Shape and Texture based Image Segmentation using Image Database

Abstract

This paper presents the overview of the methodology employed in this research work for segmenting objects in the image sequences. Image segmentation is the technique of allocating a label to each pixel in an image such that pixels with the same label share some image characteristics like foreground and background of the image pixel. The result of image segmentation is a set of segments or regions that together represent the entire image. For each image pixel, a feature vector is computed as a sequence of area moments concerning certain characteristics such as color, texture, intensity, etc. From the literature survey, it has been observed that the performance of image segmentation methods is influenced by many factors such as intensity, shape, texture and image content. Hence, a single segmentation method cannot be applied to all types of images. At the same time, not all the methods perform well for one particular image. Efficient implementation, computational times are the critical aspects of defining the performance of these methods.