A source and channel coding scheme for image transmission over noisy channel is presented in this paper. There is crude block structure in the wavelet coefficient of images. Coding each image blocks independently and reassigning its bitstream can achieve the resynchronization of decoding image in receiving terminal to improve the image's ability to resist noisy. But different image block has different character, and has different energy distribution, so it's hard to implement consistency in rate-distortion performance. Paper proposes a classification method based on high frequency subband energy distribution of wavelet. Give different codeword length to different type image block to achieve better rate distortion performance. Experimental results show that the proposed scheme is able to provide better image quality than scheme without using the classification method in range of some BER.
展开▼
