PEDESTRIAN SENSING WITH FEATURE EXTRACTION
Timothy Chamberlain, Alistair Armitage, Malcolm J. Rutter, T. David Binnie · Research Output (Edinburgh Napier University) · 2006
Automated pedestrian sensing and tracking is receiving increasing attention from researchers in a number of areas. Applications are common in security, in urban planning, and in the retail industry. A common approach is to use CCTV and visible image processing, but this involves complicated scene interpretation. This paper describes our research on monitoring the movement of pedestrians using relatively low-cost, low resolution infrared sensors. We describe our data collection system and processing techniques and introduce new data analysis software. Preliminary results show that, as well as the simple tracking of pedestrians, biometric data, such as gait measurement, becomes possible. Our work aims to provide data to improve security and to enable behavioural monitoring of pedestrians in public. This data is important to those working in pedestrian monitoring, market research and in public safety and security.