摘要
为克服在细胞图像自动判读过程中细胞重叠造成的图像分析的困难,需要将交叠在一起的细胞群分离为单个细胞。论文研究了重叠细胞图像分离的几种数学形态学方法,从理论和实验两方面分别讨论了各方法的原理、优缺点和适用性。
To overcome the difficulty caused by the overlapping of cells in the image auto-anglicizing,the touching cells need to be separated into single ones.Several morphologic algorithms of overlapped cells segmentation are present-ed in this paper.The principles,advantages and disadvantages,and applicability of the methods are discussed from both sides of theory and experiment.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第6期57-59,104,共4页
Computer Engineering and Applications
基金
浙江省自然科学基金"基于混沌理论的机器嗅觉研究"(编号:602127)资助
浙江省自然科学基金重点项目"光化学复合敏感的模式识别与集成处理系统研究"(编号:ZD0205)资助
关键词
图像分离
数学形态学
分离方法
image segmentation,mathematical morphology,segmentation methods
