摘要图像分割作为数字图像处理的基本技术之一，一直处在不断改进和发展的状态 中。图像分割的方法有很多种，它们有各自的优点与缺点，但至今仍没有方法能够 全面有效的分割全部种类的图像。近年来，人们将传统的分割方法与各领域的专业 知识相结合，用于解决专业领域的图像分割问题。还有一些根据特定的理论工具而 产生的分割方法，遗传算法就是其中的一种。遗传算法具有能够高效率寻找最优解 的优点，因此能很好的应用于图像分割中。它不仅比传统方法效率高，而且分割效 果也明显优于传统分割方法。遗传算法在图像分割中也存在着缺陷，其在适应度变 化不稳定的种群中，容易出现收敛性差或过早收敛的问题。81551

本文对 Otsu 法原理、遗传算法理论、遗传算法的特点、遗传算法在图像分割中 的应用等方面进行了研究，提出一种基于遗传算法与 Otsu 法的分割方法，并与其他 方法的分割效果进行比较。利用遗传算法的高效寻求最优解的特点来解决最大类间 方差法在多阈值情况下高耗时性的缺陷。同时还对传统遗传算法的寻优过程进行了 改进，能够加快其寻优的过程。通过该方法得到图像在不同阈值下的分割结果，进 行分割效果的比较。

通过 Matlab 仿真，本文提出的方法能够快速得到分割结果，分割的效果和效率 明显都优于传统分割方法。

毕业论文关键词：图像分割；Otsu 法；遗传算法；阈值分割

Abstract As one of the basic techniques of digital image processing, image segmentation has been in the state of continuous improvement and development。 There are many methods of image segmentation, each of them has its own advantages and disadvantages, but there is still no way to fully and effectively segment all kinds of images。 In recent years, people combine the traditional segmentation methods with the professional knowledge in various fields to solve the problem of image segmentation in different professional fields。  There are a number of different methods based on specific theoretical tools, genetic algorithm is one of them。 Genetic algorithm has the advantages of high efficiency in searching for the optimal solution, it can thus be used in image segmentation。 It is not only more efficient than the traditional method, but also is superior to the traditional segmentation method in performance。 Genetic algorithm also has defects in image segmentation。 It is easy to have the problem of convergence or premature convergence in the unstable population。

In this paper, the characteristics of the Otsu method, the theory of genetic algorithm, and genetic algorithm in image segmentation application are studied。 A segmentation method based on genetic algorithm and Otsu method is then proposed。 The performance of this algorithm is compared with that of other methods for segmentation results。 The ability of genetic algorithm to quickly find the global optima of an objective is employed to significantly improve the computational efficiency for maximizing the between class variance objective in the case of multi-threshold segmentation。 At the same time, the optimization process of the traditional genetic algorithm is improved, and the process of optimization can be accelerated。 By this method, the results of image segmentation at different thresholds are obtained, and the results are compared。

By Matlab simulation, the method proposed in this paper can get the segmentation result quickly, and the effect and efficiency of the method are better than that of traditional methods。

Key words: Image segmentation;Otsu;Genetic algorithm;Threshold segmentation

目录

第一章 绪论 1

1。1 研究背景 1