Method for extracting planes from 3D point cloud sensor data

摘要

A method extracts planes from three-dimensional (3D) points by first partitioning the 3D points into disjoint regions. A graph of nodes and edges is then constructed, wherein the nodes represent the regions and the edges represent neighborhood relationships of the regions. Finally, agglomerative hierarchical clustering is applied to the graph to merge regions belonging to the same plane.

主权项

1. A method for extracting planes from three-dimensional (3D) points, comprising the steps of:
acquiring a depth map with a 3D sensor, wherein the depth map has a two-dimensional grid of pixels and each pixel has a depth value; back-projecting the depth map to generate a cloud of 3D points; partitioning the 3D points into disjoint regions; constructing a graph of nodes and edges, wherein the nodes represent the regions and the edges represent neighborhood relationships of the regions, wherein each node includes a set of 3D points, a plane fitted to the set of 3D points, and a mean squared error (MSE) of the set of 3D points to the plane; and applying agglomerative hierarchical clustering to the graph to merge regions belonging to a same plane, wherein in each iteration of the agglomerative hierarchical clustering, a node with a minimum MSE is selected and merged with one of neighboring nodes having a minimum merging MSE among the neighboring nodes, wherein the steps are performed in a processor.