Application Of Genetic Algorithm For Digital Image Processing

Genetic algorithm, as a computational model simulating the biological evolution process of the genetic selection theory of Dar-win, is a whole new global optimization algorithm and is widely used in many fields with its remarkable characteristic of simplicity, commonability, stability, suitability for parallel processing, high-efficiency, and practibility . On the other hand, there are many op-timization problems in the field of digital image processing, such as image compression, pattern-recognition, image rectification, image segmentation, 3D image recovery, image inquiry, and or so. In fact all these problems can be generalized as the problem of searching for a global optimal solution in a large solution space, which is the classic application field of genetic algorithm. On the major premise of feasibility of this theory, this article based on the practice of fore-runners, has done some further research work about the application of genetic algorithm for image compression and pattern-recognition with a satisfactory result.This article is divided into three part . The first chapter in-troduces research history of genetic algorithm and biological back-ground briefly at the beginning , then discuss the basic realization method of the algorithm , its operators , its basic characteristic ,its applications as well as some thoughts of the author about how to use genetic algorithm to solve problems .In the second chapter,some relative basic theories in wavelet analyse is first proposed in brief , involving the cause of develop-ment of wavelet theory (that means its good ) , the definition of continuous wavelet , discrete wavelet and dyadic wavelet and the multiresolution analysis for constructing orthogonal wavelet .Then the biorthogonal wavelet and method of its application for imagecompression is discussed .Based on the research of genetic algorithm for the constructing method of wavelet filters ,the method of con-structing optimal wavelet filters for the specific image compression is put forward , so the problem about image optimal compression is solved successfully with the combination of genetic algorithm and wavelet used in image compression.In chapter three, after making clear the property of multi-variable objective optimization of pattern-recognition , realization methods of genetic algorithm, applicated for pattern-recognition of figures and images, are put forward separately and gradually for regular figure, complicated figure and image.