基于Otsu多阈值和分水岭算法的乳腺肿块分割  

Mammographic mass segmentation based on Otsu multi-threshold and watershed algorithm

作　　者：张召长[1] 戴耀东[1] 周正东[1] 

机构地区：[1]南京航空航天大学材料科学与技术学院,南京210016

出　　处：《计算机应用》2008年第S2期198-200,共3页journal of Computer Applications

摘　　要：在乳腺X线图像中,肿块大多埋没在复杂的、高密度的腺体背景中难以检出和识别。针对这一问题,提出了一种应用快速多阈值Otsu算法和改进的分水岭算法相结合的乳腺X图像肿块分割方法。首先对乳腺X图像进行形态学预处理,然后应用多阈值Otsu算法对图像进行处理,并对图像运用分水岭算法进行分割。为了避免过分割现象,对分割后的图像进行了相似区域合并。实验表明该方法对于乳腺X图像中各种类型的可疑肿块的分割都是有效的。Masses are usually difficult to be detected as they often superimpose on dense structured background of mammography.In order to solve this problem,an approach of mass detection based on fast Otsu multilevel threshold and modified watershed algorithm was presented.First,mammography was processed with morphological operators,then the method of fast Otsu multi-threshold algorithm was applied;second,the watershed algorithm was used for the segmentation.Finally,regions merging algorithm is applied to avoid over-segmentation.The experimental result show that the proposed approach is applicable for all kinds of mammographic mass detection.

关 键 词：乳腺X图像 图像分割 OTSU 分水岭 区域合并 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]