三维超声扫查截面点云灰度图像的生成方法

Generative Method of Scanning Cross Section Point Cloud Gray Image of with Three Dimensional Ultrasound

针对利用点云数据检测航空发动机叶片缺陷显示及提取困难的问题,提出一种基于平面拟合和坐标变换的叶片截面灰度图像生成方法。首先对超声扫查的截面点云数据通过平面拟合,得到点云拟合平面的法向量;接着通过空间坐标变换,使得点云数据在投影面不发生重叠和累积;然后将变换后的点云数据进行平面投影与网格划分;最后对网格中的数据进行几何分析得到截面的灰度图像,并进行了实验。结果表明,通过空间坐标变换后形成的点云截面图像比采用原始点云数据直接获得的截面图像更加清晰,明确地体现出叶片截面的缺陷和扫查轮廓,更加方便三维显示和观察以及自动化缺陷识别。

For there being problem in detecting aviation engine blade with point cloud data and extracting,a method of blade section gray image is put forward,which is based on flat fitting and coordinate transformation.First,normal vector of fitted flat is got by fitting flat of point cloud data followed ultrasound scanning;then,by space coordinate transformation,point cloud data doesn′t overlap and accumulate on projection plane;next,the point cloud data after transformation is applied with plane projection and mesh generation;finally,after the data in mesh is analyzed,gray image of section is achieved and conducted experiment.It shows that the point cloud section image with space coordinate transformation is clearer than that from original point cloud data,definitely giving defect of blade and scanned profile,so offering more convenience for three-demensional display,observation and automatic defect recognition.