基于X射线断层成像的逆向工程中轮廓序列分割算法的原理及实现

Principle and Realization of Segmenting Contour Series Algorithm in Reverse Engineering Based on X-Ray Computerized Tomography

为了实现基于X射线断层成像(CT)的逆向工程中具有参数识别的三维图像重构,提出了一种分割轮廓序列的新算法．首先通过一定角度的射线法来得到轮廓间的嵌套关系,然后采用扫描一次关系矩阵生成轮廓树的方法实现层内轮廓定位,最后运用轮廓间定量、定性的属性判定来完成层间的轮廓匹配,通过实例,本文提供的算法可以准确、快速地分割CT零件中的轮廓序列．

A new algorithm of segmenting contour series of images is presented, which can achieve three dimension reconstruction with parametric recognition in Reverse Engineering based on X-ray CT. First, in order to get the nested relationship between contours, a method of a certain angle ray is used. Second, for realizing the contour location in one slice, another approach is presented to generate the contour tree by scanning the relevant vector only once. Last, a judge algorithm is put forward to accomplish the contour match between slices by adopting the qualitative and quantitative properties. The example shows that this algorithm can segment contour series of CT parts rapidly and precisely.