Tutorial: Point Cloud Transformations

Introduction

The following tutorial cover point cloud transformations and more advanced workflows. For an introduction to the point cloud format, please see the help documentation on this FME Geometry type: Point Clouds (IFMEPointCloud), or read through the Tutorial: Getting Started with Point Clouds.

Within FME, point cloud geometries are an efficient format because they are treated much like rasters - as a single feature. This makes them useful for intermediate analysis and transformations within workflows, as they are fast and contain attributes that model continuous data, such as elevation.

Two common point cloud transformations that take advantage of these characteristics are:

• point cloud to 3D surface
• raster to point cloud

If you’re new to point clouds, the latter transformation might be surprising, but there many uses for point clouds other than reading and analyzing source point cloud data! Point clouds can be used as an intermediary or output of a 3D model in order to analyze or isolate particular components of the model, as is demonstrated in the Creating Point Clouds from 3D Models or Raster Data tutorial.

In addition, LAS files can be manipulated in FME to create 3D surfaces or building models, and raster colour images or DEMs can be transformed into point clouds for analysis or 3D visualization. We’ll take a look at all of these transformations and introduce a few key point cloud transformers in the tutorials below!

Articles

Point Cloud to 3D Terrain Model with Buildings

This article shows how to create a surface model and extract building heights using point cloud data.
 

Creating Point Clouds from 3D Models

This article shows how 3D models can be converted to a point cloud with FME using the PointCloudCombiner transformer.
 

Creating Point Clouds from Rasters

This article shows how RGB and numeric raster data can be converted to a point cloud with FME using the PointCloudCombiner transformer.
 

Using the PointCloudFilter

This article shows how to use the PointCloudFilter transformer to filter out a roof, as well as how to use mutiple filters to filter out an elevation range. 
 

Creating Boundary and Point Features from a Point Cloud

This article walks through how to create boundary polygons or point features from the classification component in point cloud data. 
 

Using the PointCloudCoercer to Convert Point Clouds

This article is a demo on how to convert point clouds into either a multipoint feature, or individual points to write out to a format where a point cloud isn't supported.
 

Setting a Point Cloud's Color Components

This article demonstrates how to set and work with a point cloud’s colour components, including applying colors from a raster, and applying expressions.
 

Setting a Point Cloud's Intensity Component

This article demonstrates how to set or correct a point cloud’s intensity component.
 

Re-classifying a Point Cloud

This article demonstrates how to set and work with a point cloud’s classification component.
 

Creating Rasters and DEMs from Point Clouds

This article demonstrates how to convert a point cloud into a grayscale raster and a digital elevation model (DEM).
 

Additional Resources

PointCloudExpressionEvaluator documentation : Has examples on how to use this transformer