A Computer Vision Approach to Object Tracking and Counting

Abstract

This paper, introduces a new method for counting people or more generally objects that enter or exit a certain area/building or perimeter. We propose an algorithm (method) that analyzes a video sequence, detects moving objects and their moving direction and filters them according to some criteria (ex only humans). As result one obtains in and out counters for objects passing the defined perimeter. Automatic object counting is a growing size application in many industry/commerce areas. Counting can be used in statistical analysis and optimal activity scheduling methods. One of the main applications is the approximation of the number of persons passing trough, or reaching a certain area: airports (customs), shopping centers and malls and sports or cultural activities with high attendance. The main purpose is to offer an accurate estimation while still keeping the anonymity of the visitors.