Pedestrian volume statistical approach based on WiFi channel state information and dynamic time wrapping

Abstract

The invention provides a pedestrian volume statistical approach based on WiFi channel state information and dynamic time wrapping for passageways and entrances and exits. Special hardware facilities are not constructed, and an existing WiFi wireless network is fully used. The method can carry out precious statistics on the pedestrian volume through the passageways or entrances and exits, is low incost and high in universality, and can work effectively under the conditions of light and no light. After a CSI data sequence is acquired, noise in the CSI data is removed first by means of a movingaverage algorithm and the CSI data after smooth denoising is normalized to -1 to 1. As it takes the crowds a certain time to pass through the passageways or entrances and exits and the time and rate passing through the passageways or entrances and exits of crowds at the same volume each time may be different, CSI data sequence templates are trained and matched by adopting a DTW algorithm to determine the accurate correspondence relationship among different CSI data sequences, so that template matching and similarity calculation are more accurate, and statistics of the pedestrian volume is moreaccurate.

Description

technical field [0001] The present invention relates to the field of people flow statistics, in particular to a method for performing people flow statistics at passages or entrances and exits based on WiFi channel state information and using a dynamic time warping algorithm. Background technique [0002] Counting the flow of people passing through passages or entrances and exits is a very important function. For example: in public places, according to the flow of people at each entrance and exit, you can understand the reasonableness of the entrance and exit settings and reasonably allocate employees; accurately count the flow of people at the passage and take measures according to the flow density Corresponding measures can effectively prevent disasters such as crowding and stampede, and improve public safety. The traditional method of counting people flow mainly adopts video surveillance method or wireless radio frequency method. Video monitoring methods have been widely ...

AI Technical Summary

Problems solved by technology

Currently the most widely used energy characteristic for measuring changes in WiFi signals is Received Signal Strength Indicator (RSSI), but due to the complexity of the indoor environment, WiFi signals have multipath effects, and each path has different delays , attenuation and phase shift, causing RSSI to be the superposition of multiple path signals, resulting in unstable RSSI and large errors when used for people flow statistics

Images