Abstract

Even though scene classification is a field that has been studied in depth, war scene classification has not been so far. Scene classification underlies many problems in visual perception such as object recognition and environment navigation. Scene classification, the classification of images into semantic categories (e.g. opencountry, mountains, highways and streets) is a challenging and important task nowadays. In this paper we are trying to classify a war scene from the natural scene. For this purpose two set of image categories are taken viz., opencountry & war tank. By using Haar and Daubechies(db4) wavelets, features are extracted from the images. The extracted features are trained and tested with (i) Artificial Neural Networks(ANN) using feed forward back propagation algorithm and (ii) Support Vector Machines(SVM) using radial basis kernel function with p=5. The comparative results are proving efficiency of Artificial Neural Networks towards war scene classification problems by using Wavelet feature extraction method. Although this study has been the first step of the research in the area of scene classification, the results have shown that the war scenes can be successfully classified from opencountry. It can be concluded that the proposed work significantly and directly contributes to scene classification and its new applications. The complete work is experimented in Matlab 7.6.0 using real world dataset.

Paper Details