Still Image Compression Method

With the fast development of information technology, the requirements on image quality, size, and transmission speed presently become more and more intense. As one of the kernel components in data compression field, image compression always attracts many researchers' attentions. Now, the new generation of image coding method, such as JPEG2000 standard, has put the efficiency of image compression on a very high level. At the same time, these methods also provide some useful functions: temporal scalability, spatial scalability and quality scalability. However, these splendent achievements still can not satisfy the people's requirement. There are a lot of works still should be done to improve the performance of image coding.Based on the existent research works, this paper respectively analyses the features of gray-scale image, multi-channel image and multi-view image. Based on the clear analysis, several novel and efficient image coding methods are proposed. The main contribution of this dissertation can be summarized as follows:1. We studied the optimal context quantization criterion for the high dimension context model from a view of conditional information entropy. Then, a novel context quantization criterion with low computational complexity is proposed.2. The paper describes an image coding algorithm based on the discrete wavelet transform and context-based adaptive arithmetic coding. A novel coding model, the combination of the spatial and frequency prediction in the wavelet domain, is proposed in the paper. At the same time, context quantization as the key part of arithmetic coding is carefully analyzed in order to obtain in suitable contexts for coding and decrease the model cost.3. Analyzes the inter-color correlation and answer two questions related to color image coding: (1) what kind of inter-color correlation exists in color images after the discrete wavelet transform? (2) How strong is it? This analysis helps us to find a most suitable inter-color context and eventually leads to a new embedded color image codec. By using the discovered inter-color context, significant performance improvement can be achieved when encoding chrominance components.4. Proposed a novel hyper-spectral image coding method based on three-dimension wavelet transform (3DWT) and context quantization. In this method, 3DWT is firstly adopted to remove the intra- and inter-spectral redundancy of hyper-spectral images. Then the correlation of wavelet coefficients is analyzed bya high dimension context predicting model. 5. In the booming field of multi-view coding, proposes a geometric predictionmethodology for accurate disparity vector (DV) predicting. Based on the new DVpredictor, this paper designs a basic framework that can be implemented in mostexisting multi-view image/video coding schemes.In conclusion, we studied the static image coding techniques and achieved some valuable results. However, static image coding is still a potential research field which is worthy of further study.