提出一种用于三维体数据体积、表面积测量的断层间边缘拟合插值算法。该算法首先对每一切片找出其轮廓线的几何形心,将边缘点坐标转化成极坐标;之后根据ρ-θ的关系对边缘点进行拟合插值,在层间边缘点建立一一对应的关系,最后通过对相邻断层间匹配边缘点极坐标中的ρ进行插值,从而生成中间层的边缘点的极坐标。与其他插值算法相比,该算法只生成并记录各断层的边缘点,减小了数据冗余,插值速度快,且精度好。在三维测量的实验中证明了该算法相对于其他插值算法在速度和精度上的优势。
A new three-dimensional image interpolation scheme for area and cubage measurement is presented. In the algorithm, firstly the geometric center of every original slice is acquired, and the coordinates of edge points are converted into polar coordinates. Then it interpolates and fits between edge points according to the function ρ = F(θ) , to create the relation between edge points in every slice. At last it interpolates the radiuspbetween every corresponding edge points in neighboring slices to create the polar coordinates of edge points in unknown slices. Compared with other interpolation schemes, the method only generates and records the edge points in each slice for computing area and cubage, so it greatly minimizes the data involvement, increases the interpolating speed and improves the accuracy. Three-dimensional image measurement experiments prove that the proposed algorithm has advantages both in speed and accuracy compared with other algorithms.