Millimeter wave radar respiration rate detection method and system based on wideband spectrum weighting

Abstract

This invention discloses a method and system for detecting breathing frequency in millimeter-wave radar based on broadband spectrum weighting. The method includes: S1, performing range-dimensional FFT on the ADC signal to obtain a range-dimensional FFT result; S2, performing motion-static separation on the range-dimensional FFT result to extract the vital signal from the static part; S3, performing motion-static separation across multiple frames to distinguish between stationary objects and slightly moving targets; S4, selecting range cells from the motion-static separation results across multiple frames and extracting the phase within the range cell; S5, performing phase unwrapping and filtering on the phase to obtain a breathing signal; S6, performing FFT calculation on the breathing signal to obtain the spectrum of the breathing signal; S7, calculating the breathing frequency based on the spectrum. This invention can improve the accuracy of breathing frequency detection.

Description

Technical Field

[0001] This invention relates to the field of millimeter-wave radar life signal detection, and in particular to a method and system for detecting respiratory frequency using millimeter-wave radar based on broadband spectrum weighting. Background Technology

[0002] Existing respiratory frequency detection methods based on millimeter-wave radar often select a single frequency point within its spectrum as the respiratory frequency value of the target being detected. However, the human respiratory frequency changes over time and is not a single frequency signal within a certain time window. Directly using the maximum peak value in the spectrum as the respiratory frequency for the entire time period may result in a relatively large error and cannot achieve accurate respiratory frequency detection.

[0003] Therefore, there is an urgent need to propose a method and system for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting in order to effectively solve the above problems. Summary of the Invention

[0004] The purpose of this invention is to propose a method and system for detecting breathing frequency using millimeter-wave radar based on broadband spectrum weighting, which can improve the accuracy of breathing frequency detection.

[0005] To address the aforementioned technical problems, this invention provides a method for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting, comprising:

[0006] S1. Perform a distance-dimensional FFT on the ADC signal to obtain a distance-dimensional FFT result;

[0007] S2. Perform dynamic-static separation on the distance dimension FFT result and extract the life signal from the static part;

[0008] S3. Separation of motion and stillness across multiple frames, distinguishing between stationary objects and slightly moving targets;

[0009] S4. Select a distance unit from the motion-static separation results between the multiple frames and extract the phase within the distance unit;

[0010] S5. Perform phase unwrapping and filtering on the phase to obtain a breathing signal;

[0011] S6. Perform FFT calculation on the respiratory signal to obtain the spectrum of the respiratory signal;

[0012] S7. Calculate the respiratory rate based on the spectrum.

[0013] Furthermore, in S2, an average value is calculated for each distance cell, and the average value is retained for static clutter removal in subsequent frames.

[0014] Furthermore, S3 includes:

[0015] S31. Apply a sliding window to the static portion extracted in S2;

[0016] S32. Calculate the average value of all distance units within the sliding window;

[0017] S33, Subtract the average value from all distance units within the sliding window;

[0018] S34. The absolute values ​​of each distance unit within the sliding window are calculated and then summed.

[0019] Furthermore, S4 includes: finding the peak with the largest amplitude from the result of destatic clutter removal in the multi-frame S3, the peak with the largest amplitude being the range cell where the micro-moving target is located, and extracting the phase within the range cell.

[0020] Furthermore, step S6 includes: windowing the respiratory signal and performing FFT calculation to obtain the spectrum of the respiratory signal.

[0021] Furthermore, S7 includes:

[0022] S71. Select the center frequency based on the breathing frequency of the previous frame and the maximum peak value within the breathing spectrum range of the current frame.

[0023] S72. Select the peak value of the spectrum within a certain range of the center frequency;

[0024] S73. Calculate the total amplitude of the peak values;

[0025] S74. Calculate the ratio of the amplitude of each peak value to the total amplitude, obtain the respiratory rate by weighting according to the ratio, and record the amplitude ratio corresponding to the maximum peak value.

[0026] S75. The final respiratory rate result is output by combining the respiratory rate calculated in S74 and the respiratory rate of the previous frame.

[0027] Furthermore, S71 includes:

[0028] S711. Within the range of the respiratory rate, find the location of the peak with the largest amplitude;

[0029] S712. Compare the frequency represented by the location of the peak with the largest amplitude with the respiratory frequency measured in the previous frame. If it is greater than the set threshold fthr, then the respiratory frequency of the previous frame is selected as the frequency; if it is less than the set threshold fthr, then the peak with the largest amplitude within the respiratory range is selected as the frequency center.

[0030] Furthermore, S72 includes:

[0031] S721. Find all peak values ​​within the respiratory rate range and calculate the frequency difference between each peak value and the center frequency;

[0032] S722. Peak values ​​that meet certain conditions are used for subsequent respiratory rate calculations. Two conditions must be met: the absolute value of the difference between the frequency corresponding to the peak value and the center frequency is less than the set threshold fthr1, and the peak amplitude is greater than the set amplitude threshold Vthr.

[0033] Furthermore, S74 includes:

[0034] S741. Calculate the proportion of each peak amplitude to the sum of amplitudes calculated in S73;

[0035] S742. Determine whether the proportion of the peak with the largest peak amplitude to the total amplitude is greater than the set threshold. If it is greater than the set threshold, directly output the frequency represented by the peak as the respiratory rate. If it is less than the set threshold, continue to calculate the respiratory rate.

[0036] S743. Calculate the respiratory rate by weighting the values ​​of each peak value.

[0037] Bfre = fre1*scale1 + fre2*scale2 + fre3*scale3...; records the maximum amplitude ratio; where Bfre represents the respiratory rate; fre1, fre2, and fre3 represent the frequencies of each peak; and scale1, scale2, and scale3 represent the percentage of each peak.

[0038] Furthermore, S75 includes: firstly, determining the weighting ratio of the historical respiratory rate value and the current calculated value based on the amplitude ratio corresponding to the maximum peak value. The larger the amplitude ratio corresponding to the maximum peak value, the larger the weighting coefficient of the current calculated value; the smaller the amplitude ratio corresponding to the maximum peak value, the smaller the weighting coefficient of the current frame calculated value.

[0039] Furthermore, this invention also proposes a millimeter-wave radar breathing frequency detection system based on broadband spectrum weighting, employing the millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting as described above, including:

[0040] The calculation module is used to perform a distance-dimensional FFT on the ADC signal;

[0041] The motion-static separation module is used to perform motion-static separation on the results of the distance-dimensional FFT, extract the life signal from the static part, and select a distance unit from the motion-static separation results across multiple frames to extract the phase within the distance unit.

[0042] The calculation module is also used to perform phase unwrapping and filtering on the phase to obtain a breathing signal; and to perform FFT calculation on the breathing signal to obtain the spectrum of the breathing signal; and to calculate the breathing frequency based on the spectrum.

[0043] Through the above technical solution, the present invention has the following beneficial effects:

[0044] A distance-dimensional FFT is performed on the ADC signal to obtain a distance-dimensional FFT result. The distance-dimensional FFT result is then subjected to motion-static separation to extract vital signals from the static portion. Motion-static separation is performed across multiple frames to distinguish between stationary objects and slightly moving targets. Range cells are selected from the motion-static separation results across multiple frames, and the phase within each range cell is extracted. The phase is then unwrapped and filtered to obtain a breathing signal. An FFT is performed on the breathing signal to obtain its spectrum. The breathing frequency is then calculated based on the spectrum. Since the breathing frequency may vary over a certain period, weighting multiple peaks in its spectrum can yield a more accurate breathing frequency, thereby improving the accuracy of breathing frequency detection. Attached Figure Description

[0045] Figure 1 This is a flowchart of a millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting in one embodiment of the present invention;

[0046] Figure 2 This is a flowchart of step S7 in a millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting in one embodiment of the present invention. Detailed Implementation

[0047] The following description, in conjunction with the accompanying drawings, provides a more detailed account of a method and system for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting, which illustrates preferred embodiments of the invention. It should be understood that those skilled in the art can modify the invention described herein while still achieving its advantageous effects. Therefore, the following description should be understood as being of general knowledge to those skilled in the art and is not intended to limit the invention.

[0048] The invention is described more specifically by way of example in the following paragraphs with reference to the accompanying drawings. The advantages and features of the invention will become clearer from the following description. It should be noted that the drawings are in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the invention.

[0049] like Figure 1 As shown, this embodiment of the invention proposes a method for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting, including:

[0050] S1. Perform a distance-dimensional FFT (Fast Fourier Transform) on the ADC signal to obtain a distance-dimensional FFT result;

[0051] S2. Perform dynamic-static separation on the distance dimension FFT result and extract the life signal from the static part;

[0052] S3. Separation of motion and stillness across multiple frames, distinguishing between stationary objects and slightly moving targets;

[0053] S4. Select a distance unit from the motion-static separation results between the multiple frames and extract the phase within the distance unit;

[0054] S5. Perform phase unwrapping and filtering on the phase to obtain a breathing signal;

[0055] S6. Perform FFT calculation on the respiratory signal to obtain the spectrum of the respiratory signal;

[0056] S7. Calculate the respiratory rate based on the spectrum.

[0057] Specifically, in S2, the average value is calculated for each distance cell, and the average value is retained for static clutter removal in subsequent frames.

[0058] In this embodiment, S3 includes:

[0059] S31. Apply a sliding window to the static portion extracted in S2;

[0060] S32. Calculate the average value of all distance units within the sliding window;

[0061] S33, Subtract the average value from all distance units within the sliding window;

[0062] S34. The absolute values ​​of each distance unit within the sliding window are calculated and then summed.

[0063] In this embodiment, S4 includes: finding the peak with the largest amplitude from the result of destatic clutter removal in the multi-frame S3, the peak with the largest amplitude being the range cell where the micro-moving target is located, and extracting the phase within the range cell.

[0064] In addition, in this embodiment, step S6 includes: windowing the respiratory signal and performing FFT calculation to obtain the spectrum of the respiratory signal.

[0065] Combination Figure 2 As shown, in this embodiment, S7 includes:

[0066] S71. Select the center frequency based on the breathing frequency of the previous frame and the maximum peak value within the breathing spectrum range of the current frame.

[0067] S72. Select the peak value of the spectrum within a certain range of the center frequency;

[0068] S73. Calculate the total amplitude of the peak values;

[0069] S74. Calculate the ratio of the amplitude of each peak value to the total amplitude, obtain the respiratory rate by weighting according to the ratio, and record the amplitude ratio corresponding to the maximum peak value.

[0070] S75. The final respiratory rate result is output by combining the respiratory rate calculated in S74 and the respiratory rate of the previous frame.

[0071] In the first embodiment, S71 includes:

[0072] S711. Within the range of the respiratory rate, find the location of the peak with the largest amplitude;

[0073] S712. Compare the frequency represented by the location of the peak with the largest amplitude with the respiratory frequency measured in the previous frame. If it is greater than the set threshold fthr, then the respiratory frequency of the previous frame is selected as the frequency; if it is less than the set threshold fthr, then the peak with the largest amplitude within the respiratory range is selected as the frequency center.

[0074] In the second embodiment, S72 includes:

[0075] S721. Find all peak values ​​within the respiratory rate range and calculate the frequency difference between each peak value and the center frequency;

[0076] S722. Peak values ​​that meet certain conditions are used for subsequent respiratory rate calculations. Two conditions must be met: the absolute value of the difference between the frequency corresponding to the peak value and the center frequency is less than the set threshold fthr1, and the peak amplitude is greater than the set amplitude threshold Vthr.

[0077] In the third embodiment, S74 includes:

[0078] S741. Calculate the proportion of each peak amplitude to the sum of amplitudes calculated in S73;

[0079] S742. Determine whether the proportion of the peak with the largest peak amplitude to the total amplitude is greater than the set threshold. If it is greater than the set threshold, directly output the frequency represented by the peak as the respiratory rate. If it is less than the set threshold, continue to calculate the respiratory rate.

[0080] S743. Calculate the respiratory rate by weighting the values ​​of each peak value.

[0081] Bfre = fre1*scale1 + fre2*scale2 + fre3*scale3...; records the maximum amplitude ratio; where Bfre represents the respiratory rate; fre1, fre2, and fre3 represent the frequencies of each peak; and scale1, scale2, and scale3 represent the percentage of each peak.

[0082] In the fourth embodiment, S75 includes: firstly, determining the weighting ratio of the historical respiratory rate value and the current calculated value based on the amplitude ratio corresponding to the maximum peak value. The larger the amplitude ratio corresponding to the maximum peak value, the larger the weighting coefficient of the current calculated value; the smaller the amplitude ratio corresponding to the maximum peak value, the smaller the weighting coefficient of the current frame calculated value.

[0083] In a specific example, after obtaining the radar ADC data, the breathing rate is obtained through step-by-step processing of the ADC data. The specific implementation steps are as follows:

[0084] Step S1: Perform a distance-dimensional FFT on the ADC data, with N points in the FFT.

[0085] Step S2: Calculate the average value for each distance cell and retain the average value for static clutter removal in subsequent frames.

[0086] Step S3: Static noise removal between multiple frames. Static noise is removed between multiple frames with a window length of 15 frames. After sliding the window, the static part retained in step S2 is placed in the last unit of the time window.

[0087] Step S31: Apply a sliding window to the static portion extracted in step S2.

[0088] Step S32: Calculate the average value of all distance cells within the sliding window.

[0089] Step S33: Subtract the average value obtained in step S32 from all distance units within the sliding window.

[0090] Step S34: Calculate the absolute value of each distance cell within the sliding window and then add them together.

[0091] Step S4: Find the peak with the largest amplitude from the results of multi-frame static clutter removal in Step S3. The peak with the largest amplitude is the range cell where the micro-moving target is located, and extract the phase within that range cell.

[0092] Step S5: Extract the obtained phase, combine it with the historical phase to perform phase unwrapping, and perform bandpass filtering on the unwrapped phase with a passband range of 0.1-0.6. The filtered phase is used to obtain the respiratory signal.

[0093] Step S6: Window the respiratory signal, and then perform a 256-point FFT after adding a Hamming window to obtain the spectrum of the respiratory signal.

[0094] Step S7: Calculate the respiratory rate based on the spectral distribution of the respiratory signal.

[0095] Step S71: Find the maximum peak value within the respiratory spectrum range. If the difference between this maximum peak value and the respiratory frequency of the previous frame exceeds 4 frequency units, select the respiratory frequency of the previous frame as the center frequency; otherwise, select the maximum peak value as the center frequency.

[0096] Step S72: Find all peaks within the respiratory spectrum range, and select peaks with an amplitude greater than 15 and a distance of less than 6 frequency units from the center frequency for subsequent respiratory rate calculation.

[0097] Step S73: Calculate the total amplitude of the selected peak values.

[0098] Step S74: Calculate the ratio of all peak amplitudes to the total amplitude sum, and record the largest amplitude ratio. If the largest amplitude ratio is greater than 0.5, select the peak with the largest amplitude as the respiratory rate; otherwise, use the amplitude ratio of each peak as a weighting coefficient to calculate the respiratory rate.

[0099] Step S75: If the largest amplitude percentage exceeds 0.5, the weighting coefficients of the current frame breathing frequency and the historical breathing frequency are 0.5 and 0.5, respectively; if the largest amplitude percentage is between 0.35 and 0.5, the weighting coefficients of the current frame breathing frequency and the historical breathing frequency are 0.3 and 0.8, respectively; if the largest amplitude percentage is less than 0.35, the weighting coefficients of the current frame breathing frequency and the historical breathing frequency are 0.1 and 0.9, respectively.

[0100] In this embodiment, a range-dimensional FFT is performed on the ADC signal to obtain a range-dimensional FFT result; motion-static separation is performed on the range-dimensional FFT result to extract the vital signal from the static part; motion-static separation is performed across multiple frames to distinguish between stationary objects and slightly moving targets; a range cell is selected from the motion-static separation results across multiple frames to extract the phase within the range cell; the phase is unwrapped and filtered to obtain a breathing signal; an FFT is performed on the breathing signal to obtain the spectrum of the breathing signal; and the breathing frequency is calculated based on the spectrum.

[0101] Furthermore, this embodiment also proposes a millimeter-wave radar breathing frequency detection system based on broadband spectrum weighting, employing the millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting as described above, including:

[0102] The calculation module is used to perform a range-dimensional FFT on the ADC signal; the motion-static separation module is used to perform motion-static separation on the result of the range-dimensional FFT, extract the vital signal from the static part; and select a distance unit from the motion-static separation results across multiple frames, extract the phase within the distance unit; the calculation module is also used to perform phase unwrapping and filtering on the phase to obtain a breathing signal; and perform FFT calculation on the breathing signal to obtain the spectrum of the breathing signal; and calculate the breathing frequency based on the spectrum.

[0103] In summary, the millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting proposed in this invention has the following advantages:

[0104] A distance-dimensional FFT is performed on the ADC signal to obtain a distance-dimensional FFT result. The distance-dimensional FFT result is then subjected to motion-static separation to extract vital signals from the static portion. Motion-static separation is performed across multiple frames to distinguish between stationary objects and slightly moving targets. Range cells are selected from the motion-static separation results across multiple frames, and the phase within each range cell is extracted. The phase is then unwrapped and filtered to obtain a breathing signal. An FFT is performed on the breathing signal to obtain its spectrum. The breathing frequency is then calculated based on the spectrum. Since the breathing frequency may vary over a certain period, weighting multiple peaks in its spectrum can yield a more accurate breathing frequency, thereby improving the accuracy of breathing frequency detection.

[0105] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for detecting breathing frequency using millimeter-wave radar based on broadband spectrum weighting, characterized in that, include: S1. Perform a distance-dimensional FFT on the ADC signal to obtain a distance-dimensional FFT result; S2. Perform dynamic-static separation on the distance dimension FFT result and extract the life signal from the static part; S3. Separation of motion and stillness across multiple frames, distinguishing between stationary objects and slightly moving targets; S4. Select a distance unit from the motion-static separation results between the multiple frames and extract the phase within the distance unit; S5. Perform phase unwrapping and filtering on the phase to obtain a breathing signal; S6. Perform FFT calculation on the respiratory signal to obtain the spectrum of the respiratory signal; S7. Calculate the respiratory frequency based on the spectrum, specifically including: S71. Selecting a center frequency based on the respiratory frequency of the previous frame and the maximum peak value within the respiratory spectrum range of the current frame; S72. Selecting spectral peak values ​​within a certain range of the center frequency; S73. Calculating the total amplitude of the peak values; S74. Calculating the ratio of the amplitude of each spectral peak value to the total amplitude, weighting the values ​​according to the ratio to obtain the respiratory frequency, and recording the amplitude ratio corresponding to the maximum peak value; S75. Outputting the final respiratory frequency result by combining the respiratory frequency calculated in S74 and the respiratory frequency of the previous frame.

2. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, In S2, the average value is calculated for each distance cell, and the average value is retained for static clutter removal in subsequent frames.

3. The method for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S3 includes: S31. Apply a sliding window to the static portion extracted in S2; S32. Calculate the average value of all distance units within the sliding window; S33, Subtract the average value from all distance units within the sliding window; S34. The absolute values ​​of each distance unit within the sliding window are calculated and then summed.

4. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S4 includes: finding the peak with the largest amplitude from the result of destatic clutter removal in S3, the peak with the largest amplitude being the range cell where the micro-moving target is located, and extracting the phase within the range cell.

5. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S6 includes: windowing the respiratory signal and performing FFT calculation to obtain the spectrum of the respiratory signal.

6. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S71 includes: S711. Within the range of the respiratory rate, find the location of the peak with the largest amplitude; S712. Compare the frequency represented by the location of the peak with the largest amplitude with the respiratory frequency measured in the previous frame. If it is greater than the set threshold fthr, then the respiratory frequency of the previous frame is selected as the frequency; if it is less than the set threshold fthr, then the peak with the largest amplitude within the respiratory range is selected as the frequency center.

7. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S72 includes: S721. Find all peak values ​​within the respiratory rate range and calculate the frequency difference between each peak value and the center frequency; S722. Peak values ​​that meet certain conditions are used for subsequent respiratory rate calculations. Two conditions must be met: the absolute value of the difference between the frequency corresponding to the peak value and the center frequency is less than the set threshold fthr1, and the peak amplitude is greater than the set amplitude threshold Vthr.

8. The method for detecting breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 1, characterized in that, S74 includes: S741. Calculate the proportion of each peak amplitude to the sum of amplitudes calculated in S73; S742. Determine whether the proportion of the peak with the largest peak amplitude to the total amplitude is greater than the set threshold. If it is greater than the set threshold, directly output the frequency represented by the peak as the respiratory rate. If it is less than the set threshold, continue to calculate the respiratory rate. S743. Calculate the respiratory rate by weighting the values ​​of each peak value. Record the maximum amplitude ratio; where Bfre represents respiratory rate; fre1, fre2, and fre3 represent the frequencies of each peak; and scale1, scale2, and scale3 represent the percentage of each peak.

9. The method for detecting the breathing frequency of millimeter-wave radar based on broadband spectrum weighting as described in claim 8, characterized in that, S75 includes: First, it is necessary to determine the weighting ratio of the historical respiratory rate value and the current calculated value based on the amplitude ratio corresponding to the maximum peak value. The larger the amplitude ratio corresponding to the maximum peak value, the larger the weighting coefficient of the current calculated value. The smaller the amplitude ratio corresponding to the maximum peak value, the smaller the weighting coefficient of the current frame calculated value.

10. A millimeter-wave radar breathing frequency detection system based on broadband spectrum weighting, employing the millimeter-wave radar breathing frequency detection method based on broadband spectrum weighting as described in any one of claims 1-9, characterized in that, include: The calculation module is used to perform a distance-dimensional FFT on the ADC signal; The motion-static separation module is used to perform motion-static separation on the results of the distance-dimensional FFT, extract the life signal from the static part, and select a distance unit from the motion-static separation results across multiple frames to extract the phase within the distance unit. The calculation module is also used to perform phase unwrapping and filtering on the phase to obtain a breathing signal; and to perform FFT calculation on the breathing signal to obtain the spectrum of the breathing signal; and to calculate the breathing frequency based on the spectrum.

Images