The Research On Content-Based Image Browse And Retrieval System

With the development of the technology of multimedia and Internet, accessing mass amounts of multimedia data is becoming a reality. However, the traditional method of the manual depict on multimedia data exists prodigious abuse. Consequently, the needs for computerized content characterization of visual information have been strongly argued in recent years because of several limitations of the manual annotations and the ambiguous nature of visual information. The Content-Based Image Retrieval(CBIR) technology can exactly meet the needs mentioned above, also has been one of most active researches in computer vision, image database, knowledge discovery and so on.In this thesis, several key technologies of CBIR have been investigated. First of all, the present and the future of CBIR are discussed. Secondly, the primary technical methods of the image retrieval are investigated, the extraction methods and comparability matching of the color, texture, and shape characteristic are analyzed separately; this portion is a key part of the full text.Because of the weak correlation of low-level features and high-level conception of images and different subjective perception of the system users, relevance feedback has been introduced to CBIR, and the system learns the requirements of different users from the feedback procedures. Based on the requires mentioned upwards, the basic idea and framework of relevance feedback are analyzed and researched in this paper, to make it more effective, rapid access to the required images, and the application of CBIR technology in the network environment are introduced briefly.Finally, a content-based image browse and retrieval system are designed to certify its validity and correctness, and the demonstrations of systematic operation result are provided. System using Delphi 6.0 for the development of tools to achieve images stored as well as...