摘要
研究了一种基于变形模型的肝轮廓提取方法 该方法以人的经验知识作为先验信息 ,利用灰度特征、肝CT序列图像特点、肝区轮廓的整体几何信息作为区域聚合依据进行肝区图像分割 实验表明该方法能够较好地克服噪声和轮廓初始位置的影响 。
The liver contour extraction is investigated based on deformable model The new method uses a prior experience,the gray attribute,the characteristics of serial CT liver images as well as the global information of image contour in segmentation The experiment of CT liver serial images shows that the method can depress the influence of noises and initial contour position so that the satisfactory segmentation results are obtained
出处
《三峡大学学报（自然科学版）》
CAS
2002年第6期529-532,共4页
Journal of China Three Gorges University：Natural Sciences
关键词
变形模型
肝
CT序列图像分割
区域生长
B spline
deformable model
region growing
serial images segmentation
