Tutorial: Getting Started with Point Clouds

Introduction

A point cloud is a collection of points with x, y, and z coordinates, which may be spaced irregularly. In FME, a point cloud is treated as a single feature, similar to way FME treats rasters, and many transformers can act on the point cloud as a whole. The points within a point cloud may have attributes that are referred to as components in FME and include properties such as intensity, color, time, etc. LiDAR is a term sometimes used to refer to point clouds but is, in fact, a technology used for capturing point cloud data.

FME can be used to translate point cloud data from one format to another using point cloud readers and writers in a workspace. Furthermore, many transformers capable of working with point clouds are also available in FME for transforming point clouds in a workspace.

For more information on point cloud geometry in FME, please see the Help on this FME Geometry: Point Clouds (IFMEPointCloud)

Point Cloud Basics

Viewing and Inspecting Point Clouds

Point cloud datasets can be opened and viewed in 2D and 3D using the FME Data Inspector. Attributes of the entire point cloud and the components of individual points within it can be inspected using the FME Data Inspector or Visual Preview.

Reading Point Clouds

Point cloud data exists in a number of formats, including both file and database, which can be read by FME in a workspace.

Writing Point Clouds

Building on the previous example, an FME Workspace can be used to translate point cloud data by reading the source data and writing it out to another format.

Thinning and Combining Point Clouds

The PointCloudCombiner is an FME transformer that can combine two or more point clouds into a single point cloud and can also combine other types of geometry into a point cloud. Point Clouds can also be thinned or decimated with the PointCloudThinner transformer.

Using the PointCloudFilter

The PointCloudFilter is an FME transformer that can filter point clouds based on an expression. Both single and multiple filters can be used to manipulate point clouds.

Removing Noise in Point Clouds

Some point clouds contain extra data that wasn’t intended to be there, such as clouds, this is called noise. Using various transformers in FME, the noise in the point cloud can be removed.