When comparing the features of points in 3-dimensional point cloud recognition process, spatial attitude transform must be taken into account due to different attitudes of point cloud, but the old algorithm of spatial attitude is of high complexity and of high computation cost. To deal with the above issues, we propose an attitude correction algorithm for similar objects. It chooses three groups of non-collinear corresponding points at the position the two similar objects are to be identified, through space transformation and rotation it unifies the attitudes of similar parts to set up the base for objects recognition. The feasibility and validity of the algorithm are proved by the correction using point cloud model of the car.%在三维点云识别过程中,由于点云姿态不同,比较点的特征时,要将空间姿态变换计算在内,旧的空间姿态算法复杂度高,计算量大.针对上述问题,提出一种相似物体的姿态矫正算法,通过在两个相似物体要进行识别的位置上选择不共线的三组对应点,经过空间变换与旋转,将相似部位的姿态统一,为物体识别做基础.通过使用汽车点云模型进行矫正,验证了该算法的可行性和正确性.
展开▼
