计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (19): 16-18.

• 博士论坛 • 上一篇    下一篇

一种新的水平集图像分割方法

任继军,何明一   

  1. 西北工业大学 电子信息学院 陕西省信息获取与处理重点实验室,西安 710072
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-07-01 发布日期:2007-07-01
  • 通讯作者: 任继军

Novel level set method to image segmentation

REN Ji-jun,HE Ming-yi   

  1. School of Electronics and Information,Northwestern Polytechnical University,Shaanxi Key Laboratory for Information Acquisition and Processing,Xi’an 710072,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-01 Published:2007-07-01
  • Contact: REN Ji-jun

摘要: 对Chan-Vese提出的基于简化Mumford-Shah区域最优划分模型和测地线主动轮廓模型在水平集框架下的物理机理进行了分析,在充分考虑其模型优点的基础上,通过构造新的能够整合局部边缘信息和全局区域信息的演化函数对上述模型所存在问题进行了针对性处理,得到了一种新的水平集图像分割模型。人工合成图像和红外光学图像的仿真结果表明,在同样的模型参数条件下,该文模型具有比传统CV模型和GAC模型更高的演化效率和分割质量。

Abstract: A new level set PDE based on the simplified Mumford-Shah model for image segmentation is proposed by Chan and Vese(CV model) and Geodesic Active Contours(GAC) mode.In order to overcome limitation of both models,the evolution function is constructed based on local edge information and the global region information and then novel level set model is presented.Experimental results of synthesis image and infrared image suggest that our model is more efficient and accurate in its segmentation operations.