Research of people counting based on face detection and area-tracking

This paper proposes a novel system for people counting based on face detection and tracking. Through the real time processing of video image gotten from monitor devices, the system detect the faces and combine with set people counting area to realize automatic people counting. In this system, firstly we set face tracking area and people counting area, then use AdaBoost and skin color character algorithm to detect faces and use camshift algorithm to track them, at last count people when the center point of tracking section passes the people counting area. Experiment result demonstrates this system has relatively high efficiency and accuracy.