Design And Implementation Of A Content Based Image Retrieval Language

With the rapid developmentof computer technology in recent years, multimediatechnology is widely used in various fields and the Internet technology has also been ahigh degree of popularity. Increasing in the application of digital image has made thedramatic growth trend in digital image data. How can we retrieve useful informationfrom a massive Internet database fast and accurately has become a challenge. To solvethis problem, the content-based image retrieval technology, which can express complexquery requests and is more accuratethan text-based image retrieval technology, hasreveived a lot of attention recently.This paper proposes a content-based retrieval language named IQL (short forImage Query Language). Firstly, it conducts a deep analysis on image retrieval language.On the basis of this, we decided to implement image retrieval by the query expression,which was generated from combining the operation of logical and spatial relationshipamong images. Then we designed the grammar of IQL. We implement the interpreter ofIQL and the image retrieval system based on IQL combined with Hadoop frameworkfinally.Some experimental results have shown the effectiveness, efficiency and flexibilityof the proposed IQL system.