Comparative Analysis in Content Based Image Retrieval System Using Color and Texture

Searching digital images from a large database can be solved by the Content Based Image Retrieval System based on the extraction of features such as color, texture and shape of the image. CBIR is the most efficient image retrieval method. Most of techniques for CBIR use numerical representations called feature vectors to allow d searching in large image collections. Analyzing images based on the colors is one of the most widely used techniques since the color does not depend on image size or orientation. Texture analysis is mainly used to identify the image given by the shape, size, brightness etc. The proposed method uses image features such as color and texture to retrieve the images from database. In addition, it uses auto color correlogram for identifying images based on the feature and Spatial texture for identifying images based on the texture. Moreover, this system has operated on a Corel database containing 1000 general-use color images. The study shows that the performance and accuracy of retrieving images in the large database.