A New Relevance Feedback based Approach for Efficient Image Retrieval

Abstract

The rapid growth of digital image data summons the need for an effective and efficient content-based image searching system. Such systems should address the needs of the end user and should deliver the relevant images based on the search criteria. In order to meet this requirement, the content-based image search technique should capture the color and texture information. The performance of the algorithm can be enhanced using relevance feedback method. In this paper, a content-based image retrieval method based on image and its complement is presented. The similarity between the images is identified using an approach based on most significant highest priority (MSHP) principle or using a new distance measure which belongs to minkowski family. The retrieval rate is enhanced by relevance feedback technique based on kmeans algorithm. The approach is tested on Simplicity test dataset and a comparable performance was achieved