Method for precisely quantifying gait parameters using wi-fi

Abstract

This invention relates to a method for accurately quantifying gait parameters using Wi-Fi, comprising: (S1A) establishing a Wi-Fi space using a transmitter and a receiver, wherein the transmitter transmits a first wireless signal; (S2A) moving a human body from a starting point (S) to an ending point (E) within the Wi-Fi space, and receiving a second wireless signal using a receiver, and extracting channel state information from it. (S3A) Preprocess the Channel State Information (CSI) signal to obtain a denoised Channel State Information (CSI) signal; (S4A) Extract features from the denoised Channel State Information (CSI) signal to obtain human Channel State Information (CSI) gait parameters, including the original gait velocity curve, and the instantaneous gait velocity at the nth second on the original gait velocity curve is vf; (S5A) Obtain the correction multiplier (R); and (S6A) Use the correction multiplier (R) to correct the original gait velocity curve to obtain the corrected gait velocity curve, wherein the instantaneous gait velocity at the nth second on the corrected gait velocity curve is vg, and vg = vf. R.

Claims

1. A method for accurately quantifying gait parameters using Wi-Fi, comprising: (S1A) establishing a Wi-Fi space with a transmitter and a receiver, wherein the transmitter is used to transmit a first wireless signal; (S2A) moving a human body from a starting point (S) to an ending point (E) in the Wi-Fi space, and receiving a second wireless signal with the receiver, and extracting a channel state information (CSI) signal from it, wherein the second wireless signal is formed by the reflection of the first wireless signal by the human body; (S3A) preprocessing the CSI signal to obtain a CSI denoised signal; (S4A) performing feature extraction on the CSI denoised signal to obtain a human body CSI gait parameter, including an original gait speed curve, wherein the instantaneous gait speed at the nth second on the original gait speed curve is; (S5A) obtaining a correction multiplier (R) according to the following formula: , where: D is the straight-line distance between the starting point (S) and the ending point (E), A1 is the sum of the straight-line path lengths from the starting point (S) to the transmitter and the receiver respectively, and A2 is the sum of the straight-line path lengths from the ending point (E) to the transmitter and the receiver respectively; and (S6A) the original gait speed curve is corrected using the correction multiplier (R) to obtain a corrected gait speed curve, wherein the instantaneous gait speed at the nth second on the corrected gait speed curve is, and.

2. The method as described in claim 1, wherein: The transmitter and the receiver are connected by a line of sight (LoS), wherein: the length of the line of sight (LoS) is L, and 6 m ≥ L ≥ 2 m; and the extension direction of the line connecting the starting point (S) to the ending point (E) forms an angle θ with the line of sight (LoS), and 90° ≥ θ ≥ 45°.

3. The method as described in request item 1, wherein: A line of sight (LoS) is formed between the transmitter and the receiver, and the starting point (S) and the ending point (E) are located on the perpendicular line of the line of sight (LoS).

4. The method as described in claim 1, wherein step (S3A) includes: filtering high-frequency noise and low-frequency DC components in the channel state information (CSI) signal with a filter; and performing principal component analysis on the channel state information (CSI) signal to extract signals associated with human gait.

5. The method as described in claim 1, wherein step (S4A) comprises: performing a short-time Fourier transform (STFT) on the Channel State Information (CSI) noise-reduced signal to obtain a time-frequency map; tracking high-energy regions in the time-frequency map to form a trunk velocity profile on the time-frequency map; and converting the time-frequency map into a velocity-time graph to convert the trunk velocity profile into the original gait velocity curve and obtain the human channel state information (CSI) gait parameters.

6. A method for accurately quantifying gait parameters using Wi-Fi, comprising: (S1B) establishing a Wi-Fi space with a transmitter and a receiver, wherein the transmitter is used to transmit a first wireless signal; (S2B) causing a human body to move from a starting point (S) to an ending point (E) in the Wi-Fi space, and receiving a second wireless signal with the receiver, and extracting a channel state information (CSI) signal from it, wherein the second wireless signal is formed by the reflection of the first wireless signal by the human body; (S3B) Extract from the Channel Status Information (CSI) signal: the angle of arrival (AOAS) from the starting point (S) to the receiver, the angle of arrival (AOAE) from the ending point (E) to the receiver, the sum of the time taken for the first wireless signal to travel from the transmitter to the starting point (S) and the time taken for the second wireless signal to travel from the starting point (S) to the receiver (TOFS), and the sum of the time taken for the first wireless signal to travel from the transmitter to the ending point (E) and the time taken for the second wireless signal to travel from the ending point (E) to the receiver (TOFE); (S4B) Preprocess the Channel Status Information (CSI) signal to obtain a Channel Status Information (CSI) noise-reduced signal; (S5B) Perform feature extraction on the Channel Status Information (CSI) noise-reduced signal to obtain a human Channel Status Information (CSI) gait parameter, which includes an original gait speed curve, and the instantaneous gait speed at the nth second on the original gait speed curve is; (S6B) Obtain a correction multiplier (R) according to the following formula: , where c is the speed of light; and (S7B) using the correction multiplier (R) to correct the original gait velocity curve to obtain a corrected gait velocity curve, wherein the instantaneous gait velocity at the nth second on the corrected gait velocity curve is, and.

7. The method as described in claim 6, wherein: The transmitter and the receiver are connected by a line of sight (LoS), wherein: the length of the line of sight (LoS) is L, and 6 m ≥ L ≥ 2 m; and the extension direction of the line connecting the starting point (S) to the ending point (E) forms an angle θ with the line of sight (LoS), and 90° ≥ θ ≥ 45°.

8. The method as described in claim 6, wherein: A line of sight (LoS) is formed between the transmitter and the receiver, and the starting point (S) and the ending point (E) are located on the perpendicular line of the line of sight (LoS).

9. The method as described in claim 6, wherein step (S4B) comprises: filtering high-frequency noise and low-frequency DC components in the Channel State Information (CSI) signal with a filter; and performing principal component analysis on the Channel State Information (CSI) signal to extract signals associated with human gait.

10. The method as described in claim 6, wherein step (S5B) comprises: performing a short-time Fourier transform (STFT) on the Channel State Information (CSI) denoised signal to obtain a time-frequency map; tracking high-energy regions in the time-frequency map to form a trunk velocity profile on the time-frequency map; and converting the time-frequency map into a velocity-time graph to convert the trunk velocity profile into the original gait velocity curve and obtain the human channel state information (CSI) gait parameters.

Images