Abstract—The main research challenge for a security and surveillance system is to create a real-time fully autonomous system that is also robust. In this research, a robust approach for real-time object detection and tracking in a dynamic scene using a moving camera is presented. The detection of the moving object and the tracking of the detected object are accomplished using a modified version of the enhanced SURF algorithm. This includes a color feature also to achieve a more accurate and robust results. This approach is able to track the detected object while reentering the scene after being absent for a short period of 4 or 5 frames. The regular SURF, enhanced SURF, and the current approach are implemented and the results are compared for speed and accuracy.

Index Terms—object detection, speeded-up robust features (SURF), scale- and rotation- invariant, object tracking

Cite: Saad A. Yaseen and Sreela Sasi, "Robust Algorithm for Object Detection and Tracking in a Dynamic Scene," Journal of Image and Graphics, Vol. 2, No. 1, pp. 41-45, June 2014. doi: 10.12720/joig.2.1.41-45