Effective Location-Based Geo-tagged Image Retrieval for Mobile Culture and Tourism Education

Abstract

With the development of Web 2.0 and location-based technologies, location- based image retrieval and indexing is now possible. This can provide high-quality support for field visits and teaching, which is an integral part for culture and tourism education. In the state-of-the-art retrieval methods, geo-tag and visual feature-based image retrieval has not been considered together so far. In this paper, we present an efficient location-based image retrieval method by conducting the search over combined geo-tag and visual-feature spaces. In this retrieval method, a cost-based query optimization scheme is proposed to optimize the query processing. Different from conventional image retrieval methods, our proposed retrieval algorithm combines the above two functions as uniform measure. Comprehensive experiments are conducted to testify the effectiveness and efficiency of our proposed retrieval and indexing methods respectively.