UWB radar-based kick feature extraction method and system, and storage medium

Abstract

The invention relates to the technical field of radar electronics, and provides a UWB radar-based kicking feature extraction method and system and a storage medium, and the method comprises the steps: carrying out the recognition of a kicking motion through employing a UWB radar, carrying out the sampling analysis of original sampling data through employing a range gate as an interval, obtaining to-be-analyzed data, carrying out the feature extraction, and carrying out the feature extraction; the method comprises the following steps: acquiring accurate initial kicking feature data as a basis, performing data fluctuation analysis on to-be-analyzed data to obtain fluctuation features, and finally calculating target kicking features based on the fluctuation features and the initial kicking feature data to realize interference suppression; the kicking function is achieved based on the UWB technology, compared with a capacitive sensor, the kicking system has the advantages of being higher in precision, high in response speed, high in anti-interference capability and adaptive to diversified environments, has the advantages of being high in signal-to-noise ratio, low in calculation power and good in stability, and can further improve the recognition accuracy of kicking signals.

Description

Technical Field

[0001] The present invention relates to the field of radar electronic technology, and in particular to a kick feature extraction method, system and storage medium based on UWB radar. Background Art

[0002] Currently, the vehicle's kick-trunk function is mainly implemented based on capacitive sensors, which has the following limitations:

[0003] 1. The detection accuracy of capacitive sensors for target objects is greatly affected by environmental factors such as humidity and temperature changes, which may cause unstable detection distance or large errors, affecting the accuracy of the kicking action.

[0004] 2. The response speed of the capacitive sensor may be limited by circuit design and signal processing, and may not be able to meet the needs of kicking actions that require a quick response.

[0005] 3. Electromagnetic interference or other sensor signals in the surrounding environment may affect the stability and accuracy of the capacitive sensor, thereby affecting the reliability of the kick function.

[0006] 4. Capacitive sensors may drift or fail during long-term use, especially under high-intensity use or harsh environmental conditions, and may require frequent maintenance and calibration.

[0007] 5. Currently, there are a few methods that use UWB to detect kicking functions: such as combining UWB signals with deep learning models. For example, Chinese patent document CN115273229A discloses a method, device, storage medium, sensor and vehicle for motion recognition. Its implementation uses ultra-wideband UWB (Ultra Wide Band) radar to recognize human kicking movements, and combines deep learning technology with UWB radar to solve the problems of misidentification and misaction in recognition technology. However, this method requires the establishment of a large number of test samples for model training. Since data processing is relatively complex, the required computing power is high, and the hardware requirements of the equipment are high. Summary of the invention

[0008] The present invention provides a kick feature extraction method, system and storage medium based on UWB radar, which solves the technical problems of high function maintenance cost caused by poor stability of existing kick function implementation solutions and high hardware equipment cost caused by high data processing difficulty.

[0009] In order to solve the above technical problems, the present invention provides a kick feature extraction method based on UWB radar, comprising:

[0010] Using the radar mode to perform signal sampling to obtain original sampling data, and performing sampling and analysis on the original sampling data at intervals of a range gate to obtain data to be analyzed;

[0011] Extracting features from the data to be analyzed to obtain initial kick feature data;

[0012] A data fluctuation analysis is performed on the data to be analyzed to obtain a fluctuation feature, and a target kick feature is calculated based on the fluctuation feature and the initial kick feature data.

[0013] This basic scheme adopts UWB radar to identify kicking actions. First, the original sampling data is sampled and analyzed at intervals of the range gate to obtain the data to be analyzed, and then feature extraction is performed to obtain accurate initial kicking feature data as a basis. Then, data fluctuation analysis is performed on the data to be analyzed to obtain fluctuation characteristics. Finally, the target kicking characteristics are calculated based on the fluctuation characteristics and the initial kicking feature data to achieve interference suppression. The kicking function is realized based on UWB technology. Compared with capacitive sensors, it has higher accuracy, faster response speed, stronger anti-interference ability and the advantage of adapting to diverse environments. It has good stability and can further improve the recognition accuracy of kicking signals.

[0014] In a further embodiment, the method of sampling signals using a radar mode to obtain raw sampled data includes:

[0015] Ultra-wideband UWB uses radar mode to obtain frame sampling data at a preset sampling rate;

[0016] The frame sampling data is read, the real part I and the imaginary part Q of the channel impulse response CIR of each frame are obtained, the amplitude value of each CIR sampling point is calculated, and the original sampling data is obtained.

[0017] This scheme calculates the amplitude value of each CIR sampling point based on the real part I and imaginary part Q of the channel impulse response CIR of each frame to obtain the original sampling data. By extracting the amplitude value data that obviously reflects the kicking signal process, it provides an effective data basis for kicking feature extraction.

[0018] In a further embodiment, the sampling and analysis of the original sampled data at intervals of a range gate to obtain the data to be analyzed includes:

[0019] Obtaining N range gate data from each frame of the original sampled data;

[0020] Calculate the optimal range gate sequence number based on the range resolution and installation height;

[0021] Determine at least two target range gate sequence numbers according to the number of frames of the real-time data window and the optimal range gate sequence number;

[0022] The range gate data corresponding to the target range gate serial number is obtained, and the data is fused according to a preset weight to obtain the data to be analyzed.

[0023] This scheme increases the accuracy and efficiency of target detection by simultaneously checking the range gate data of multiple target range gate serial numbers and fusing the data according to the coefficients of preset weights, thereby providing reliable target range information for the application.

[0024] In a further embodiment, feature extraction is performed on the data to be analyzed to obtain initial kick feature data, including:

[0025] Based on the data to be analyzed, calculating the standard deviation within the signal window;

[0026] Based on the data to be analyzed, calculating the root mean square value within the signal window;

[0027] Initial kick feature data is calculated based on the standard deviation and the root mean square value.

[0028] This scheme calculates the standard deviation and root mean square value based on the data to be analyzed, and then calculates the initial kick feature data. The standard deviation and root mean square value are excellent in avoiding the influence of extreme values ​​to provide more stable statistics, which is crucial for the stability and reliability of the data, and is conducive to detecting kick features and further improving the stability of kick signal recognition.

[0029] In a further embodiment, performing data fluctuation analysis on the data to be analyzed to obtain fluctuation characteristics comprises:

[0030] Based on the data length of the sliding window, sliding window processing is performed on the real-time data of the initial kicking feature data in sequence to obtain feature signal data;

[0031] Calculate the target sliding window length according to the preset sampling rate;

[0032] A window is set according to the target sliding window length, and variance analysis is performed on the characteristic signal data to obtain a set of fluctuation characteristics.

[0033] In a further embodiment, the calculating of the target kick feature based on the fluctuation feature and the initial kick feature data comprises:

[0034] Obtaining a fluctuation minimum value from the fluctuation feature as a noise suppression parameter;

[0035] Perform interference suppression on the characteristic signal data according to the noise suppression parameter to obtain a target kicking feature;

[0036] Wherein, the noise suppression parameter is,

[0037] flu_min=min(flu_var)

[0038] Where flu_min is the noise suppression parameter, flu_var is the fluctuation characteristic, and min() means taking the minimum value.

[0039] In a further embodiment, the calculation formula of the fluctuation characteristic is as follows:

[0040]

[0041] Wherein, flu_var is the fluctuation characteristic, n is the total number of characteristic signal data in the window, is the average of the characteristic signal data in the window, x i is the i-th characteristic signal data in the window.

[0042] This scheme uses weight coefficients to realize data fusion of specific range gate units, and performs preliminary kick feature SR extraction through standard deviation and root mean square. Subsequently, the fluctuation degree of the signal is used to obtain noise suppression parameters according to the sliding window, and the final feature is obtained in combination with the preliminary kick feature. The kick feature can be highlighted, which is beneficial to subsequent kick detection and has the advantages of high signal-to-noise ratio and low computing power.

[0043] In a further embodiment, the target sliding window length is calculated as follows:

[0044] Win_var=fs*0.1

[0045] Where Win_var is the target sliding window length and fs is the preset sampling rate.

[0046] This solution sets a suitable target sliding window length based on a preset sampling rate, thereby taking into account the allocation of time and accuracy during the sampling process, and the range and accuracy of signal processing.

[0047] The present invention provides a kick feature extraction system based on UWB radar, comprising a memory and a processor, wherein the memory is coupled to the processor;

[0048] Wherein, the memory is used to store program data;

[0049] The processor is used to execute the program data to implement the kick feature extraction method based on UWB radar as described above.

[0050] The kick feature extraction system in this scheme adopts various modules to implement various steps in the kick feature extraction method, providing a hardware basis for the kick feature extraction method and facilitating the implementation of the method.

[0051] The present invention also provides a storage medium on which a computer program is stored, and the computer program is used to be loaded by the kick feature extraction system based on UWB radar as described above to implement the kick feature extraction method based on UWB radar as described above. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flow chart of a kick feature extraction method based on UWB radar provided by an embodiment of the present invention;

[0053] Figure 2 It is a workflow diagram provided by an embodiment of the present invention; DETAILED DESCRIPTION

[0054] The following specifically illustrates the implementation mode of the present invention in conjunction with the accompanying drawings. The embodiments are provided for illustrative purposes only and cannot be understood as limiting the present invention. The accompanying drawings are provided for reference and illustration only and do not constitute a limitation on the scope of patent protection of the present invention, because many changes can be made to the present invention without departing from the spirit and scope of the present invention.

[0055] Example 1

[0056] The embodiment of the present invention provides a kick feature extraction method based on UWB radar, such as Figure 1 , Figure 2 As shown, in this embodiment, steps S1 to S5 are included:

[0057] S1. Use radar mode to sample signals and obtain original sampling data, including:

[0058] S11, ultra-wideband UWB uses radar mode to obtain frame sampling data at a preset sampling rate;

[0059] The preset sampling rate may be adjusted according to actual needs, for example, adjusted to a preset sampling rate fs=100 Hz.

[0060] S12, reading the frame sampling data, obtaining the real part I and the imaginary part Q of the channel impulse response CIR of each frame, calculating the amplitude value of each CIR sampling point, and obtaining the original sampling data.

[0061] The formula for the amplitude value is as follows:

[0062]

[0063] This embodiment calculates the amplitude value of each CIR sampling point based on the real part I and imaginary part Q of the channel impulse response CIR of each frame to obtain the original sampling data. By extracting the amplitude value data that obviously reflects the kicking signal process, an effective data basis is provided for kicking feature extraction.

[0064] S2, sampling and analyzing the original sampled data at intervals of a range gate to obtain data to be analyzed, including:

[0065] S21, obtaining N range gate data from each frame of the original sampled data;

[0066] S22. Calculate the optimal range gate sequence number based on the range resolution and installation height, as follows:

[0067]

[0068] Where N_best_tap is the optimal range gate sequence number, H is the installation height, and Tap_res is the range resolution.

[0069] S23, determining at least two target range gate sequence numbers according to the number of frames of the real-time data window and the optimal range gate sequence number;

[0070] Specifically, the kicking features are often concentrated in several adjacent TAPs, so the data of multiple TAPs are selected for kicking detection to determine the three target range gate sequence numbers. In the kicking feature analysis, the number of frames of the real-time data window Win_curtime_size = 128. The data of the three range gates data1, data2, and data3 are analyzed in real time, and the target range gate sequence numbers are N_best_tap-1, N_best_tap, and N_best_tap+1 respectively.

[0071] S24, obtaining the range gate data corresponding to the target range gate serial number, and performing data fusion according to a preset weight to obtain data to be analyzed.

[0072] Among them, the preset weights can be customized according to actual conditions; or, they can be dynamically set with reference to historical / standard data.

[0073] For example, Data_cur=0.8*data1+1.2*data2+0.8*data3.

[0074] In the formula, Data_cur represents the data to be analyzed.

[0075] This embodiment increases the accuracy and efficiency of target detection by simultaneously checking the range gate data of multiple target range gate serial numbers and performing data fusion according to preset weight coefficients, thereby providing reliable target range information for the application.

[0076] S3, extracting features from the data to be analyzed to obtain initial kick feature data, including:

[0077] S31. Based on the data to be analyzed, the standard deviation within the signal window is calculated to determine the degree of fluctuation of the signal, as follows:

[0078]

[0079] Where n is the total number of frame sampling data in the window, x i is the sampling data of the i-th frame in the window, is the average value of the frame sampling data within the window.

[0080] S32. Based on the data to be analyzed, the root mean square value in the signal window is calculated to determine the energy of the signal, as follows:

[0081]

[0082] S33, calculating the initial kick feature data according to the standard deviation and the root mean square value, as follows:

[0083] SR=std*rms

[0084] SR = (std*rms) is used as a feature to detect kicking. When a kicking action occurs, the std standard deviation will increase in amplitude, and rms is used to assist in enhancing the signal energy at this time, which is beneficial to detecting the kicking feature.

[0085] This embodiment calculates the standard deviation and root mean square value based on the data to be analyzed, and then calculates the initial kick feature data. The standard deviation and root mean square value are excellent in avoiding the influence of extreme values ​​to provide more stable statistics, which is crucial to the stability and reliability of the data, is conducive to detecting the kick features, and further improves the stability of kick signal recognition.

[0086] S4, performing data fluctuation analysis on the data to be analyzed to obtain fluctuation characteristics, including:

[0087] S41, based on the data length of the sliding window, sequentially perform sliding window processing on the real-time data of the initial kicking feature data to obtain feature signal data;

[0088] Specifically, characteristic parameters are calculated for real-time data with a length of Win_curtime_size according to a sliding window, and the data length of the sliding window is Win_fea_size=20, so that a characteristic signal SR with a data length of (Win_curtime_size-Win_fea_size+1) is obtained.

[0089] S42, calculating the target sliding window length according to the preset sampling rate;

[0090] In this embodiment, the calculation formula of the target sliding window length is as follows:

[0091] Win_var=fs*0.1

[0092] Where Win_var is the target sliding window length and fs is the preset sampling rate.

[0093] This embodiment sets a suitable target sliding window length based on a preset sampling rate, thereby taking into account the allocation of time and accuracy during the sampling process, and the range and accuracy of signal processing.

[0094] S43, setting a window according to the target sliding window length, performing variance analysis on the characteristic signal data, and obtaining a set of fluctuation characteristics.

[0095] In this embodiment, the calculation formula of the fluctuation characteristic is as follows:

[0096]

[0097] Wherein, flu_var is the fluctuation characteristic, n is the total number of characteristic signal data in the window, is the average of the characteristic signal data in the window, x i is the i-th characteristic signal data in the window.

[0098] The window is set according to the target sliding window length, and variance analysis is performed on the characteristic signal data to obtain a set of fluctuation characteristics. When the kicking action actually occurs, the entire action generally takes less than 1 second. Therefore, there is a period of small fluctuation in the characteristic signal data. The minimum value of the fluctuation is obtained as the noise suppression parameter to suppress the interference signal, which can achieve strong anti-interference ability and adapt to diverse environments.

[0099] S5, calculating the target kicking feature based on the fluctuation feature and the initial kicking feature data, including:

[0100] Obtaining a fluctuation minimum value from the fluctuation feature as a noise suppression parameter;

[0101] Perform interference suppression on the characteristic signal data according to the noise suppression parameter to obtain a target kicking feature;

[0102] Wherein, the noise suppression parameter is,

[0103] flu_min=min(flu_var)

[0104] Where flu_min is the noise suppression parameter, flu_var is the fluctuation characteristic, and min() means taking the minimum value.

[0105] The calculation formula of the target kick feature is as follows:

[0106] SR_fea=SR / flu_min

[0107] Wherein, SR is the characteristic signal data obtained in step S41.

[0108] In this embodiment, the target kicking feature obtained is applied to various subsequent kicking signal analyses.

[0109] This embodiment uses weight coefficients to achieve data fusion of specific range gate units, and performs preliminary kick feature SR extraction through standard deviation and root mean square. Subsequently, the signal fluctuation degree is used to obtain noise suppression parameters according to the sliding window, and the final feature is acquired in combination with the preliminary kick feature. The kick feature can be highlighted, which is beneficial to subsequent kick detection and has the advantages of high signal-to-noise ratio and low computing power.

[0110] The embodiment of the present invention adopts UWB radar to identify kicking actions. First, the original sampling data is sampled and analyzed at intervals of a range gate to obtain data to be analyzed, and then feature extraction is performed to obtain accurate initial kicking feature data as a basis. Then, data fluctuation analysis is performed on the data to be analyzed to obtain fluctuation characteristics. Finally, the target kicking characteristics are calculated based on the fluctuation characteristics and the initial kicking feature data to achieve interference suppression. The kicking function is realized based on UWB technology. Compared with capacitive sensors, it has higher accuracy, faster response speed, stronger anti-interference ability and the advantage of adapting to diverse environments, and has good stability, which can further improve the recognition accuracy of kicking signals.

[0111] Example 2

[0112] An embodiment of the present invention provides a kick feature extraction system based on UWB radar, comprising a memory and a processor, wherein the memory is coupled to the processor;

[0113] Wherein, the memory is used to store program data;

[0114] The processor is used to execute the program data to implement the kick feature extraction method based on UWB radar as described in Example 1 above.

[0115] Since the above modules are mainly used to execute the steps in Example 1, and the more specific contents of steps S1-S6 have been described in Example 1, they will not be repeated in this example.

[0116] The kick feature extraction system in this embodiment uses various modules to implement various steps in the kick feature extraction method, providing a hardware foundation for the kick feature extraction method and facilitating the implementation of the method.

[0117] Example 3

[0118] The embodiment of the present invention further provides a storage medium on which a computer program is stored, and the computer program is used to be loaded by the kick feature extraction system based on UWB radar as described in the above embodiment 2 to implement the kick feature extraction method based on UWB radar as described in the above embodiment 1. The storage medium may be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0119] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this embodiment in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A kick feature extraction method based on UWB radar, characterized in that: include: Using the radar mode to perform signal sampling to obtain original sampling data, and performing sampling and analysis on the original sampling data at intervals of a range gate to obtain data to be analyzed; Extracting features from the data to be analyzed to obtain initial kick feature data; A data fluctuation analysis is performed on the data to be analyzed to obtain a fluctuation feature, and a target kick feature is calculated based on the fluctuation feature and the initial kick feature data.

2. A kick feature extraction method based on UWB radar as claimed in claim 1, characterized in that: The method of sampling the signal using the radar mode to obtain the original sampling data includes: Ultra-wideband UWB uses radar mode to obtain frame sampling data at a preset sampling rate; The frame sampling data is read, the real part I and the imaginary part Q of the channel impulse response CIR of each frame are obtained, the amplitude value of each CIR sampling point is calculated, and the original sampling data is obtained.

3. A kick feature extraction method based on UWB radar as claimed in claim 2, characterized in that: The sampling and analysis of the original sampled data at intervals of the range gate to obtain the data to be analyzed includes: Obtaining N range gate data from each frame of the original sampled data; Calculate the optimal range gate sequence number based on the range resolution and installation height; Determine at least two target range gate sequence numbers according to the number of frames of the real-time data window and the optimal range gate sequence number; The range gate data corresponding to the target range gate serial number is obtained, and the data is fused according to a preset weight to obtain the data to be analyzed.

4. A kick feature extraction method based on UWB radar as claimed in claim 1, characterized in that: The step of extracting features from the data to be analyzed to obtain initial kick feature data includes: Based on the data to be analyzed, calculating the standard deviation within the signal window; Based on the data to be analyzed, calculating the root mean square value within the signal window; Initial kick feature data is calculated based on the standard deviation and the root mean square value.

5. A kick feature extraction method based on UWB radar as claimed in claim 4, characterized in that: The performing data fluctuation analysis on the data to be analyzed to obtain fluctuation characteristics includes: Based on the data length of the sliding window, sliding window processing is performed on the real-time data of the initial kicking feature data in sequence to obtain feature signal data; Calculate the target sliding window length according to the preset sampling rate; A window is set according to the target sliding window length, and variance analysis is performed on the characteristic signal data to obtain a set of fluctuation characteristics.

6. A kick feature extraction method based on UWB radar as claimed in claim 5, characterized in that: The calculating of the target kicking feature based on the fluctuation feature and the initial kicking feature data comprises: Obtaining a fluctuation minimum value from the fluctuation feature as a noise suppression parameter; Perform interference suppression on the characteristic signal data according to the noise suppression parameter to obtain a target kicking feature; Wherein, the noise suppression parameter is, flu_min=min(flu_var) Where flu_min is the noise suppression parameter, flu_var is the fluctuation characteristic, and min() means taking the minimum value.

7. A kick feature extraction method based on UWB radar as claimed in claim 5, characterized in that: The calculation formula of the fluctuation characteristics is as follows: Wherein, flu_var is the fluctuation characteristic, n is the total number of characteristic signal data in the window, is the average of the characteristic signal data in the window, x i is the i-th characteristic signal data in the window.

8. The kick feature extraction method based on UWB radar as claimed in claim 4, characterized in that: The calculation formula of the target sliding window length is as follows: Win_var=fs*0.1 Where Win_var is the target sliding window length and fs is the preset sampling rate.

9. A kick feature extraction system based on UWB radar, characterized in that: comprising a memory and a processor, wherein the memory is coupled to the processor; Wherein, the memory is used to store program data; The processor is used to execute the program data to implement the kick feature extraction method based on UWB radar as described in any one of claims 1-8.

10. A storage medium having a computer program stored thereon, characterized in that: The computer program is used to be loaded by the UWB radar-based kick feature extraction system according to claim 9 to implement the UWB radar-based kick feature extraction method according to any one of claims 1-8.

Images