The Research And Implementation Of Image Compression System Based On EBCOT And ROI Coding Technologies

Region-of-Interest (ROI) image coding technique means to compress the interesting regions in an image without loss or with little loss, and to compress the uninteresting (background) regions with much loss. Based on this idea a high compression ratio can be obtained and the important information can be preserved in the image. ROI coding technique provides a new way to solve the contradiction between the compression ratio and the image quality.The key techniques and difficulties in EBCOT realization are researched, including DWT, independent block coding, rate-distortion optimization and code organization, and a gray image compression system is realized. The system possesses several characteristics, such as good compression performance, SNR and resolution retractility, ROI coding supporting and so on. Some optimization measures are adopted in the system. Experiments show that the system increases the image coding efficiency.In this thesis, a novel ROI coding method is proposed, which could shift the rate-distortion slopes of the ROI code-blocks, and dynamically reduce the priority of background wavelet coefficients in the ROI code-blocks according to the target bit-rate. The proposed method could not only avoid the low efficiency caused by increased bit-plans, but also achieve fast ROI quality improvement and provide early background appearance for ROI assessment. Furthermore, the proposed method enables the flexible adjustment of image quality between ROI and background, and can reconstruct ROI with higher quality before receiving the entire image, and can support arbitrarily shaped ROI coding without coding the shape. Finally, experiments have done to validate the proposed method, and the results are satisfying.