电子、通信与自动控制

基于变分水平集的显微注射过程细胞分割

展开
  • 1.华南理工大学 自动化科学与工程学院,广东 广州 510640; 2.华南理工大学 自主系统与网络控制教育部重点实验室,广东 广州 510640; 3.华南理工大学 机械与汽车工程学院,广东 广州 510640
秦传波(1982-),男,博士生,主要从事模式识别与智能系统研究.E-mail:tenround@163.com

收稿日期: 2013-10-30

  网络出版日期: 2014-02-19

基金资助

广东省科技厅国际合作项目(2012B050600011);广东高校国际科技合作项目(粤教科函[ 2010] 119);高等学校博士学科点专项科研基金新教师课题(20120172120032);广州市科信局国际合作项目(2012J5100001)

Cell Segmentation in the Process of Cell Microinjection Based on Variation Level Set

Expand
  • 1.School of Automation Science and Engineering,South China University of Technology,Guangzhou 510640,Guangdong,China;2.Key Laboratory of Autonomous System and Network Control of the Ministry of Education,Guangzhou 510640,Guangdong,China;3.School of Mechanical and Automotive Engineering,South China University of Technology, Guangzhou 510640,Guangdong,China
秦传波(1982-),男,博士生,主要从事模式识别与智能系统研究.E-mail:tenround@163.com

Received date: 2013-10-30

  Online published: 2014-02-19

Supported by

广东省科技厅国际合作项目(2012B050600011);广东高校国际科技合作项目(粤教科函[ 2010] 119);高等学校博士学科点专项科研基金新教师课题(20120172120032);广州市科信局国际合作项目(2012J5100001)

摘要

对于显微注射过程中的细胞质分割问题,传统水平集模型无法在多个目标连成一体的情况下分割出细胞质,且分割速度慢.为此,文中首先依据形状差异性特点获取细胞和注射针的联合轮廓,而后结合变分水平集理论增加边缘约束条件,改进能量泛函模型,实现对细胞质的完整分割,同时大幅提升水平集收敛速度.实验结果表明,文中算法可以快速实现细胞质的独立分割,同时避免对其他目标的错误分割.

本文引用格式

秦传波 杜启亮 田联房 张勤 . 基于变分水平集的显微注射过程细胞分割[J]. 华南理工大学学报(自然科学版), 2014 , 42(3) : 15 -20,34 . DOI: 10.3969/j.issn.1000-565X.2014.03.003

Abstract

So far as the cytoplasm segmentation in the process of cell microinjection is concerned,the traditionallevel set models can not directly segment cytoplasm from multiple adhered targets.Additionally,they are slow insegmentation.In order to solve these problems,a new cell segmentation method based on variation level set is pro-posed.In this method,the joint contour of cell and injection pipette is acquired according to the difference inshape.Then,by combining the variation level set theory,an extra edge constraint is added to improve the energyfunctional model.A complete cell segmentation is thus achieved and the convergence speed of the variation level setis significantly improved.Experimental results indicate that the proposed method helps to perform the independentcytoplasm segmentation rapidly and at the same time avoid wrong segmentation of other targets.

文章导航

/