摘要
为克服传统snake模型不能适应结构复杂的解剖图像、初始轮廓必须充分接近物体边缘的缺点 ,本研究将基于梯度向量流 (GVF)的snake模型用于可视人计划 (VHP)图像中骨组织的分割 ,并修改梯度向量流 (GVF)模型 ,使之适用于彩色图像 ;针对VHP彩色解剖图像数据量巨大的特点 ,将多尺度思想应用到snake模型中 ,以提高处理速度。这种方法提高了计算效率 ,节省了 70 %分割时间 ,得到了理想的精确度 ,对研究解剖结构、组织定量化测定等具有较高的实用意义。
Traditional snake model can not be applied in complicated anatomic image, and the original contour must be close to the real edge of the object. In order to overcome these disadvantages, gradient vector flow (GVF) snake model was used to segment the bone in VHP images, modifying the gradient vector flow model to adapt to the color anatomic image in VHP image. Because of the huge data of the color anatomic image, multiply scale was also used in GVF snake model to speed up processes and improve results. This method improved the computation efficiency, saved 70% of the segmentation time and attained the higher precision, suggesting its significant value for practical application in anatomic structure research and quantified tissue measurement.
出处
《中国医学影像技术》
CSCD
2004年第12期1940-1943,共4页
Chinese Journal of Medical Imaging Technology
基金
本课题受西安市科技攻关基金HJ0 4 0 33
航空科学基金(0 2I530 71 )
高等学校博士点基金 (2 0 0 4 0 6990 1 5)资助。
关键词
可视人计划
图像分割
活动轮廓模型
梯度向量流
多尺度
Visible human project
Image segmentation
Active contour models
Gradient vector flow
Multiply scale
