An improved image retrieval system using optimized FCM & multiple shape, texture features

Abstract

Retrieval of user interested images based on pictorial queries is an interesting and challenging task. This paper proposes an Improved Region based image retrieval system using FCM & multiple shape, texture features. The Proposed system uses Fuzzy c-means clustering algorithm for image segmentation. Local Binary Pattern (LBP), Hu moments and Radial Chebyshev Moments are used in this work. For similarity comparison City block distance is used. The experimental results presents a comparative analysis of the proposed image retrieval system with existing system using multiple features. The Experimental results also prove that the proposed improved method provides better precision. The precision is increased from 85 to 88 percentage as per the recorded results. The precision is calculated by the ratio of number of similar images retrieved to the number of actual images retrieved from database. The Proposed method is tested by using COIL database which is freely available in web.