基于MATLAB的图像分割算法研究与仿真.doc

PAGE I 摘 要 随着信息化程度的加深,图像作为包含大量信息的载体形式越来越能体现出它强大的信息包含能力。图像分割是进行图像分析的关键步骤，也是进一步理解图像的基础。图像分割在图像工程中位于图像处理和图像分析之间。图像分割就是把图像分割成一系列有意义、各具特征的目标或区域的过程。虽然图像分割算法有很多种，但是到目前位置还没有通用的图像分割算法。 论文综述了阈值分割、区域生长、分裂合并和边缘检测图像分割算法。首先介绍了图像分割的研究意义及国内外研究现状。接着介绍了阈值分割、区域生长、分裂合并和边缘检测图像分割算法的基本原理及其MATLAB仿真结果。最后介绍了MATLAB图形用户界面设计。 MATLAB仿真结果表明：各种图像分割算法各有其优缺点及其适用范围。因此，在对一幅图像进行分割之前，首先要对图像进行判断分析，然后再选择恰当的分割算法，这样才能得到比较令人满意的分割效果。 关键词：图像分割；阈值分割；区域生长；分裂合并；边缘检测；MATLAB PAGE IV PAGE III Abstract As a media containing great amount of information, image embodies its powerful ability with the deepening of informationization. Image segmentation is not only a key step before image analysis, but also the basis of further understanding image. Image segmentation locates between image process and image analysis. The process of dividing an image into a series of meaningful and characteristic objects or regions is called image segmentation. Although there are many kinds of image segmentation algorithms, there is no general image segmentation algorithm by now. This paper summarizes such image algorithms as threshold segmentation, region growing ，split and merge as well as edge detection. Firstly, research meaning and state around the world of image segmentation is introduced. Secondly, basic principle and MATLAB simulation results of threshold segmentation, region growing, split and merge as well as edge detection image segmentation algorithm is introduced. Finally, graphic user interface of MATLAB is introduced. The MATLAB simulation results demonstrate that different image segmentation algorithms are of different advantages and disadvantages as well as its application. Therefore, image should be analyzed first and then appropriate image segmentation algorithm is selected before segmentation so as to acquire satisfactory results. Key words: image segmentation；threshold segmentation; region growing; split and merge ; edge detection; MATLAB 目 录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc230703484 1 引 言 PAGEREF _