Research Of Image Matching Method Based On Improved Genetic Algorithm

As one of the main branches of evolutionary computation, genetic algorithm uses principle of biological evolution to solve application problem in our real live. Genetic algorithm was used in many subjects with its many advantages, such as simplicity, essential parallelism and good robust. Image matching is an important research direction of image processing and computer vision. It is a mapping in space domain and lighting, its purpose is finding the optimum coordinate conversion in space domain and lighting conversion. But the traditional matching method, such as template match, when there are differences between template image and matching image in noises effects and lighting etc, matching algorithm can not satisfy with time and accuracy at the same time. This paper summarizes the principle and the development direction of the genetic algorithm. Based on the traditional matching method, an improved genetic algorithm, which introduces niche and adaptive genetic algorithm, was presented. The improved algorithm was used for imaging matching was successful.This thesis will mainly complete works as follow:Firstly, introduce the biological background and its improvement history, based on the basic principle of the genetic algorithm, and presents the basic genetic algorithm's operation procedure. After realizing and analysis the present and improvement, a new improvement genetic algorithm was presented.Secondly, introduce the relevant theory and analysis the traditional represented algorithm, on account of the problems in the traditional area of image matching, a image matching algorithm based on improvement genetic algorithm was introduced.Thirdly, use the standard genetic algorithm in traditional image matching and verified its effectiveness. Then use improved algorithm in image matching, the result indicated that the local and global optimization are stronger, and the searching performance is better than the standard.This work combine with genetic algorithm and image matching, solve the problems in traditional image matching, promoted matching efficiency and have stronger significance in practical application.