基于三维点云的车辆残余变形求解

3D point cloud-based calculation of residual deformation of vehicle

为提高事故车辆车身残余变形的测量速度及精度,采用三维激光扫描技术分别获取事故车及原型车的点云数据,利用改进的最近点迭代(ICP)算法配准原型车身与已变形车身2者间相同部件的点云;就变形部位对原型车和变形车进行比较,得到变形轮廓和基准线轮廓,并按照6点法求得残余变形。结果表明,用改进ICP算法能够较大提升处理效率;利用点云数据所得残余变形量较手工方法更客观准确。

In order to overcome the shortcomings of the conventional car crush measure methods which were low in result accuracy and influenced by subjective factors very easily,3D laser scan technology was used to obtain damaged car and the original car point clouds quickly and accurately. Point clouds were registered for the same components in the two cars by the improved ICP algorithm. After a deformed region comparison was made between the two cars,the deformed contour and based line were obtained. Finally the residual deformation was calculated by six positions method. The results show that the improved ICP algorithm is not limited by the initial position and has better efficiency compared to the conventional ICP algorithm.