Introduction

There are two types of point clouds: organized and unorganized. These describe point cloud data stored in a structured manner or in an arbitrary fashion, respectively. An organized point cloud resembles a 2-D matrix, with its data divided into rows and columns. The data is divided according to the spatial relationships between the points. As a result, the memory layout of an organized point cloud relates to the spatial layout represented by the xyz-coordinates of its points. In contrast, unorganized point clouds consist of a single stream of 3-D coordinates, each coordinate representing a single point. You can also differentiate these point clouds based on the shape of their data. Organized point clouds are M-by-N-by-3 arrays, with the three channels representing the x-, y-, and z- coordinates of the points. Unorganized point clouds are M-by-3 matrices, where M is the total number of points in the point cloud.

Unorganized to Organized Conversion

Most deep learning segmentation networks, such as SqueezeSegv1/v2, RangeNet++, and SalsaNext, process only organized point clouds. In addition, organized point clouds are used in ground plane extraction and key point detection methods. This makes organized point cloud conversion an important preprocessing step for many Lidar Toolbox™ workflows.

You can convert unorganized point clouds to organized point clouds by using the pcorganize function. The underlying algorithm uses spherical projection to represent the 3-D point cloud data in a 2-D (organized) form. It requires certain corresponding lidar sensor parameters, specified using the lidarParameters object, in order to convert the data.