空心涡轮叶片CAD模型与CT切片模型的配准研究

A Study of Registration Method between Hollow Turbine Blades CAD Model and CT Slice Model

针对利用三维CT设备对空心涡轮叶片进行无损检测时存在的定位偏差问题,提出了一种基于三维数字样品的图像模型匹配方法,这对于充分利用有关叶片的先验知识,建立叶片无损检测专家系统具有重要的意义。该方法首先通过图像模型的一阶矩,进行灰度重心匹配;然后将叶片的图像模型与其CAD理论模型之间的残差平方和定义为优化目标函数,采用黄金分割方法搜索其极小值;同时利用叶片上预先指定的结构特征,求出空间变换的位置匹配参数和角度匹配参数,最后根据匹配参数精确地建立起从叶片CT切片模型到其CAD理论模型的空间变换。仿真结果表明:该方法不依赖于初始参数的选取,可达到满意的精度。

There are some position and orientation problems when a3D CT equiment undertakes nondestructive testing to a hollow turbine blade.In allusion to this,a registration method based on a3D digital sample is put forward.It is of significance to make use of blade's information and establish a blade nondestructive testing expert system.First the method proceeds to match gray centroid using one-order moment of image model.Then deviation square sum between blade's image model and CAD model is defined as optimized object function.Authors search the minimum by the golden section method and make use of specified structure character to get the position and orientation registration parameters of space transform.Finally space transform from blade's CT slice model to its CAD model is established.Simulation results indicate that the method doesn't depend on preliminary parameters and it can achieve satisfactory degree.