Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (28): 196-198.DOI: 10.3778/j.issn.1002-8331.2009.28.059

• 图形、图像、模式识别 • Previous Articles     Next Articles

Realization of infrared image segmentation based on improved Chan-Vese model

ZHANG Yi-xiong1,LI Xi-ying2,3   

  1. 1.Physics and Electronic Engineering Department,Hanshan Normal University,Chaozhou,Guangdong 521041,China
    2.School of Engineering,Sun Yat-sen University,Guangzhou 510275,China
    3.State Key Laboratory of Optoelectronic Materials and Technologies,Sun Yat-sen University,Guangzhou 510275,China
  • Received:2009-04-14 Revised:2009-06-08 Online:2009-10-01 Published:2009-10-01
  • Contact: ZHANG Yi-xiong

Chan-Vese可变模块法快速实现红外图像分割

张奕雄1,李熙莹2,3   

  1. 1.韩山师范学院 物理与电子工程系,广东 潮州 521041
    2.中山大学 工学院,广州 510275
    3.中山大学 光电材料与技术国家重点实验室,广州 510275
  • 通讯作者: 张奕雄

Abstract: This paper analyzes the speed of the Chan-Vese method.However,as the simplified model of Mumford-Shah,Chan-Vese method will result in large amount of calculation,seriously impact the speed of curve deformation,due to the full-range changes of level set function in Chan-Vese model processing.This paper draws on the variable masking method to raise the speed of Chan-Vese model.The experiment result of the infrared image segmentation demonstrates the efficiency and effectiveness of the improved method.

Key words: improved Chan-Vese method, infrared image, image segmentation

摘要: Chan-Vese方法作为Mumford-Shah的简化模型,因其水平集函数在演化过程中的全区域变化,会造成计算量极大,严重影响曲线形变速度;即使采用窄带法,在像素大的图像处理中,演化速度也极大地受到限制。提出一种改进Chan-Vese可变模块法快速实现红外图像分割,提出了形变曲线的不同节点根据前一次迭代演化情况,采用不同大小的模块运算,从而极大地提高了Chan-Vese模型演化速度。通过对红外图像的实验仿真,证明了该改进C-V方法能达到理想的效果。

关键词: 改进Chan-Vese方法, 红外图像, 图像分割

CLC Number: