Shape Based Descriptors for Image Retrieval

-Image retrieval based on its content is the application of computer vision techniques to the problem of searching digital images in large databases. Images need to be described by certain features to search an image from database. During the last several years; content based image retrieval appeared as influential tool to proficiently retrieve digital images visually related to query image. The basic idea behind this approach is representation of image as feature vector and to measure the similarities between the images with distance between their corresponding feature vectors according to some metrics. The finding of correct features to represent images with, as well as the similarity metric that groups visually similar image together, are important landmark in construction of any Content Based Image Retrieval (CBIR) system. Shape is one of the most important features in CBIR. Many shape representations and a retrieval method exists. In this paper, various shape based descriptors are discussed for retrieving images.