Research On Image Content Retrieval Based On The Complex Wavelet Transform

A large number of images have been produced with the development and application of multimedia technology. How to analyze, store and retrieve these images become an imperative problem. The content-based image retrieval technology can find the image that contain specific content directly from the image library, according to image content features and feature combinations, and it has become a hotspot in study to effectively resolve above problem.In this paper, according to the hot point of the current research of CBIR, based on the analyzing and discussing for the key techniques of CBIR, we mainly research the texture image retrieval based on dual-tree complex wavelet. The complex wavelet transform principles and characteristics are introduced, and through DT-CWT implement the decomposition and reconstruction of the two-dimensional image. The rotation-invariant texture characteristics are studied deeply, a new rotation invariant texture extraction algorithm using DT-CWT and moment invariant is proposed. The feasibility of the algorithm has been demonstrated through two experiments. A rotation-invariant texture images retrieval sysem based on DT-CWT is develped. This system implements the texture image retrieval after eliminating the interference caused by the texture image rotation. Finally, performance analysis have been done for the retrieval system.