A Study On Image Compression Coding Technology

In this article, we studied on universal method of lossless and lossy image coding, and presented solutions for such problems respectively: criterion for performance evaluation of integer wavelet transforms, coding structure with multi-scalability, and low complexity entropy coding algorithm.Wavelet transforms that map integer to integer (Integer Wavelet Transform) are completely reversible even under finite precision condition, therefore, they become the common-core of lossless and lossy image coding. Wavelet transforms possess the ability of multi-resolution analyzing, and the transform coefficients preserve the information of position in space field. Coding structure making use of these properties can provide image coding and decoding with scalabilities of resolution, fidelity and random access.Multi-level coefficients entropy was presented instead of single level coefficients weighted entropy, and as a new criterion for performance evaluation of integer wavelet transforms, it obtained better result which is nearer to bit rate than old one. Without attached to an entropy coder, the new method can provide more exact evaluation for integer wavelet transform.Based on blocking and bit-plane, a coding structure with multiple scalabilities was constructed. Group testing have been introduced into wavelet transform image coding, and exhibited its low complexity relative to traditional method based on arithmetic coding. Presented a low complexity entropy coding algorithm, and integrated it into above-mentioned coding structure, all that lead to a brand-new solution for integer transform image coding. Obtained very satisfactory results with software realization.