中文摘要：

工件自动配准在柔性生产装配中至关重要,其中装配工件CAD模型数据和实测点云数据的配准是关键技术之一。针对装配工件和CAD模型的配准问题,提出了一种基于工件四特征点的粗配准算法。获得配准物体CAD模型数据和实测点云数据后,该方法包含四步：点云数据取样,特征四点集提取,特征一致四点集寻找,计算变换一致性矩阵。以正方体为模型的仿真实验结果显示该配准方法正确可行,以维纳斯石膏像作为实验对象进行配准实验,当点云数据为80000点时,点到点的误差均值为0.0622mm。该算法稳定,配准结果可以作为精确配准算法ICP（Iterative Closest Point）等的叠代初值。

英文摘要：

Work-pieces automation registration is critical in flexible manufacture assembly,and the registration of CAD model with measurement data cloud of its part is one of key technology. Acoarse registration algorithm based on four feature points of work-piece is proposed to accomplish its registration. After CAD model data and actual measurement data of a work-piece are acquired,the method consists of the following four steps： sampling of point clouds data,extracting of four feature points set,searching of congruent four feature points set andcomputing of transformation congruent matrix.Simulation experiment of the registration algorithm is carried out using the cube as the theoretical model,and the result show that the algorithm is verified correct.Another experiment is done with Venus plaster figure,and the mean-error of point-to-point distance is 0.062 mm accessing measurement data more than 80000 points. The experiment also shows that the algorithm is more efficient and robust,and can further be used as preparatory step of precision registration such as Iterative Closest Point（ICP）algorithm.