CRCSI Project 5.1 Support Tools for Spatial Data Mining and Agent-Based Modelling Project Leader – Dr David Tien, Charles Sturt University
Funding received $470,000
Participants Charles Sturt University
There are 3 main approaches to the question of spatial search and image retrieval: text-based search on annotated images (TBSAN), content-based image retrieval and semantic based image retrieval. Currently, there are a number of web search engines, such as Google Image Search, Yahoo Image, AltaVista Image, Ditto, and PicSearch to name just a few. However, these engines are TBSAN. Most of these are based on the technique described in P. Enser’s 1995 article, “Pictorial Information Retrieval” in the Journal of documentation. This, in fact, is not spatial image retrieval, but a text search on image descriptions. Textual representation of an image is often ambiguous and non-informative of the actual image content. Filenames may be misleading, adjacent text is difficult to define, and a word may contain multiple senses.
The aim of the Project 5.1 is go beyond the current scope used by Google and Yahoo etc and integrate Content-Based Image Retrieval (CBIR) into the search ability. Although CBIR is not a new idea, the research interests and publications, fuelled by the internet search needs, only took off after 1997.