A method for detecting the number of consecutive passers-by based on wifi Doppler frequency shift

Abstract

The invention proposes a method for detecting the number of people passing through continuously based on WiFi Doppler frequency shift, and extracts Doppler spectrum from CSI information according to the principle that a human body will produce a Doppler effect on WiFi wireless signals when passing through a WiFi detection area. According to the maximum energy curve in the Doppler spectrum, combined with the threshold value, it is proposed to judge whether there are people passing through. Then according to the zero frequency shift energy curve in the Doppler spectrum, combined with the threshold value to judge the number of consecutive passes. A Doppler spectrum segmentation method based on optimal ergodic rotation mapping is proposed, and the continuous passing Doppler spectrum is divided according to the number of passes. It is proposed to extract features from the segmented Doppler spectrum and get the specific number of people passing each time through a classifier. The number of people passing each time is added up to obtain the number of consecutive people passing. In the experimental environment, the present invention can achieve 100% detection accuracy of the number of passing times and higher than 90% detection accuracy of the number of passing people, which is higher than the existing WiFi continuous passing number detection scheme.

Description

technical field [0001] The present invention relates to continuous passage detection and continuous passage number detection technology, in particular to a continuous passage detection and continuous passage number detection technology based on WiFi. Background technique [0002] Passage detection refers to counting the number of people passing through passages or entrances and exits within a period of time. The automatic detection and accurate counting of the number of people is very important for many places, such as: retail stores, shopping malls, visitor centers, libraries, museums, and public transportation, etc., these places need to detect and monitor the flow of people. For retail stores, accurate detection of the number of people passing at the entrance and exit can help retailers understand the number of customers in the store, thereby assisting store management. For indoor public places, the status of people entering and leaving can be determined through the dete...

AI Technical Summary

Problems solved by technology

This solution uses WiFi CSI as the original data, extracts the Doppler spectrum from it, and then performs traffic and number detection based on the WiFi Doppler spectrum. When it is used for discontinuous traffic, the detection accuracy of the number of people is high, but when it is used for continuous traffic, That is to say, when the front and rear distances between the passing people are relatively short, the Doppler effect caused by the passing of the front and rear people will interfere with each other, so that the Doppler spectrum of the front and rear people passing by will overlap significantly, resulting in the inability to carry out effective traffic detection. and people counting

Images