Date of Publication :9th August 2017

Abstract: With the increasing demand for need of retrieval of images based on different aspects, properties, characteristics there has been tremendous research going on in the field of Content-based image retrieval (CBIR), also known as query by image content (QBIC). Content here is referred to visual features ie colour, texture, shape, spatial layout of an image which needs to be extracted. This paper proposes a novel approach wherein content based image retrieval is based on color and texture primarily implementing color histogram for color based and gabor filter for texture based retrieval. Colour histogram comprises of two colour models(RGB and HSV). For texture based, Frequency and orientation representations of Gabor filters are similar to those of the human visual system, and they have been found to be particularly appropriate for representing and discriminating texture well . Main aim of this paper is to implement a new technique that is quadtree segmentation on query and database images and then apply feature extraction ie color histogram on all images and match the query and database images using similarity criteria. Similarity metric used in this paper is Euclidean distance. Quad tree is a segmentation technique which divides image into homogeneous blocks. Hence the name Quadhistogram is a new revised technique to implement colour based retrieval under local colour histogram which divides the images into homogeneous blocks and then compute histogram of each block

Reference :