The Research On Progressive Image Transmission Methods Based On Region Of Interest

The demand for image information has increased dominantly with the mature of Internetconstruction, and the image technology has become a hot spot in the image processingresearch. For the image with large amount of information, such as full visual image, it issignificant to research on improving the image transmission efficiency, and allowing users toget quick access of the image information, so as to obtain more image information withlimited bandwidth resources. In order to improve the access speed of the image information,the image compression ratio is increased. However, the increased image compression ratiowill lead to the degradation of image quality. Thus how to increase the image compressionratio and guarantee the image quality at the same time has become the difficulty of the imagetransmission technology.Progressive image transmission technology first performs compression to the integratedimage information, and then transmits this information with priority during the imagetransmission. The receive client can achieve the rough information of the image quickly withlimited encoding bits. Then, the client can determine whether perform compression to thedetail image information according to the requirements. The detail image information qualitycan be continuously improved with the achievement of following image encoding. It alsoprovides the termination of image transmission according to the requirement of users, whichcan keep the good balance between guarantee quality of image transmission and improve theimage compression efficiency.For the transmitted image, most of the people usually interest in some specific region ofthe image, which is defined as the region of interest (ROI). The ROI contains more valuableinformation to the users. On the basis of progressive image transmission, we perform priorityencoding to the ROI according to the information requirements of the users. Thus, the receiveclient can quickly obtain the required valuable information of ROI and determine whether tocontinue transmit the encoded image information, further improve the image transmissionefficiency.The wavelet transform is performed as the image pre-processing in this thesis, and thecompressed encoding of ROI is performed to the wavelet coefficient of image. Then the integrated information of the image and the ROI information is in the front of transmittedencoding. The progressive image information can be established between two independentcomputer by internet.The encoding and decoding methods of progressive image transmission are firstlyintroduced in this thesis. Secondly, the embedded compression encoding methods based onwavelet transform theory are summarized. The improved SPECK compression encodingmethod is proposed. It can reduce the encoding complexity and improve the process speed ofSPECK encoding and decoding. Thirdly, the ROI encoding methods are analyzed, and therespective advantages and disadvantages are analyzed by simulation. Finally, the progressiveimage transmission interface of the server-side and client is designed by hybrid programmingof Visual C++6.0software and MATLAB interface functions. It can realize the progressiveimage transmission between two independent computers.