Transformers in General and Point Cloud

The idea behind transforming point cloud is as simple as with any other geometry - any transformer that can make a sensible transformation to a point cloud, should work and produce some reasonable output when it receives a point cloud geometry.


This means that FME is "smart" enough to "understand" what is a correct transformation in each particular case. For example, with such transformers as Scaler, Rotator, or Offsetter, the whole point cloud (every single point in it) will be transformed according to the transformer's parameters (see Point Cloud Scenario 9. Offset Rotate Scale).


Bufferer will use point cloud extents to produce a buffer polygon.


BoundsExtractor and CoordinateExtractor will add extent coordinates or coordinates of a specified vertex of the bounding box respectively.


Clipper and Tiler will clip the incoming point cloud according to clipper boundaries or tile sizes (check Point Cloud Scenario 3. Tiling For Delivery for the examples).


ImageRasterizer and NumericRasterizer understand how to make rasters from point cloud based on intensity/color or poin elevation respectively. Visit Point Cloud Scenario 11. Point Cloud Rasterization to see how it works.


All transformers manipulating only attributes (such as AttributeCreator, AttributeSetter, AttributeValueMapper or list transformers) or both geometry and attributes (such as Clipper) work as they do with any other geometry.


Note however, that although it may look like point clouds can sometimes work as polygons, they are not polygons - they do not build a continuous coverage.


In FME 2011, not all transformers are smart enough to work with point cloud directly. For example, Clipper may produce unpredictable results with point clouds as clippers, in this situation we should use BoundngBoxReplacer first.


If you think that some transformer should work differently with point cloud geometry, please let us know (support@safe.com).

Transformers specific to Point Cloud

For FME 2011, we created several specific point cloud transformers:

• PointCloudCoercer
• PointCloudCombiner
• PointCloudCreator
• PointCloudPropertyExtractor
• PointCloudSplitter
• PointCloudThinner

These transformers either manipulate point clouds without changing geometry, or make point clouds from other geometries, or make other geometries from point clouds. In the the various Point Cloud Scenario articles I talk about these transformers in detail.