Tracking Methods for Multipurpose Video Processing Applications

Abstract

Video target tracking is a complex and time consuming process because of huge data that is contained in video stream. In this paper, a real-time software is developed to track any motion in a video zone for surveillance/security applications. First, an interested object is detected, isolated (segmented), and localized among different objects of video frames. Secondly, three tracking procedures are applied to find the efficient one that satisfies the real-time requirements for tracking the target with minimum response time. The tracking model consists of two main parts, the web camera which represents the video acquisition device that monitors any motion in a dedicated zone. The second part is the software that controls and processes the video stream; it captures the frames and tracks the new target center coordinates. The center values that will be found by tracking algorithms are very essential for other video/image applications like radar system in military applications, road traffic control applications, or kidney stone ultrasound in the medical applications. The experimental results found that the response time and targeting accuracy of the three methods are very close, the maximum difference in a response time was 0.0421 sec for threshold ϴ = 0.039.