Moving Objects Detection Using Fuzzy Color Difference Histogram Mechanism

In the presence of dynamic background (such as camouflage, ripples in water) detecting of moving objects became the very toughest job. In order to subtract the background we are using three techniques in this paper i.e. CDH, FCDH, FCH. Initially the CDH (color difference histogram) technique used in order to remove the dynamic background so that it will help us to detect the moving objects by minimizing the false error occurred due to the non-stationary background. This paper proposes an algorithm FCDH (fuzzy color difference histogram) which uses the concept of fuzzy C-means clustering (FCM) and updates CDH for reducing the effects of intensity variations because of some fake motions. This algorithm tested with number of videos available in public.