Level Set Based Image Coding Method

Digital image coding plays a very important role in the field of image processing and communication. Its purpose is to minimize the amount of data of the image. The first generation image coding method is based on waveform and has satisfactory results. The second generation coding method based on human visual is also increasingly becoming systematic and shows a good performance, including Wavelet Transform Coding which is popular recently, edge-based "Alternating Projection" algorithm of Mallat and the TV inpainting model coding based on Partial differential Equations. However, another method which uses level set and AMLE method to achieve image coding and compression is also studied in this thesis.In this thesis, we first introduce the applications of PDE in image coding and the principles of image interpolation. Based on the principle of AMLE, the principles of the existing program proposed by Caselles which apply level set and AMLE method to image coding are analyzed. The shortage of current scheme is also pointed out in this thesis. The primary works we do are as follows:First, an improved coding scheme based on the level set method, is proposed on the basis of the analysis. The scheme adapts layer-by-layer processing method on the data of level set to achieve image reconstruction and coding. We also discuss the feasibility and the numerical realization of this theory.Second, Chain coding is also applied in this coding system. The principle of the program which encodes the location of image level set boundaries is discussed. We make a preliminary estimation about coding system proposed in this paper.Finally, the methods of improving the compression performance are also studied in this thesis based on the improved scheme.