Treffer: Content-Initiated Organization of Mobile Image Repositories

Considerable research has been done on the contentbased image delivery and access in distributed repositories. As noted in the literature, there is always a tradeoff between the image quality and the access speed. In addition, the overall performance is greatly determined by the distribution of the image data, specially, in a heterogeneous environment. In this paper, a semantic-based image access scheme for a distributed, mobile, heterogeneous database infrastructure, the Ubiquitous Content Summary Model, is presented that addresses both the data quality and performance issues. With the ability of summarizing the content information and guiding the data distribution, the proposed solution is distinguished by its mathematical representation and concise abstraction of the semantic contents of image data, which are further integrated to form a general overview of a image data source and its application of word relationships to construct a hierarchical meta-data based on the summary schemas allowing imprecise queries. Furthermore, it achieves the optimal performance in terms of searching cost. The fundamental structure of the proposed model is presented.