目标组织轮廓的三次非均匀B样条逼近

Cubic Non-uniform B-Spline Approximation for Object Contour

提出断层医学图像中目标组织轮廓最少数据点描述的样条曲线建模方法 .首先采用多边形逼近将图像分割和轮廓跟踪所获得的目标组织轮廓线上几乎共线的部分离散数据点去除 ;再以剩下的轮廓数据点作为型值点 ,采用周期 B样条曲线逼近的方法 ,得到控制误差条件下最少数据点描述的目标组织轮廓曲线 ,并依照 IGES标准 。

A compact object boundary spline curve modeling method for sectional medical images was proposed. Polygonal approximation is first applied to the discrete points collected from image segmentation and contour tracing, and the points that are almost collinear with others are deleted as a result. In the second stage, periodic dosed B-spline approximation is adopted to further reduce the data points, and the most compact B-spline boundary curve within the given tolerance is obtained. Boundary spline curves can be output as IGES file to provide a geometric model for successive procedures such as medical implant design and bone tissue finite element method analysis.