An Efficient Image Clustering and Content Based Image Retrieval Using Fuzzy K Means Clustering Algorithm

Abstract

The construction of large database with thousands of data has been facilitated by the developments in data storage and image acquisition technologies. Suitable information system requires proper handling of these datasets in efficient manner. Content-Based Image Retrieval (CBIR) is commonly used system to handle these datasets. Basis on the image substance CBIR extracts the images that are relevant to the user given query image from large image databases. Many of the CBIR systems retrieval of the result are corresponding to feature similarities for user given query, ignoring the similarities among images in database. These existing CBIR system measures the feature similarities by using k means algorithm, but the traditional k-means algorithm mostly depends on the selection of initial centers values, the algorithm normally uses random procedures to get them and it degrades the performance of the CBIR retrieval results.  To overcome the problem of initial centroid random  selection process  in K means clustering algorithm use the fuzzy logic based feature similarities information with K means clustering algorithm to image retrieval system. Combining both low-level and high-level visual features, the fuzzy k means algorithm entirely measures the features similarities information between the images in larger dataset. Fuzzy k means clustering algorithm optimizes the relevance results from conventional image retrieval system by firstly clustering the related images in the images database to improve the effectiveness of images retrieval system.
Copyright © 2014 Praise Worthy Prize - All rights reserved.