A video is a collection of sequential images with a constant time interval. Videos can provide more information about real world objects where scenario changes with respect to time. Because of complicated occlusions, and disordered background, Identifying moving objects and their tracking is a challenging problem for many computer vision applications, especially in complex real world scenes that commonly involves multiple objects. Therefore, we need some automated devices to handle the video for monitoring the moving objects. Many algorithms have already been developed to automate the monitoring of the moving objects. In this paper, we propose a novel approach for tracking multiple objects via using background subtraction based on Gaussian Mixture Model (GMM). The proposed model in this paper is effectively working for dynamic background, fast light change environment, repeated motion. For tracking the objects we used Kalman filter. This filter successfully tracks moving objects even in full occlusion cases. Hence lower processing time while maintaining competitive performance in terms of recall and precession is the main feature of the proposed approach in this paper.
Object detection, Background subtraction, Gaussian mixture Model, Tracking, Kalman filter, Shadow Detection.
Nisha Pal; Suneeta Agarwal, Detection and Tracking of Moving Objects in Surveillance System, HCTL Open International Journal of Technology Innovations and Research (IJTIR), Volume 14, April 2015, eISSN: 2321-1814, ISBN (Print): 978-1-62951-946-3.