摘要
文章研究了基于窄带水平集的曲线演化方法并应用于心脏的MRI图像分割。分析了窄带的生成技术,提出了基于模板的距离函数生成方法;针对MRI图像的特点,给出了一种分割MRI图像的水平集速度函数,以处理图像中出现的区域灰度不一致性以及弱边界现象。MRI图像的分割实验证明了该文算法的有效性。
The curves evolving method based on narrow band level set is studied and applied in segmentation of cardiac MRI in this paper.By analyzing the construction algorithm of the narrow band,a new method based on distant function template is proposed.Because region inhomogeneity,weak edges may emerge in the MRI images frequently,a velocity function is proposed to address the problems.The segmentation results of the MRI images demonstrate the effectiveness of the algorithm listed in the paper.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第27期46-48,72,共4页
Computer Engineering and Applications
基金
香港特区政府研究资助局项目(编号:CUHK/4180/01E
CUHK1/00C)
