一种板型物体混叠场景的快速分割算法

A Fast Segmenting Method for Scenes with Stacked Plate-Shaped Objects

针对多个板型物体混叠摆放的场景,提出了一种快速有效的分割算法。该算法充分利用有序点云的特点,将自顶向下以及自底向上的分割策略结合,根据三维点的空间位置和法向量,利用随机采样一致性(RANSAC)算法从三维点云数据中快速提取平面点集;然后将提取的平面点集所对应的图像坐标映射为二值图像,通过连通区域分析将其分割为多个连通的平面区域;接着利用"胶水"算法对这些区域进行快速合并,并对较大的弱连接连通区域进行断裂修正,得到最终的分割结果。实验结果表明:与区域生长算法相比,所提算法的分割结果更优,且算法效率大幅提升。

A fast and efficient segmentation algorithm is proposed for scenes in which multiple plate-shaped objects are placed in an overlapping manner. The algorithm makes full use of the characteristics of the ordered point cloud, and combines the top-down and bottom-up segmentation strategies. The Random Sample Consensus(RANSAC) algorithm is used to quickly extract the three-dimensional planar point set according to the spatial position and normal vector of the three-dimensional point from the three-dimensional point cloud. The image coordinates corresponding to the extracted planar point set are mapped into a binary image, and are divided into a plurality of connected planar regions by the connected region analysis. Then, the glue algorithm is used to quickly merge these regions, and the larger weakly connected regions are subjected to the fracture correction, so as to obtain the final segmentation result. The experimental results show that compared with the region growing algorithm, the proposed algorithm can obtain better segmentation results, and the algorithm efficiency is greatly improved.