Fast Planar Surface Detection in Point Clouds Using Voxel-grid-center-constrained Iterative Adaptive Plane Extraction

Plane detection is a key pre-processing step in point clouds for many applications, such as buildings registration and automous cars. Many methods have been proposed for plane detection in point clouds, but most of them cannot meet the requirement of speedability. Therefore, in this paper, we propose a voxel-grid-center-constrained iterative adaptive plane extraction method that can meet the speedability without losing the accuracy of the detected planes. First, the point cloud is subdivided using an octree, and then a voxel grid is randomly selected for coplanar testing to find the plane in the point clouds. Second, in the plane detection phase, a voxel grid center constraint strategy is used to improve the efficiency of plane detection. Finally, an iterative adaptive plane extraction technique is used to improve the robustness. Experimental results demonstrate that the proposed method can achieve high performance on the synthesis datasets as well as on the real datasets.