Detecting and Tracking of Multiple People in Video based on Hybrid Detection and Human Anatomy Body Proportion

Abstract

This paper addresses problems of detection and tracking of moving multiple people in a video stream. Detecting and tracking are fundamental tasks for future research into Human Computer Interaction (HCI). Detecting and Tracking multiple people in video are considered time consuming processes due to the amount of data a video contains, illumination changes, complex backgrounds and occlusions that occur as soon as people change orientations over time. This study focus on developing a fully automated system aims to Detecting and tracking multiple people in video, by analyzes sequential video frames based on hybrid detection algorithm, and tracking based on human body structure. The performance of the proposed system is tested through a series of experiments and human computer interaction application based human detection, tracking and identification. Identification is based on new clustering method mentioned in this paper. General Terms Video Processing, Computer vision systems, Human detection and tracking, Clustering.