Research On The Grey Face Image Encoding Based On Wavelet Transform

In recent years , wavelet analysis has developed greatly in theory and application .In many fields ,such as phonetic analysis ,signal processing, quantum mechanics and non-linear science ,it has been widely applied and especially on the image compression fields ,which is focused on by experts home and abroad .So the study on wavelet theory is important means on the image compression .In this dissertation ,main focus of research is used to process face gray images compression based on wavelet analysis .Firstly this dissertation introduce wavelet theory and multiresolution analysis ,and then from two keys link of images compression: wave filter(wavelet transform), coquarzation encoding(algorithm), four respect ,the feature and construction of orthogonal PR-FIR.: the choice of wavelet bases(filter) ,the characteristic and the coefficient of wavelet transformation for digital images and algorithm ,this dissertation studies images compress based on wavelet transform systematically and detailedly ,and give new improvement and enhancement .The experiment result are given ,it shows that the new methods is simple and reliable.Firstly ,a brief survey of wavelet theory is given, the research background > present situation and research results ,both domestic and abroad ,of the method of image compression are surveyed.In chapter one ,we first discuss wavelet theory and multiresolution analysis , and focus our studies on the discrete processing of the continuous wavelet transform and the decomposition and reconstruction with dyadic wavelet based on multiresolution analysis.In chapter two , I discuss the relation between orthogonal wavelet bases and FIR filter ,and analyze characteristic of compactly supported orthogonal and biorthogonal wavelet ,and then I study the methods of wavelet bases construction from two angles which are frame of multiresolution analysis and characteristic of filter ,and constructbiorthogonal wavelet bases of compactly supported wavelet with B-spline and matrix .This dissertation put forward to and approve the condition that mirror filter become convergent wavelet , and the experiment result are given ,it shows the new methods is reasonable .In chapter three ,firstly we process digital images by wavelet transformation and from three aspects which are keeping of linear phase ,simpleness and particularity of calculation and efficiency of images compression discuss the choice of wavelet bases .we study effect of the orthogonalization ,symmetry regularity ,high order vanished moments of wavelet and the main-wave to side-wave ratio of wavelet in images compression , and the experiment result are given .In chapter four , firstly we process face images by wavelet transformation and reconstruct sub-images with coefficient of sub-band , and the we discuss the feature of reconstructing sub-images and of distribution of coefficient of sub-bands on the basis of characteristic of statistical data from experiment and of face images .It provides convenience for subsequence images compression.In chapter five ,from two angle which I improved efficiency of compression and reduced distortion of reconstruction ,we study algorithms of coding ^specially SPIHT algorithm , and give analysis and improvement in bit assignment ,data organization and aggregation divide by statistical data from experiment .conjoining idea of embedded coding ,we give new algorithm of coding .the experiment result show it superior to SPIHT algorithm .Wavelet theory and application in images are not perfect in many fields , Now a number of researchers continue to study in the theory and application of wavelet .