Splicing Of The Fundus Image Based On Wavelet Analysis

Wavelet analysis is an effective analytical instrument, which has aroused the highly attention of scientists and researchers in various fields and subjects, resulting in notable achievements. And digital image processing by the aid of wavelet analysis is one of the hot issues of wavelet research and application.After a brief description of Fourier transform and Gabor transform, the paper introduces the history of wavelet development and the basic theoretical knowledge of wavelet, and especially Mallat Algorithm is interpreted in detail. Mallat Algorithm is the achievement of QMF (Quadrature Mirror Filter) and wavelet transform. It carries out the fast wavelet's arithmetic and its function is beneficial to wavelet just like FFT to Fourier transform. Mallat Algorithm makes the complex wavelet to be implemented in computer.In this thesis, we mainly discuss the topic in three parts: image preprocessing, image matching, image mosaic. After having a research on wavelet analysis in two-dimensional signals, it points out that it is a good method that compounds wavelet method with image traditional matching. At the beginning of the work, the thesis advances some methods to build up the image and pick up the module to match another image. Finally, and smooth operator is used to make the images mosaic really smooth and the computer program of the work is completed. These methods contain higher practical value.