Scene Image Analysis using GLCM & Gabor Filter

In this paper, images of realworld natural scenes and manmade structures of different depth are taken. With increase in image depth , roughness increases in case of man-made structures whereas natural scene images become smooth, thus reducing the local roughness of the picture. Such kind of specific arrangement produces a particular spatial pattern of dominant orientations and scales that can be described using Gabor filter as it gives the local estimate of frequency content in an image. Here various techniques are used i.e. grey level co-occurrence matrices (GLCM), Gabor filters, combined GLCM and Gabor filters. Here the real scene images are classified in four classes such as near natural, near manmade, far natural and far manmade .Gabor filter only classify into low energy and high energy scenes. So the combination of Gabor filter and GLCM are used for classification in to four classes. In the proposed method i.e. the combination of Gabor and GLCM, first Gabor and PNN is used for classification between two groups high energy (such as near natural & far manmade) and low energy( such as near manmade & far natural) and then the GLCM and PNN is used for classification of subgroups. Keywords— Ga bor f i l t er s , GLCM, scene classification,