Gustafson-Kessel algorithm in content based image retrieval for Malaysia tourism website

Abstract

In large areas of information technology, huge collections tourism images are being created and publish anywhere especially in blog or in personal website. Many of these collections are the photo that being taken around the globe using digital camera. We know the one way of indexing and searching these digital images collections was using keyword metadata, or simply by browsing. Nowadays, Content Based Images Retrieval (CBIR) is the way to assist the system to retrieve the related images. When the users are not satisfied with their query results, the relevance feedback (RF) retrieval is one of the solutions for this problem. The user needs to use this system in multiple time in order to increase the retrieval performance. In this paper, we concentrate on relevant feedback approach based and Gustafson-Kessel (GK) clustering approach in order to evaluate the availability this prosedur to adapt into tourism website for the image retrieval results using users feedback. From the experiments, we have found that the RF method using Gustafson-Kessel (GK) clustering can improve the retrieval performance of the Tourism CBIR system even if we are using a large set of image datasets with a variety of images especially in tourism image database.