Abstract

In the field of remote sensing, image segmentation methods becomes more and more important to analyze the images. From the images, Image Segmentation extracts the similar features based on their color, intensity. Then using different segmentation techniques groups them together. The consideration of the segmentation quality is of great significance. This paper presents a technique to segment the satellite images into different regions based on the properties of color and texture present in the image. In this technique, the implicit integration of color attributes and texture attributes are performed separately and combined them during the segmentation process. The Texture superpixels method improves the superpixels decomposition approach and locally set the spatial regularity of superpixels, to automatically adapt the image contents, thus improving the accuracy of segmentation.