2、,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AbstractIn an image, the scene is often alargenumberof targets, reflectedin many are- asi
3、n theimage.Imagesegmentation is animportant imageanalysistechniqueofwiththeluminancecomponent of the image,simpleandfast.Butitignores alarge partoftheinformation in the image: color,so the poorsegmentationresults.Researchon colorimagesegmentation hasbeen the focus of imageproc
4、essing,which usesa variety of color spacemodel,makingthe segmentation morecomprehensiveandmoreaccurate.Thispaperfirst describes the traditionalimagesegmentation and clusteringalgorithm topartition,andthen focuseson a segmentationmethod basedon the K- meansclusteringalgorithmfo