GVF Snake模型在细胞核图像分割中的应用  

Application of improved GVF Snake model in automatic segmentation of cell nucleus

作　　者：张培培[1] 包振健[2] 张更路[1] 

机构地区：[1]济南军区总医院医学工程科,山东济南250031 [2]北京凌云光视数字图像技术有限公司,北京100097

出　　处：《中国医学装备》2010年第4期11-14,共4页China Medical Equipment

摘　　要：在对骨髓细胞核分割技术研究的基础上,提出了一种改进的GVF Snake模型应用于细胞核分割。该模型针对骨髓细胞图像特有的双轮廓结构,通过在传统的梯度向量场中添加动态的轮廓牵引力,克服了传统GVF Snake模型存在的初始轮廓选取对临界点敏感的问题,在骨髓细胞核分割的应用中,消除了细胞外轮廓对细胞核分割的影响。Based on the study of bone marrow cell nucleus segmentation technique,an improved GVF Snake is proposed and applied to cell nucleus segmentation.According to the dual-contour structure of the bone marrow cell image,a dynamic contour traction is imported into the traditional Gradient Vector Flow,through which the problem that the initial contour is allergic to the critical points in the traditional GVF is solved.Then affection of the outer contour is overcome and the cell nucleus can be split successfully in the application of bone marrow cell nucleus segmentation.

关 键 词：图像分割 骨髓细胞图像 梯度向量场 主动轮廓线模型 

分 类 号：TH788[机械工程—仪器科学与技术]