Competent Tracking of Moving Object Using Affine & Illumination Insensitive Template Matching

摘要

Moving object detection & tracking in real world scene is becoming significant problem in today's era. The extensive study in this area is motivated by potential number of applications of object tracking. In this paper, we analyze a method for motion segmentation & tracking of nonstationary objects that uses the complete image information using the affine flow equations. These affine flow parameters are then combined with illumination insensitive template matching tracker to efficiently segment & track the moving objects. Experimental results performed on various database videos shows the method effectively tracks the object without miss detection of the target compared to other methods.