Learning 3D Point Cloud Histograms

In this paper the authors show how using histograms based on the angular relationships between a subset of point normals in a 3D point Cloud can be used in a machine learning algorithm in order to recognize different classes of objects given by the 3D point clouds. This approach extends the work done by Gary Bradski at Willow Garage on point clouds recognition by applying a machine learning approach to learn the histograms. This approach has been tested on a database of 44 types of IKEA models with 40 samples for each type of object.