Multi-Region Level Set Image Segmentation Based on Image Cartoon-Texture Decomposition Model

摘要：

This paper presents a multi-region level set image segmentation method based on image cartoon-texture decomposition model. The image feature is extracted by using the image decomposition method. We represent the regions by the level set functions with constraint. The coupled Partial Differential Equations (PDE) related to the minimization of the functional are considered through a dynamical scheme. A modified region competition factor is introduced to guarantee no vacuum and non-overlapping between the neighbor regions, it also speed up the cure evolution functions, the final segmentation can be achieved after several iterations. Several experiments are conducted on both synthetic images and natural images, the results illustrate that the proposed multi-region segmentation method is fast and less sensitive to the initializations.

关键词： Multi-region Image Segmentation Level Set Image Decomposition Model PDE Texture

作者: Xue-Min Yin Jian-Ping Guo Chong-Fa Zhong Yu-Hua Yao Zhe Zhang Yi Wei

作者单位: Astronaut Center of China

会议类型: 国际会议

会议名称: 2011 International Symposium on Bioelectronics and Bioinformatics(第二届国际生物医学电子学与生物信息学学术会议 ISBB 2011)

会议地点: 苏州

会议语种:英文

页码: 92-95

在线出版日期: 2011-11-03（万方平台首次上网日期，不代表论文的发表时间）