Block-Based Adaptive Threshold Wavelet Image Coding

With the wide application of Internet and the increase of wireless video transmission needs, processing capacity and transmission capacity of the image data expand fast, while the existing hardware facilities for the image processing and transmission are hard to keep pace with its development, which brings tremendous burden on communication bandwidth and storage medium. Therefore, it is an inevitable choice for the image data to adapt coding compression technology for processing to reduce transmission information. The embedded coding has unique advantages, such as the endorsement of progressive transmission, the halt at any point of coding; fill the goals and objectives of the distortion rate requirements. It has been applied to progressive image transmission, image browsing and image transmission on the Internet more and more.Among so many embedded image coding algorithms, the EZW, SPIHT and SPECK are the best representatives. The EZW algorithm is regarded as a great leap, it uses zerotree data structure to realize the effective organization of the wavelet coefficients. The SPIHT algorithm causes a better performance by improving the EZW's data structure. This paper focuses on the SPECK algorithm, because it fully takes the advantage of energy concentration of the wavelet coefficients and the characteristics of energy decrease with the increase in scale. It combined the quad-tree division method and the bit-plane coding method to obtain a better compression performance and received more attention. SPECK uses block independent coding with high coding speed, but there still exists some improvement to be done as for the achieving speed and complexity.An improved algorithm in the sequence of process and the importance test and etc is presented in this dissertation. Besides the high coding efficiency, the improved block-based adaptive threshold wavelet embedded coding algorithm preserves the properties of the SPECK and has high coding efficiency. Simulation results show the correctness of the improved algorithm, and testified the achieving speed and high coding efficiency.