Performance Evaluation of Alpha-Beta and Kalman Filter for Object Tracking

Abstract

object tracking is an important field of research of image processing widely used in computer vision, video image processing, pattern recognition, and artificial intelligence. Video based object tracking is a challenging problem which involves lot of difficulties like object to scene and object to object occlusions, abrupt object motion, and camera motion. The tracking process involves in two phases. The first phase includes the background separation for detection of moving object. In the second phase tracking of the detected object is done using the filters like Alpha-Beta filter and Kalman filter. Alpha-Beta filter is one of the traditional techniques used for tracking which can be solved in iterative, decentralized manner. Another most popular technique used for tracking is the one that uses Kalman filter with measurements (often noisy) of position of object to be tracked as input to it. It is very much needed for the new filter designer to design object tracking algorithm and performance comparison of different object tracking algorithms should be known. A perfect comparison of object tracking algorithms under different noisy conditions and implementation issues are not discussed in the literature. Hence the author proposes performance comparison by taking parameters like RMSE, SNR, error estimations, and computational complexity under different noise conditions.