Content Based Image Retrieval in Similarity Integrated Network Using Combined Ranking Algorithm

Abstract

In image rich websites like flickr, Photobucket retrieving images in such a large network is very useful but also very tedious process. It also consumes more time and the retrieved contents are not exactly relevant always. In this paper, three algorithms have been proposed to improve the performance of such sites. To compare the similarity of the images efficiently, HMok-SimRank algorithm is used. It’s derived from similarity algorithm. Integrated Weighted Similarity Learning(IWSL) is used to integrate meta information descriptions with image content. Finally, Ranking algorithm is used to rank the images for the order of retrieval. Benefits of our proposed system applied in flickr are experimentally shown in terms of both relevance and speed.

5 Figures and Tables

Cite this paper

@inproceedings{Rajendran2015ContentBI, title={Content Based Image Retrieval in Similarity Integrated Network Using Combined Ranking Algorithm}, author={Pradeep S Rajendran}, year={2015} }