Malaysia travel destination images search engine using Geometric Moment Invariants

Abstract

Content-based image retrieval (CBIR) is an alternative way to achieve effective image retrieval results which fulfill the user’s requirement or query since images have become a demanding media. This paper proposes a Malaysia’s travel destination search engine (MTDSE) system for image retrieval. In the MTDSE, the effectiveness and the accuracy of the image retrieval process will be determined through features extraction using moment geometry. The keywords from user will be used to match with the image database to search for the related images. The results from the text-based retrieval will be shown to user. The user will choose one of the images that he or she wanted. The selected image will proceed to the edge detection process using Sobel operator. Features extracted from the edge detection process will be used for the matching process in order to retrieve relevant images. The precision and recall metrics have been used to evaluate the effectiveness and the accuracy of the image retrieval prototype. From the experiments, the prototype system has returned the related images that match with user’s interest.