Detection and Tracking of Multiple Moving Objects for Surveillance and Security System

In this paper, detection and tracking of moving object with shadow removal, automated counting and alarm generation was proposed. Improved Gaussian mixture model was used for background subtraction and modeling as it robustly deals with lighting changes, repetitive motions, clutter and slowly moving objects. Shadow detection and removal was performed using improved HSV color space method. Morphological operations were used for noise removal in post processing step. Kalman filter was used for tracking of moving objects with occlusion handling. In the last step, automated counting of moving objects and generation of alarm when motion was detected were performed. Experimental results show that proposed approach was able to detect and track moving objects with effective shadow removal and counting.