Volume : III, Issue : IV, April - 2014

An Enhanced Content Based Image Retrieval Using Three Region Colour Histogram of The Image

L. Jayanthi, Dr. K. Lakshmi

Abstract :

Large number of images are created everyday due to the advances in data storage and image acquisition technologies. In order to deal with these data, it is necessary to develop suitable information systems to efficiently manage these collections. Image searching and retrieval of desired images is one of the most important services that support the system with image collections like educational image resources, medical images, finger prints, satellite images, photo collections, museum pictures etc. Content Based Image Retrieval system searches the most similar images of a query image that involves in compå the feature vectors of all the images in the data base with that of the query image The color histogram is one of the most important techniques used in Content Based Image Retrieval(CBIR). It’s efficient to compute and effective in searching results. Most commercial CBIR systems use color histograms as one of the features. The weakness of the colour histogram based method for measuring image content is that the space information will be missing in a typical colour histogram. It means, the colour histogram will not say where that color was originally present in that image. If the colour feature is blue, we may not be able to identify whether it is sky or sea. In this work, we address CBIR using three region (top, middle and bottom) colour histogram of the image. The combination of these three histograms will indirectly include the space information within it. We have used the Semantic Sensitive Integrated Matching for Picture Liaries database to evaluate the performance of the system. The average precision is used as a metric to measure the performance and compared the performance of our method with four other CBIR methods using Integrated Region Matching(IRM), Fuzzy Integrated Region Matching (F-IRM), Colour Salient Points (CSP) and Ripplet Transform and Fuzzy Relevance Feedback Mechanism (F-RFM) which used the same data set and evaluation method. According to the arrived results, the performance of the proposed three region colour histogram based CBIR system was significantly better than other compared methods.

Keywords :

Article: Download PDF   DOI : 10.36106/ijsr  

Cite This Article:

L.Jayanthi, Dr. K.Lakshmi An Enhanced Content Based Image Retrieval Using Three Region Colour Histogram of The Image International Journal of Scientific Research, Vol.III, Issue. IV Apr 2014

Number of Downloads : 1106

References :