Efficient technique for color image segmentation towards the enhancement of image retrieval

Abstract

Content Based Image Retrieval CBIR system performs search and retrieval operations in image database that has been developed in recent years CBIR has a very vital impact in web application The main objective of this thesis work is to increase the accuracy level of image retrieval Since the image database is very huge more attention has to be given during retrieval process in i the quality of the image ii features that are to be selected for newlineimage retrieval iii object detection methodology and iv enhancement in the retrieval process The above said processes are analyzed in four chapters and are as follows Since the retrieval process is based on web applications where the color images are available on the internet reduction of additive noise for digital color images is important and is performed in chapter 3 From the literature till recent publications it is found that wavelet based fuzzy filter suppresses the noise very well by keeping its features intact eg edge color components distances and so on The visual results illustrate that the wavelet newlinebased fuzzy filter method restores the original color components distances much better than other filters like median wiener linear smoothing filters and so on In this research work wavelet based fuzzy filter is used newline