Some Approaches for Land Use Land Cover Classification

Abstract

newline Land Use Land Cover LULC classification is used in planning and utilization of natural resources and their management Recent advances in satellite and airborne sensors provide high resolution images with detailed texture edges and contours Synthetic Aperture Radar SAR is a powerful imaging method due to its active sensor technology SAR images provide fine texture of the surface under all weather conditions throughout the day and night SAR image plays a vital role in LULC classification than the other space borne and air borne images But SAR image contains a lot of speckle noise Speckle noise is randomly spread along with the surface texture and affects the identification of features Classification accuracy depends on the extracted features Despeckling filters reduce the noise considerably with loss of information So SAR images are enhanced by fusing with other sensor images Designing a despeckling filter and fusion of images is still a challenging task in the context of preserving desired features for classification The aim of this work is to despeckle SAR to fuse SAR and PAN images and to classify land covers maximum A Posterior MAP estimation is used in reducing noise without loss of information Fusion is achieved using feature based rule to obtain enhanced features without artifacts Based on the enhanced features land cover classification is attempted The SAR and Panchromatic PAN images from the space borne and airborne sensors are used in capturing the land surface.