基于动态方向梯度矢量流蛇模型的图像分割  

Medical Image Segmentation Based on Dynamic Directional Vector Flow Snake Model

作　　者：邬立保[1] 董育宁[1] 俞海平[2] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京鼓楼医院,江苏南京210008

出　　处：《电视技术》2009年第S1期139-141,157,共4页Video Engineering

摘　　要：提出了一种改进的基于动态方向梯度矢量流蛇模型的MR脑肿瘤图像分割新方法,实验结果表明该方法能够有效地分割提取出脑肿瘤图像。为了对MR脑肿瘤图像形状特征进行描述,提出了一种改进的基于Zernike矩的快速算法,为进一步诊断病情提供了有效的帮助。A new scheme for segmentation and distillation from MR encephaloma image sequences based on improved dynamic gradient vector flow(DDGVF) snake model is proposed in this paper.Experimental results prove that it segment the encephaloma image sequences effectively.For further description of encephaloma image sequences characteristic which could be helpful in diagnosing the patients,an improved fast algorithm based on Zenike moments is also proposed.

关 键 词：活动轮廓模型 SNAKE模型 梯度矢量流 动态方向梯度矢量流 Zenike矩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]