Skip to main content

Hi there,

I have created a Mosaic Dataset in ArcGIS Pro from downloaded 1m DSM GeoTIFFs. This displays fine in ArcGIS Pro.  The boundary of the mosaic is clipped to the extent of the data I want to analyze.  

I need to convert this to LAS to run some ArcGIS GP tools. So I have done this using the Mosaic Dataset as an input in FME:

Esri Geodatabase Raster  → Point Cloud Combiner → LAS Writer

The output shows some incorrect data striping and looks like it missed some areas. In ArcGIS Pro:

In FME Inspector:

I don’t see any out of memory errors or issues with reading the Mosaic data in the FME logs.

Any ideas?

Thanks,

Marc

Have you checked the parameters on the Reader, assuming its a fGDB source https://docs.safe.com/fme/html/FME-Form-Documentation/FME-ReadersWriters/geodatabase/GEODATABASE_FILE_RASTER_DATASET_reader.htm

there's sheetings in the parameter for method… order...etc

Then also read up in detail about https://docs.safe.com/fme/html/FME-Form-Documentation/FME-Transformers/Transformers/pointcloudcombiner.htm

Rasters are converted as follows:

  • The x and y components are created from the cell ground coordinates, or columns and rows if the raster has no coordinate system.
  • The first selected numeric band will become the z component.
  • The first selected bands with red/green/blue/gray interpretations will become the color_red/color_green/color_blue components.
  • Additional selected bands will also be preserved. If the band has a name, the component name will be the band name. If the band has no name, the component name will be bandN, where N is the band index.
  • Nodata may optionally be extracted

Thanks for your response.

I have left everything in the mosaic dataset reader as default, also the pointCloudCombiner.  As far as I can tell, this should read all the rasters in the mosaic and convert them to a point cloud.  There is nothing complicated about the tiffs, they are simply single band DSMs.  There is no overlapping data, so the mosaic parameters should not matter. 

For reference, I read in the tiffs natively using a csv exported from the mosaic footprints and a  featureReader, and converted these to LAS and the output is perfect.