摘要
为了高精度提取涡轮叶片内外形面点云数据,提出了基于叶片3D-ICT图像的轮廓追踪方法.对轮廓追踪与细化过程中出现的干扰环、孤立点、轮廓片断的处理进行了分析;为了进一步提高轮廓追踪的精度,利用矩匹配方法进行二次追踪,使得轮廓坐标精度达到亚像素级.实验结果表明,该方法可以得到叶片内外形面有序、密集的高精度点云数据.
In order to extract the internal and external surface point clouds of the turbine blade with high accuracy, a contour tracing method based on the 3D-ICT images of the turbine blade is proposed. The techniques to deal with the fake closed contours, isolated points, contour segments during contour tracing and refinement are analyzed. In order to improve the accuracy of the contour coordinates, a moment matching contour tracing algorithm is introduced, which achieves the sub-pixel accuracy. The experimental results prove that through this method, the ordered, dense and accurate point cloud data of the internal and external surface are finally obtained.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2007年第9期1212-1217,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
航空科学基金(03C51079)
中国博士后科学基金(2002037036).
关键词
图像处理
点云数据
矩匹配边界算子
工业计算机断层扫描
涡轮叶片
image processing
point cloud data
moment-matching edge operator
industrial computedtomography
turbine blade
