array, if the number of envelopes is less than the preset number of envelopes, go to step M, otherwise continue to step K;

[0094] Step K: extracting the distance dimension feature means less than the second preset distance threshold in the feature mean matrix to construct a distance dimension feature matrix;

[0095] Step L: Count the number of peaks in the distance dimension feature matrix. If the number of peaks is less than a first preset number, proceed to step M; otherwise, proceed to step N.

[0096] Step M: The kicking target is detected in the current point cloud.

[0097] Step N: No kicking target is detected in the current point cloud.

[0098] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0099] The following is an embodiment of the device of the present invention. For details not described in detail, reference may be made to the corresponding method embodiments described above.

[0100] Figure 3The structure diagram of the radar-based kick signal recognition device provided by an embodiment of the present invention is shown. For the convenience of description, only the part related to the embodiment of the present invention is shown, which is described in detail as follows:

[0101] like Figure 3 As shown, the radar-based kick signal recognition device 100 includes:

[0102] The point cloud information acquisition module 110 is used to acquire the point cloud information of the echo signal detected by the radar; the point cloud information includes distance dimension features, angle dimension features and speed dimension features;

[0103] The feature mean matrix construction module 120 is used to calculate the feature mean of the points whose distance dimension features satisfy the preset conditions in each frame of point cloud, and store the feature mean corresponding to the frame of point cloud into the feature mean matrix;

[0104] The centroid position calculation module 130 is used to divide the detection area of ​​the radar in the angle dimension into grids, determine the angle range of each grid, and determine the centroid position of the point cloud trajectory according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix;

[0105] A distance dimension array calculation module 140 is used to extract the distance dimension feature mean of each frame of point cloud in the feature mean matrix to obtain a distance dimension array if the centroid position is located in the detection area and the speed dimension feature mean of each frame of point cloud in the feature mean matrix is ​​located in a preset speed range;

[0106] The kick target determination module 150 is used to determine whether a kick target is detected according to the distance dimension array.

[0107] In one embodiment, the feature mean matrix construction module 120 includes:

[0108] Extracting points in the frame point cloud whose distance dimension features are less than or equal to a first preset distance threshold as target points;

[0109] The feature means corresponding to various features of all target points in the frame point cloud are calculated, and the feature means corresponding to various features of the frame point cloud are stored in the feature mean matrix.

[0110] In one embodiment, the radar-based kick signal recognition device 100 further includes a counting module for:

[0111] If the frame point cloud is an empty frame or the distance dimension feature of the frame point cloud is greater than a first preset distance threshold, the first count value is added to the first value;

[0112] If the first count value reaches the first count threshold, stop updating the feature mean matrix, and continue to execute the step of gridding the detection area of ​​the radar in the angle dimension and determining the angle range of each grid; otherwise, return to the step of obtaining the point cloud information of the echo signal detected by the radar.

[0113] In one embodiment, the centroid position calculation module 130 includes:

[0114] A grid division unit, used to divide the detection area of ​​the radar in the angle dimension into a plurality of sub-areas, and to perform grid division on each sub-area to determine the angle range of each grid;

[0115] A point cloud quantity counting unit, used to determine the number of point clouds located in each grid according to the angle range of each grid and the angle dimension feature mean of each frame of point cloud in the feature mean matrix;

[0116] A first grid value assignment unit is used for setting, for each grid, a value corresponding to the grid to a second value if the number of point clouds in the grid is greater than a preset threshold corresponding to the sub-region where the grid is located, otherwise it is set to a third value, and a grid matrix is ​​constructed, wherein the grid matrix includes the values ​​corresponding to all grids;

[0117] A second grid value assignment unit is used for updating the value of each grid in the grid matrix to a fourth value if the values ​​of the grid and its adjacent grids are both the second value;

[0118] The centroid position calculation unit is used to extract the angle range of all grids whose values ​​are the fourth value in the updated grid matrix, and calculate the centroid position of the point cloud trajectory according to the angle range of all grids whose values ​​are the fourth value.

[0119] In one embodiment, the angle dimension feature includes an azimuth feature and an elevation feature; the angle range includes an azimuth range and an elevation range; the centroid position includes an azimuth average range and an elevation average range; and the centroid position calculation unit includes:

[0120] Averaging the azimuth ranges of all grids whose values ​​are the fourth value to obtain an average azimuth range;

[0121] The pitch angle ranges of all grids whose values ​​are the fourth value are averaged to obtain an average pitch angle range.

[0122] In one embodiment, the detection area includes a plurality of sub-areas;

[0123] The distance dimension array calculation module 140 includes:

[0124] If the centroid position is located in one of the sub-areas and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located in a preset speed range, the distance dimension feature mean of each frame point cloud in the feature mean matrix is ​​extracted to obtain a distance dimension array.

[0125] In one embodiment, the kick target determination module 150 includes:

[0126] For each distance dimension feature mean in the distance dimension array, the difference between the distance dimension feature mean and the next distance dimension feature mean is calculated; if the difference between the distance dimension feature mean and the next distance dimension feature mean is less than a first preset difference, the frame corresponding to the distance dimension feature mean is used as the envelope start frame; if the difference between the distance dimension feature mean and the next distance dimension feature mean is greater than a second preset difference, the frame corresponding to the next distance dimension feature mean is used as the envelope end frame;

[0127] Taking the distance dimension feature mean between adjacent envelope start frames and envelope end frames as an envelope, and counting the number of envelopes in the distance dimension array;

[0128] If the number of envelopes is less than the preset number of envelopes, it is determined that a kick target is detected;

[0129] If the number of envelopes is greater than the preset number of envelopes, extracting the distance dimension feature means less than a second preset distance threshold in the feature mean matrix to construct a distance dimension feature matrix;

[0130] The number of peak values ​​in the distance dimension feature matrix is ​​counted, and if the number of peak values ​​is less than a first preset number, it is determined that a kick target is detected.

[0131] As can be seen from the above embodiments, the embodiment of the present invention first obtains the point cloud information of the echo signal detected by the radar; for each frame of point cloud, the feature mean of the point whose distance dimension feature meets the preset conditions in the frame of point cloud is calculated, and the feature mean corresponding to the frame of point cloud is stored in the feature mean matrix; then the detection area of ​​the radar in the angle dimension is gridded, the angle range of each grid is determined, and the centroid position of the point cloud trajectory is determined according to the angle range of each grid and the angle dimension feature mean of each frame of point cloud in the feature mean matrix; if the centroid position is located in the detection area and the speed dimension feature mean of each frame of point cloud in the feature mean matrix is ​​located in the preset speed range, the distance dimension feature mean of each frame of point cloud in the feature mean matrix is ​​extracted to obtain the distance dimension array; finally, according to the distance dimension array, it is judged whether the kicking target is detected. Through the above scheme, the present embodiment can comprehensively identify the kicking signal according to the angle dimension feature, speed dimension feature and distance dimension feature of the point cloud information detected by the radar, avoid false alarms, and improve the recognition accuracy of the kicking signal.

[0132] Figure 4 FIG. 1 is a schematic diagram of a radar provided by an embodiment of the present invention. Figure 4 As shown, the radar 4 of this embodiment includes: a processor 40, a memory 41, and a computer program 42 stored in the memory 41 and executable on the processor 40. When the processor 40 executes the computer program 42, the steps in the above-mentioned embodiments of the radar-based kick signal recognition method are implemented, for example Figure 1 Alternatively, when the processor 40 executes the computer program 42, the functions of each module / unit in the above-mentioned device embodiments are realized, for example, Figure 3 The functions of the modules 110 to 150 are shown.

[0133] Exemplarily, the computer program 42 may be divided into one or more modules / units, which are stored in the memory 41 and executed by the processor 40 to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program 42 in the radar 4.

[0134] The radar 4 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The radar 4 may include, but is not limited to, a processor 40 and a memory 41. Those skilled in the art will appreciate that Figure 4 It is only an example of radar 4 and does not constitute a limitation of radar 4. It can include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal can also include input and output devices, network access devices, buses, etc.

[0135] The processor 40 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0136] The memory 41 may be an internal storage unit of the radar 4, such as a hard disk or memory of the radar 4. The memory 41 may also be an external storage device of the radar 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the radar 4. Further, the memory 41 may also include both an internal storage unit and an external storage device of the radar 4. The memory 41 is used to store the computer program and other programs and data required by the terminal. The memory 41 may also be used to temporarily store data that has been output or is to be output.

[0137] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0138] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0139] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0140] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0141] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0142] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0143] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned radar-based kick signal recognition method embodiment can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0144] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for identifying kicking signals based on radar, It is characterized in that include: Acquire point cloud information of the echo signal detected by the radar; the point cloud information includes distance dimension features, angle dimension features and speed dimension features; For each frame of point cloud, calculate the feature mean of the points in the frame of point cloud whose distance dimension features meet the preset conditions, and store the feature mean corresponding to the frame of point cloud into the feature mean matrix; Divide the detection area of ​​the radar in the angle dimension into grids, determine the angle range of each grid, and determine the centroid position of the point cloud trajectory according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix; If the centroid position is located within the detection area and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located within a preset speed range, then extracting the distance dimension feature mean of each frame point cloud in the feature mean matrix to obtain a distance dimension array; It is determined whether a kicking target is detected according to the distance dimension array.

2. The radar-based kick signal recognition method according to claim 1, It is characterized in that The step of calculating the feature mean of the points in the frame point cloud whose distance dimension features meet the preset conditions, and storing the feature mean corresponding to the frame point cloud into the feature mean matrix includes: Extracting points in the frame point cloud whose distance dimension features are less than or equal to a first preset distance threshold as target points; The feature means corresponding to various features of all target points in the frame point cloud are calculated, and the feature means corresponding to various features of the frame point cloud are stored in the feature mean matrix.

3. The radar-based kick signal recognition method according to claim 1 or 2, It is characterized in that Before dividing the detection area of ​​the radar in the angle dimension into grids and determining the angle range of each grid, the method further includes: If the frame point cloud is an empty frame or the distance dimension feature of the frame point cloud is greater than a first preset distance threshold, the first count value is added to the first value; If the first count value reaches the first count threshold, stop updating the feature mean matrix, and continue to execute the step of gridding the detection area of ​​the radar in the angle dimension and determining the angle range of each grid; otherwise, return to the step of obtaining the point cloud information of the echo signal detected by the radar.

4. The radar-based kick signal recognition method according to claim 1, It is characterized in that The detecting area of ​​the radar in the angle dimension is gridded, the angle range of each grid is determined, and the centroid position of the point cloud trajectory is determined according to the angle range of each grid and the angle dimension feature mean of each frame of the point cloud in the feature mean matrix, including: Dividing the detection area of ​​the radar in the angle dimension into a plurality of sub-areas, and dividing each sub-area into grids to determine the angle range of each grid; Determine the number of point clouds located in each grid according to the angle range of each grid and the angle dimension feature mean of each frame of point cloud in the feature mean matrix; For each grid, if the number of point clouds in the grid is greater than the preset threshold corresponding to the sub-region where the grid is located, the value corresponding to the grid is set to the second value, otherwise it is set to the third value, and a grid matrix is ​​constructed, wherein the grid matrix includes the values ​​corresponding to all grids; For each grid in the grid matrix, if the values ​​of the grid and its adjacent grids are both the second value, updating the value of the grid to a fourth value; The angle range of all grids whose values ​​are the fourth value in the updated grid matrix is ​​extracted, and the centroid position of the point cloud trajectory is calculated according to the angle range of all grids whose values ​​are the fourth value.

5. The radar-based kick signal recognition method as claimed in claim 4, It is characterized in that The angle dimension feature includes an azimuth feature and an elevation feature; the angle range includes an azimuth range and an elevation range; the centroid position includes an azimuth average range and an elevation average range; The calculating the centroid position of the point cloud trajectory according to the angle range of all grids whose values ​​are the fourth value includes: Averaging the azimuth ranges of all grids whose values ​​are the fourth value to obtain an average azimuth range; The pitch angle ranges of all grids whose values ​​are the fourth value are averaged to obtain an average pitch angle range.

6. The radar-based kick signal recognition method as claimed in claim 4, It is characterized in that The detection area includes a plurality of sub-areas; If the centroid position is located within the detection area and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located within a preset speed range, the distance dimension feature mean of each frame point cloud in the feature mean matrix is ​​extracted to obtain a distance dimension array, including: If the centroid position is located in one of the sub-areas and the speed dimension feature mean of each frame point cloud in the feature mean matrix is ​​located in a preset speed range, the distance dimension feature mean of each frame point cloud in the feature mean matrix is ​​extracted to obtain a distance dimension array.

7. The radar-based kick signal recognition method according to claim 1, It is characterized in that The step of judging whether a kick target is detected according to the distance dimension array comprises: For each distance dimension feature mean in the distance dimension array, the difference between the distance dimension feature mean and the next distance dimension feature mean is calculated; if the difference between the distance dimension feature mean and the next distance dimension feature mean is less than a first preset difference, the frame corresponding to the distance dimension feature mean is used as the envelope start frame; if the difference between the distance dimension feature mean and the next distance dimension feature mean is greater than a second preset difference, the frame corresponding to the next distance dimension feature mean is used as the envelope end frame; Taking the distance dimension feature mean between adjacent envelope start frames and envelope end frames as an envelope, and counting the number of envelopes in the distance dimension array; If the number of envelopes is less than the preset number of envelopes, it is determined that a kick target is detected; If the number of envelopes is greater than the preset number of envelopes, extracting the distance dimension feature means less than a second preset distance threshold in the feature mean matrix to construct a distance dimension feature matrix; The number of peak values ​​in the distance dimension feature matrix is ​​counted, and if the number of peak values ​​is less than a first preset number, it is determined that a kick target is detected.

8. A radar, It is characterized in that The method comprises a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program. It is characterized in that When the computer program is executed by a processor, the steps of the method as claimed in any one of claims 1 to 7 are implemented.

10. A car, It is characterized in that Comprising the radar as claimed in claim 8.

Images