基于K均值聚类算法的彩色图像分割改进算法毕业论文.doc

成绩本科毕业论文（设计）题目：基于K-均值聚类算法的彩色图像分割改进算法教务处制二○一二年六月摘 要在一幅图像中,景物往往有众多的目标组成,反映在图像中是众多的区域。图像分割属于图像处理中一种重要的图像分析技术。图像分割的传统方法是对灰度图像分割，处理图像的亮度分量，简单快速。但却忽略了图像中很大一部分信息：色彩，因此分割效果不佳。对彩色图像分割的研究一直是图像处理的焦点，它采用各种颜色空间模型，使得图像分割更全面，更精确。本论文首先介绍了传统的图像分割与聚类算法分割，然后重点介绍一种基于K-均值聚类算法的图像改进分割方法。实验结果表明, 改进的分割方法能够实时稳定的对目标分割提取，分割效果良好。关键词：K-均值聚类；图像分割； 聚类算法Abstract In?an?image,?the?scene?is?often?a large number of?targets,?reflected in?many ?are-?as in?the image. Image segmentation?is?an important?image analysis technique of with the luminance component?of?the?image，simple and fast. But it ignores?a large?part of the information?in?the?image:?color, so?the?poor segmentation results.Researchon?color image segmentation?has been?the?focus?of?image processing, which?uses a?variety?of?color?space model, making the?segmentation?more comprehensive and more accurate.This paper first?describes?the?traditional image segmentation?and?clustering algorithm?to partition, and then?focuses on?a?segmentation method ?based on?the ?K-?means clustering algorithm for image?improvement. The?experimental results show that?the?segmentation?method can?improve?real-time stability of?segmentation?to ext-ract?the target?partition?to good effect.Key words: K-means clustering;?image segmentation;?clustering algorithm目录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc325567131序言 PAGEREF _Toc325567131 \h 1HYPERLINK \l _Toc3255671321图像分割综述 PAGEREF _Toc325567132 \h 1HYPERLINK \l _Toc3255671331.1 图像分割技术的现状和发展情况 PAGEREF _Toc325567133 \h 1HYPERLINK \l _Toc3255671341.2 图像分割主要研究方法 PAGEREF _Toc325567134 \h 2HYPERLINK \l _Toc3255671352 K-均值聚类算法 PAGEREF _Toc325567135 \h 2HYPERLINK \l _Toc3255671362.1 聚类概念 PAGEREF _Toc325567136 \h 2HYPERLINK \l _Toc3255671372.2 K-均值聚类算法 PAGEREF _Toc325567137 \h 2HYPERLINK \l _Toc3255671383 基于K-均值聚类的彩色图像分割算法及改进 PAGEREF _Toc325567138 \h 3HYPERLINK \l _Toc3255671393.1引言 PAGEREF _Toc325567139 \h 3HYPERLINK \l _Toc3255671403.2 图像特征提取 PAGEREF _Toc325567140 \h 4HYPERLINK \l _Toc3255671413.2.1 颜色特征的提取 PAGEREF _