A Novel Way of Tracking People in an Indoor Area

摘要

Tracking people in an indoor area is a technically challenging problem, and has many interesting applications. One of the scenarios being, tracking customers in a big shopping mall. This real time location information can be used for a variety of needs. In this paper we present a novel way of achieving this by matching people based on the image of the lower part of their body (pant/leg and shoes). Our approach is novel in 2 ways: there are no per-customer costs i.e. nothing vieeds to be changed at the customer side. Also, as we use the image of lower part of the body, there should be potentially no privacy related issues, unlike face recognition.