Kick threshold adjustment method based on UWB radar, storage medium and equipment

Abstract

The invention relates to the technical field of vehicle-mounted intelligent auxiliary driving, and provides a UWB radar-based kicking threshold value adjustment method, a storage medium and equipment, after kicking action feature data is obtained, a noise level and a noise coefficient are further calculated to obtain a threshold value set, and a target detection threshold value is obtained by combining the threshold value set with a threshold value adjustment coefficient; therefore, the obtained target detection threshold value can be flexibly adjusted according to the characteristics of the actual kicking action and the environmental condition, the applicability and flexibility of the system are effectively improved, and then the accuracy and reliability of kicking detection are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle-mounted intelligent assisted driving technology, and in particular to a kick threshold adjustment method based on UWB radar, a storage medium and a device. Background Art

[0002] At present, capacitive sensors or ultra-wideband technology are often used in the automatic opening system of the car trunk to realize kick detection. However, these methods still have some shortcomings in practical applications.

[0003] Capacitive sensors mainly rely on detecting the capacitance change caused by the human body approaching to determine whether to perform the opening action. Although this method is low-cost, it is prone to false detection in complex environments. For example, when a user walks near the rear of the car, the system may mistake it for a kicking action, thereby erroneously triggering the trunk to open. In addition, capacitive sensors are sensitive to environmental factors such as humidity and temperature changes, which will also affect the accuracy and reliability of detection.

[0004] In contrast, UWB technology uses radio waves to achieve high-precision distance measurement and has better anti-interference capabilities against environmental interference. However, existing UWB-based kick detection algorithms still face challenges in handling diverse usage scenarios. Specifically, these algorithms are prone to false detection when faced with non-target actions such as reciprocating walking, resulting in false triggering of the system.

[0005] In addition, the fixed threshold method lacks flexibility and is difficult to maintain consistent detection performance across different users and in different environments, while manually setting the threshold cannot achieve real-time updates or adapt to different detection environments. Summary of the invention

[0006] The present invention provides a kick threshold adjustment method, storage medium and device based on UWB radar, which solves the technical problem that the existing kick detection technology cannot automatically adjust the kick threshold, thereby causing the kick detection to be unable to perform scene adaptation and poor detection accuracy.

[0007] In order to solve the above technical problems, the present invention provides a kick threshold adjustment method based on UWB radar, comprising:

[0008] Using the UWB radar function to collect kicking data, and performing feature extraction on the kicking data to obtain kicking features;

[0009] Calculating a threshold adjustment coefficient according to the kicking feature;

[0010] Based on the kicking characteristics, a reference unit and a protection unit are determined, and a noise level and a noise coefficient are calculated;

[0011] Calculating a threshold set corresponding to the current window according to the noise level and the noise coefficient;

[0012] A target detection threshold is obtained according to the threshold set and the threshold adjustment coefficient.

[0013] After obtaining the characteristic data of the kicking action, this basic scheme further calculates the noise level and noise coefficient to obtain the threshold set, and obtains the target detection threshold by combining the threshold set with the threshold adjustment coefficient. In this way, the obtained target detection threshold can be flexibly adjusted according to the characteristics of the actual kicking action and environmental conditions, which effectively improves the applicability and flexibility of the system, and thus significantly improves the accuracy and reliability of kicking detection.

[0014] In a further embodiment, the calculating of the threshold adjustment coefficient according to the kicking feature comprises:

[0015] Obtaining the characteristic length of the kicking feature and determining a proportionality coefficient;

[0016] Calculating an average value of the kicking features, and further calculating a square root of the average value;

[0017] The kicking features are summed, and the threshold adjustment coefficient is calculated by combining the square root and the proportional coefficient.

[0018] This solution determines the corresponding proportional coefficient based on the actual length of the kicking feature data collected; at the same time, based on the kicking feature, calculates the square root of the average value to measure the discreteness of the data; sums the kicking features; combines the square root, the sum of the data and the proportional coefficient to calculate the threshold adjustment coefficient, and starts from the kicking feature data itself to make targeted adjustments to the threshold adjustment coefficient, which can be more in line with the actual current kicking feature recognition.

[0019] In a further embodiment, the calculating the noise level comprises:

[0020] Perform parameter settings and determine the number of reference units and protection units;

[0021] Determining a threshold starting position in the kick feature according to the number of reference units, the number of protection units and the feature length of the kick feature;

[0022] Traverse each detection unit, calculate the left and right reference noises and take the minimum value as the noise level.

[0023] In a further embodiment, the noise level is calculated as follows:

[0024] avg_soca=min((sum(Kick_fea(k-Nguard-Nref:k-Nguard-1))),sum(Kick_fea(k+Nguard+1:k+Nguard+Nref))) / (2*Nref);

[0025] Wherein, avg_soca represents the noise level, min() represents the minimum value, sum() represents the sum of all parameters, mean() function represents the average value, Nref is the number of reference units, Nguard is the number of guard units, Kick_fea is the kick feature; k is the current detection unit.

[0026] This solution can effectively reduce the impact of noise on the signal by calculating the left and right reference noises and taking the minimum value as the noise level, thereby improving the quality of the signal.

[0027] In a further embodiment, the calculation of the noise figure includes:

[0028] Based on the kick feature, the square of the sum of the powers of all the reference units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the reference unit energy;

[0029] Based on the kicking feature, the square of the sum of the powers of all the protection units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the protection unit energy;

[0030] A noise figure is calculated based on the reference cell energy and the protection cell energy.

[0031] In a further embodiment, the noise factor is calculated as follows:

[0032] alpha_soca=alpha*(log(bg_erg_ref) / log(bg_erg_gud)).

[0033] Wherein, alpha_soca represents the noise coefficient, alpha is a preset reference value, bg_erg_ref is the reference unit energy, bg_erg_gud is the protection unit energy, and log() represents a logarithmic operation.

[0034] This scheme calculates the reference unit energy and the protection unit energy by using the power average of the reference unit and the protection unit, which are used to calculate the noise factor. Since the reference unit usually selects signals in an environment or conditions similar to those of the detection unit for calibration and comparison, the influence of environmental noise on the measurement results is eliminated or reduced; the protection unit is used to isolate the detection unit to prevent external interference, which further improves the accuracy of the noise factor calculation, can more accurately estimate the system noise level, and maintain the stability and reliability of the noise factor calculation.

[0035] In a further embodiment, the threshold value set is calculated as follows:

[0036] T_soca=alpha_soca*avg_soca

[0037] Among them, the number of thresholds is last_num-first_num+1;

[0038] Where T_soca represents the threshold set, alpha_soca is the noise coefficient, avg_soca is the noise level, first_num and ast_num are the threshold starting positions.

[0039] By calculating the threshold set according to the noise coefficient and noise level, the CFAR processing can adapt to different environmental conditions, improve the accuracy and reliability of signal detection, and ensure the performance of the radar system and the correct detection of targets in complex environments.

[0040] In a further embodiment, obtaining a target detection threshold according to the threshold set and the threshold adjustment coefficient includes:

[0041] Performing averaging processing on the threshold set to obtain a threshold average value representing an average level of the thresholds;

[0042] According to the threshold average value and the threshold adjustment coefficient, the target detection threshold corresponding to the current kicking feature is calculated. The calculation formula of the target detection threshold is as follows:

[0043] T_soca_fit=T_cofe / F_cofe

[0044] Where T_soca_fit represents the target detection threshold, T_cofe represents the threshold average, and F_cofe represents the threshold adjustment coefficient.

[0045] This scheme averages the threshold set to represent the average level of the threshold and reduces excessive fluctuations in the threshold. It then introduces a threshold adjustment coefficient derived from the kicking feature to calculate the target detection threshold, which can dynamically adjust the threshold in real time based on the feature.

[0046] The present invention provides a storage medium, on which a computer program or instruction is stored, and the computer program or instruction is executed by a processor to implement the steps of the UWB radar-based kick threshold adjustment method as described above. The storage medium may be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0047] The present invention also provides a device, comprising: a processor, a memory, and a computer program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed by the processor, the steps of the UWB radar-based kick threshold adjustment method as described above are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 The present invention provides a system flow chart of a method for adjusting a kick threshold based on a UWB radar according to an embodiment of the present invention. DETAILED DESCRIPTION

[0049] 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.

[0050] Example 1

[0051] An embodiment of the present invention provides a kick threshold adjustment method based on UWB radar, such as Figure 1 As shown, in this embodiment, it includes:

[0052] S1, using the UWB radar function to collect kicking data, and extracting features from the kicking data to obtain kicking features Kick_fea;

[0053] The specific feature extraction method is a conventional technology in the art and will not be described in detail in this embodiment.

[0054] S2A, calculating a threshold adjustment coefficient according to the kicking feature, including:

[0055] Obtaining the characteristic length of the kicking feature and determining a proportionality coefficient;

[0056] Calculating an average value of the kicking features, and further calculating a square root of the average value;

[0057] The kicking features are summed, and the threshold adjustment coefficient is calculated by combining the square root and the proportional coefficient.

[0058] Specifically, the calculation formula of the threshold adjustment coefficient is as follows:

[0059] F_cofe=a*sqrt(mean(Kick_fea))*sum(Kick_fea)

[0060] Wherein, F_cofe is the threshold adjustment coefficient, Kick_fea represents the kick feature, a represents the proportional coefficient related to the length of the kick feature Kick-fea, sqrt() represents the square root, and sum() represents the calculation of the sum of all parameters.

[0061] Among them, the proportional coefficient a can be adaptively adjusted according to the relationship with the kick feature kick-fea length, and the specific adjustment range can be simulated according to the training data.

[0062] This embodiment determines the corresponding proportional coefficient based on the length of the kicking feature data actually collected; at the same time, based on the kicking feature, calculates the square root of the average value to measure the discreteness of the data; sums the kicking features; calculates the threshold adjustment coefficient by combining the square root, the sum of the data and the proportional coefficient, and performs targeted threshold adjustment coefficient adjustment based on the kicking feature data itself, which can be more in line with the actual current kicking feature recognition.

[0063] S2B, calculating the threshold of each sub-window according to the kicking feature to obtain a threshold set;

[0064] B1. Based on the kicking feature, determine the reference unit and the protection unit, and calculate the noise level and noise coefficient;

[0065] In this embodiment, the calculating of the noise level includes:

[0066] B11. Set parameters to determine the number of reference units and the number of protection units; for example, determine the number of cfar reference units Nref=5 and the number of protection units Nguard=35.

[0067] A certain number of reference cells are selected around the detection cell. These reference cells are used to estimate the level of background noise and interference. The number and location of reference cells are crucial for accurately estimating background noise.

[0068] B12, determining a threshold starting position in the kick feature according to the number of reference units, the number of protection units and the characteristic length of the kick feature;

[0069] Specifically, the threshold starting position is (first_num=Nref+Nguard+1, last_num=Nz-Nref-Nguard) where Nz is the characteristic length for cfar detection.

[0070] B13, let k be the current detection unit (i.e. the current threshold calculation point), first num ≤k≤last num ; Traverse each detection unit, calculate the left and right reference noises and take the minimum value as the noise level.

[0071] In this embodiment, the calculation formula of the noise level is as follows:

[0072] avg_soca=min((sum(Kick_fea(k-Nguard-Nref:k-Nguard-1))),sum(Kick_fea(k+Nguard+1:k+Nguard+Nref))) / (2*Nref);

[0073] Wherein, avg_soca represents the noise level, min() represents the minimum value, sum() represents the sum of all parameters, mean() function represents the average value, Nref is the number of reference units, Nguard is the number of guard units, Kick_fea is the kick feature; k is the current detection unit.

[0074] This embodiment calculates the left and right reference noises and takes the minimum value as the noise level, wherein the noise level is used to estimate the noise and interference level of the environment surrounding the detection unit, thereby effectively reducing the impact of noise on the signal and improving the quality of the signal.

[0075] In this embodiment, the calculation of the noise coefficient includes:

[0076] B21. Based on the kicking feature, the square of the sum of the powers of all the reference units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the reference unit energy. The calculation formula is as follows;

[0077] bg_erg_ref=((sum(Kick_fea(k-Nguard-Nref:k-Nguard-1))^2+(sum(Kick_fea(k+Nguard+1:k+Nguard+Nref)))^2) / (2*Nref))

[0078] B22. Based on the kicking feature, the square of the sum of the powers of all the protection units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the protection unit energy. The calculation formula is as follows;

[0079] bg_erg_gud=((sum(Kick_fea(k-Nguard:k-1))^2+(sum(Kick_fea(k+1:k+Nguard)))^2) / (2*Nguard))

[0080] B23. Calculate a noise coefficient according to the reference unit energy and the protection unit energy.

[0081] In this embodiment, the calculation formula of the noise coefficient is as follows:

[0082] alpha_soca=alpha*(log(bg_erg_ref) / log(bg_erg_gud)).

[0083] Wherein, alpha_soca represents the noise coefficient, alpha is a preset reference value, bg_erg_ref is the reference unit energy, bg_erg_gud is the protection unit energy, and log() represents a logarithmic operation.

[0084] In this embodiment, the log function is used to suppress excessive data jumps, and the reference unit is used to divide the protection unit. The purpose is that if the current position is at the peak of the kicking position, alpha_soca will be reduced at this time, thereby improving the detection rate.

[0085] This embodiment uses the power average of the reference unit and the protection unit to calculate the reference unit energy and the protection unit energy for calculating the noise factor. Since the reference unit usually selects a signal in an environment or condition similar to that of the detection unit for calibration and comparison, the influence of environmental noise on the measurement result is eliminated or reduced; the protection unit is used to isolate the detection unit to prevent external interference, further improving the accuracy of the noise factor calculation, being able to more accurately estimate the system noise level, and maintaining the stability and reliability of the noise factor calculation.

[0086] B3. Calculate a threshold set corresponding to the current window according to the noise level and the noise coefficient.

[0087] In this embodiment, the calculation formula of the threshold set is as follows:

[0088] T_soca=alpha_soca*avg_soca

[0089] Among them, the number of thresholds is last_num-first_num+1;

[0090] Where T_soca represents the threshold set, alpha_soca is the noise coefficient, avg_soca is the noise level, first_num and ast_num are the threshold starting positions.

[0091] In this embodiment, by calculating the threshold set according to the noise coefficient and the noise level, the CFAR processing can adapt to different environmental conditions, improve the accuracy and reliability of signal detection, and ensure the performance of the radar system and the correct detection of targets in complex environments.

[0092] S3. Obtaining a target detection threshold according to the threshold set and the threshold adjustment coefficient, including:

[0093] Performing averaging processing on the threshold set to obtain a threshold average value representing an average level of the thresholds;

[0094] According to the threshold average value and the threshold adjustment coefficient, the target detection threshold corresponding to the current kicking feature is calculated. The calculation formula of the target detection threshold is as follows:

[0095] T_soca_fit=T_cofe / F_cofe

[0096] Where T_soca_fit represents the target detection threshold, T_cofe represents the threshold average, and F_cofe represents the threshold adjustment coefficient.

[0097] This embodiment averages the threshold set to represent the average level of the threshold, while reducing excessive fluctuations in the threshold; then, a threshold adjustment coefficient derived from the kicking feature is introduced to calculate the target detection threshold, which can dynamically adjust the threshold in real time based on the feature.

[0098] In this embodiment, the target detection threshold is used in synchronization with the feature to determine whether there is a kicking action. For example, the signal amplitude of the detection unit is compared with the calculated threshold. If the signal amplitude exceeds the threshold, it is considered that the target is detected; otherwise, it is considered that there is no target.

[0099] After obtaining the characteristic data of the kicking action, the embodiment of the present invention further calculates the noise level and the noise coefficient to obtain a threshold set, and obtains the target detection threshold by combining the threshold set with the threshold adjustment coefficient; in this way, the obtained target detection threshold can be flexibly adjusted according to the characteristics of the actual kicking action and environmental conditions, which effectively improves the applicability and flexibility of the system, and thus significantly improves the accuracy and reliability of kicking detection.

[0100] Example 2

[0101] An embodiment of the present invention provides a storage medium, on which a computer program or instruction is stored, and the computer program or instruction is executed by a processor to implement the steps of the kick threshold adjustment 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.

[0102] Example 3

[0103] An embodiment of the present invention also provides a device, comprising: a processor, a memory, and a computer program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the kick threshold adjustment method based on UWB radar as described in Example 1 above.

[0104] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A kick threshold adjustment method based on UWB radar, characterized in that: include: Using the UWB radar function to collect kicking data, and performing feature extraction on the kicking data to obtain kicking features; Calculating a threshold adjustment coefficient according to the kicking feature; Based on the kicking characteristics, a reference unit and a protection unit are determined, and a noise level and a noise coefficient are calculated; Calculating a threshold set corresponding to the current window according to the noise level and the noise coefficient; A target detection threshold is obtained according to the threshold set and the threshold adjustment coefficient.

2. A kick threshold adjustment method based on UWB radar as claimed in claim 1, characterized in that: The calculating of the threshold adjustment coefficient according to the kicking feature comprises: Obtaining the characteristic length of the kicking feature and determining a proportionality coefficient; Calculating an average value of the kicking features, and further calculating a square root of the average value; The kicking features are summed, and the threshold adjustment coefficient is calculated by combining the square root and the proportional coefficient.

3. A kick threshold adjustment method based on UWB radar as claimed in claim 2, characterized in that: The calculating noise level comprises: Perform parameter settings and determine the number of reference units and protection units; Determining a threshold starting position in the kick feature according to the number of reference units, the number of protection units and the feature length of the kick feature; Traverse each detection unit, calculate the left and right reference noises and take the minimum value as the noise level.

4. A kick threshold adjustment method based on UWB radar as claimed in claim 3, characterized in that: The noise level is calculated as follows: avg_soca=min((sum(Kick_fea(k-Nguard-Nref:k-Nguard-1))),sum(Kick_fea(k+Nguard+1:k+Nguard+Nref))) / (2*Nref); Wherein, avg_soca represents the noise level, min() represents the minimum value, sum() represents the sum of all parameters, mean() function represents the average value, Nref is the number of reference units, Nguard is the number of guard units, Kick_fea is the kick feature; k is the current detection unit.

5. A kick threshold adjustment method based on UWB radar as claimed in claim 4, characterized in that: The calculation of the noise figure includes: Based on the kick feature, the square of the sum of the powers of all the reference units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the reference unit energy; Based on the kicking feature, the square of the sum of the powers of all the protection units on both sides of the detection unit is calculated within the detection window, and then the power average is calculated as the protection unit energy; A noise figure is calculated based on the reference cell energy and the protection cell energy.

6. A kick threshold adjustment method based on UWB radar as claimed in claim 5, characterized in that: The noise factor is calculated as follows: alpha_soca=alpha*(log(bg_erg_ref) / log(bg_erg_gud)). Wherein, alpha_soca represents the noise coefficient, alpha is a preset reference value, bg_erg_ref is the reference unit energy, bg_erg_gud is the protection unit energy, and log() represents a logarithmic operation.

7. A kick threshold adjustment method based on UWB radar as claimed in claim 1, characterized in that: The calculation formula of the threshold set is as follows: T_soca=alpha_soca*avg_soca Among them, the number of thresholds is last_num-first_num+1; Where T_soca represents the threshold set, alpha_soca is the noise coefficient, avg_soca is the noise level, first_num and ast_num are the threshold starting positions.

8. The method for adjusting the kick threshold based on UWB radar as claimed in claim 1, characterized in that: Obtaining a target detection threshold according to the threshold set and the threshold adjustment coefficient includes: Performing averaging processing on the threshold set to obtain a threshold average value representing an average level of the thresholds; According to the threshold average value and the threshold adjustment coefficient, the target detection threshold corresponding to the current kicking feature is calculated. The calculation formula of the target detection threshold is as follows: T_soca_fit=T_cofe / F_cofe Where T_soca_fit represents the target detection threshold, T_cofe represents the threshold average, and F_cofe represents the threshold adjustment coefficient.

9. A storage medium, characterized in that: The storage medium stores a computer program or instruction, and the computer program or instruction is executed by a processor to implement the steps of the UWB radar-based kick threshold adjustment method as described in any one of claims 1-8.

10. A device, characterized in that: include: A processor, a memory, and a computer program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the kick threshold adjustment method based on UWB radar as described in any one of claims 1 to 8.

Images