Repair Model Based On The Total Variation Multiscale Image Coding System

As we all known, image coding and compressing techniques play an important role in the image storage, processing and transmission. Presently, image coding technologies with using the discrete cosine transform (DCT) and the discrete wavelet transform (DWT) have been widely put into application, all of which belong to the first generation coding system based on the waveform. During the end of last century, the second generation coding system based on the object has become one of the focuses in the image processing field, which is developed to improve image coding technologies with better performance. And due to the remarkable achievements of image reconstruction made in the recent twenty years resorting to partial differential equation (PDE), this novel technique has also been applied in the second generation coding system. Among those image reconstruction methods, such as the "Absolutely Minimizing Lipschitz Extension Based On The Image Level" (V.Caselles, 1998) and the "Total Variation Inpainting Model" (T.Chen, 2002), are the important achievements in this research field.In this thesis, a multi-scale image coding and compressing scheme system is presented based on the total variation inpainting model. The algorithm and function of various component modules are discussed, and the system's main performance is also analyzed. The scheme's core technology is realized by the combination of the predictive coding with total variation inpainting model. The advantage of this scheme is that it not only can form nested bit stream to meet the needs of different information channel capacity, but also can improve the coding efficiency and decoding efficiency. Besides this image coding and compressing scheme, an improved Canny edge detection algorithm based on the curvature-weighted gradient module (CWGM) is proposed to meet the needs of the edge image coding, which is realized by the combination of the gradient modular and curvature of isophotes into an edge characteristic.The results of simulation experiment with adopting typical images show that the algorithm of various functions module is practicable, and can achieve satisfactory coding efficiency. So the scheme proposed in this thesis can be treated as a foundation for further developing an applicable high compression ratio image coding system.