Multilayer Architecture for Content-Based Image Retrieval Systems

In this paper, we present a novel architecture for content-based image retrieval systems. Effective storing, browsing and searching collections of images is one of the most critical challenges of computer science. The design of architecture for storing such data requires a set of tools and frameworks such as database and service-oriented frameworks. With the vast growth of the Internet multimedia content, web services for CBIR systems are highly desirable. The proposed solution is based on a multi-layer architecture, which allows replacing any component without recompilation of other components. The approach is elastic and highly scalable. For experimental purposes we implemented the SURF local interest point detector as a feature extractor and k-means clustering as the image indexer. We also adopted an external solution, i.e. the CEDD descriptor into our system. The presented architecture is intended for content-based retrieval experiments as well as for real-world CBIR desktop and web applications.