Color Still Image Combined Coding Based On Wavelet Transform

With the development of multimedia applications, digital video techniques and transmission images in heterogeneous networks (such as smart cell phone, long-distance browsing and searching for images, global multi-user virtual environment sharing, and so on), image processing and coding techniques become increasing more important. Especially for image acquisition system, image storage and image transmission process, image processing and image coding technology has been widely employed.Wavelet analysis is becoming one of the important techniques used in image coding due to its good performance in time domain. Vector quantization relates to different research areas, such as information theory, coding theory, communication theory, security technology, signal processing , optimization theory, fuzzy theory, matrix analysis, neural network, wavelet transform, vision model, topology, stochastic process, prediction technique, pattern recognition, and so on. Arithmetic coding is adopted by many coding standards. One of its main advantages is that the model of the input symbols is adaptive. The arithmetic coding is an efficient lossless coding.This thesis titled"Color Still Image Combined Coding based on Wavelet Transform"first illustrates some related image coding theory, and then focuses on the discussion about three useful coding techniques; they are wavelet analysis and transform, arithmetic coding, and vector quantization respectively. Based on the discussion, we jointly used wavelet transform, arithmetic coding, vector quantization and the new coding scheme proposed in the color image processing, in this scheme low-frequency coefficients are encoded by adaptive arithmetic coding and high-frequency coefficients are encoded by vector quantization. During the vector quantization process, we improve the codebook design algorithm for hierachical vector quantization codebook and finally optimize the codebook. In this way, the computational complexity is greatly reduced with a better reconstructed image quality. The proposed system has been implemented by C language. The experiment results show that our proposed algorithm has a much better performance than JPEG coding standard, with the same compression rate, our proposed algorithm can obtain a better PSNR value than JPEG.