Image retrieval using color and texture binary patterns

Abstract

Image retrieval plays a major role in security systems to extract the images with similar features or patterns, to retrieve the relevant images in web search engines, in industries to detect crack in the manufactured parts, in architectural designs to find same texture patterns and so on. To accomplish efficiency in all the fields of image processing, the effective image retrieval mechanism is imminent. In this paper, we proposed a method based on the combination of binary texture patterns and color features. A Local Binary Pattern (LBP) plays an important role in extracting the binary texture features and color histogram. This feature is implemented to identify and extract features of prominent object present in an image. Using different statistical measures, similarity measures are calculated and evaluated. Image retrieval based on color or texture is a trivial task. Identifying objects of prominence in an image and retrieving image with similar features is a complex task. Finding prominent object in an image is difficult in a background image and is the challenging task in retrieving images. The Implementation results proved that proposed method is effective in recalling the images of same pattern or texture.