| Recent years have witnessed the rapid growth of digital images due to theincreasing power of computing and the fast development of Internet. How to retrievethe images in which people are interested from those is becoming an issue which needssolving urgently. Content-based Image Retrieval (CBIR) is the key technology toaddress this urgent issue. CBIR is a technology for retrieving images on the basis ofimage features such as color, texture and shape, etc. However, the content features ofimages are complicated and different at all, which make the retrieval efficiency ofCBIR systems lower. According to this, MPEG (Moving Picture Experts Group) ofISO/IEC developed the ISO/IEC standard MPEG-7. Based on MPEG-7, the dissertation gives two image retrieval methods based oncolor histogram and color-space histogram, and implemented the Image RetrievedModel System. This dissertation includes four sections as follows: Section â… researches CBIR. By introducing the background of CBIR, thisdissertation gives several image retrieval methods based on color, texture and shapefeatures. Section â…¡ introduces the MPEG-7 standard with the main elements of theMPEG-7's standard, MPEG-7 parts and the scope of the standard, etc. This sectionemphasizes on the basic structures and descriptors that cover following basic visualfeatures: color, texture, shape, motion, localization, and face recognition. Then thissection analyses the relation of CBIR and the MPEG-7's standard. Section â…¢ gives an image retrieval method based on color histogram. Accordingto this, the dissertation gives an improved image retrieval method based on color-spacehistogram, which overlapped sub-regions to get color histogram in each region. Section IV implements the Image Retrieved Model System using two imageretrieval methods based on color histogram and color-space histogram. This sectionalso gives the system design, system implement and the results of the experiment. Inthe end, this section analyses the results.
|
PDF Full Text Request