Cluttered Background Removal in Static Images with Mild Occlusions

Cluttered background removal in static images with mild occlusion remains still a challenging task for object identification or classification problems. In most of the real-world images, vehicle objects with cluttered background containing trees, road views, buildings, people, etc tent to be a noisy data or leads to the problem of clarity. The background feature covers the major portion of the image. Classification of objects fails due to cluttered background features and occluded features. This paper presents a novel approach for background removal in static images containing car object with cluttered background and mild occlusion. The morphological operations like region filling technique, background subtraction along with mapping function are used to extract the region of interest being the vehicle object. A critical evaluation of the proposed approach with the University of Illinois, UrbanaChampaign (UIUC) standard database is presented.