Object Classification Under Partially Cluttered Background Using Statistical Based Features

Object classification under partially cluttered background is a difficult task in still images. The challenging task in this problem is the classification of objects invariant to size and pose with partially cluttered environment containing natural scenes. This paper addresses the issues to classify sample objects from caltech-101 standard database containing airplanes and motorbikes. The threshold technique with background subtraction is used to segment the background region to extract the object of interest. The background segmented image with region of interest is divided into equal sized blocks of sub-images. The statistical features are extracted from each sub-block. The features of the objects are fed to the back-propagation neural classifier. Thus the performance of the neural classifier is compared with various categories of block size. Quantitative evaluation out perform the previous work in the literature with an improved results of 92.4% due to absence of occlusions. A critical evaluation of our approach under the proposed is presented.