Robust Real-Time Object Tracking Under Background Clutter

Abstract

In this paper we propose a novel method for detecting target object in presence of background clutter such as leaves of trees and changing illumination condition in real-time video and then tracking the object. A small movement in the background such as leaves of trees and changing illumination condition affects the performance of the automated tracking system. Here in this paper, we reduce the spatial and intensity resolution of an image to minimize the effects caused by the background clutter. A 3-frame-differencing method is employed to detect a moving object. Fuzzy c-means clustering is used to segment the moving object from the background. The proposed method yields superior performance as compared to the other existing methods.