形态学与测地距离函数应用于CT断层图像插值

Morphology and geodesic distance function applied in CT faulted image interpolation

现有的插值方法在进行断层图像插值时,要么不能兼顾灰度和形状的变化,要么计算量太大。为解决这一问题,提出一种基于数学形态学、测地距离函数与向量积点匹配相结合的插值算法。通过对已知两幅断层图像进行分割,得到目标与背景的区域后,利用数学形态学与测地距离函数相结合的方法确定出插值图像目标与背景区域的轮廓,然后在每个轮廓内运用向量积点匹配插值方法得到插值点的值,从而得到整个插值图像。实验结果表明,该算法插值出的图像不仅视觉效果好而且使计算量极大减少。

In the case of image interpolation for 3D volume models,existing methods either lack of the capability of interpolation gray levels and shapes at the same time,or need higher computation cost.In order to solve the problem,image interpolation methods based on the mathematical morphology,geodesic distance function and vector on point matching are proposed.The original images are segmented object and background with the threshold segmentation.Then the contours of every region in the interpolated image are determined with the help of mathematical morphology and geodesic distance function.Finally,the values of interpolated points in the contours are obtained with the vector on point matching interpolation.Further,the whole image is obtained.Experiments prove that the proposed algorithm greatly improves the quality of image and has much lower computation cost.