Estimation of Pedestrian Flow Based on Wi-Fi Data and Video Cameras

Abstract

This paper is devoted an analysis using WIFI MAC addresses and video data. A framework is developed for processing and analysis of the data from WIFI scanners, particularly electronic devices with unique MAC addresses, for estimation of pedestrian flow and walking time. Experiments are conducted in the campus of The Hong Kong Polytechnic University (PolyU) to collect relevant data and to investigate the detection performance of the WIFI scanners together with the signal strength of electronic devices. Pedestrian flow is estimated based on video data by moving observer method and WiFi data by the developed framework, respectively. Comparison is also made between the estimated number of pedestrians using WIFI data with the actual number of pedestrians extracted from video records.