Survey On Content Base Image Retrieval System
Swarnalata V. Wankhede, Rashmi V. Sawalakhe · 2013
this paper shows a novel framework for combining all the three features color, texture and shape information. Also it achieves higher retrieval efficiency using dominant color feature. The image and its complement are divided into non- overlapping tiles of same size. The features drawn from co-occurrence conditional histograms between the image tiles and corresponding complement tiles. In RGB color space, act as local descriptors of color, shape and texture. In this paper apply the integration of the above combination, then cluster based on like properties. Based on dominant colors and retrieve the similar images. The information of Image is captured in terms of edge images this image computed using Gradient Vector Flow fields. Constant moments are then used to store the shape features.