Background:
Our organisation procures a number of photogrammetric meshes from external vendors. The meshes we have are in SLPK format primarily and often rather large in size (100GB+) with irregular boundaries. Out of necessity, one contiguous capture may occasionally be received in multiple SLPK files of a jigsaw fit.
To help our users browse our catalogue of data, spatial extent metadata in the form of 2D footprints is essential. We request 2D footprint shapefiles from vendors, but sometimes things fall through the cracks, or the footprints don't reflect a bisected dataset.
Question:
Is there a workflow to produce 2D footprints from an SLPK?
I've tried a number of approaches, including smashing an SLPK file open in 7-zip and poking around at the lowest levels of the pyramid structure, but to no avail. I can see a potential workflow using the SLPK To LAS tool in the latest ArcGIS then rinsing it through a TIN, flattening, dissolving, etc, but that seems excessively computationally complex. I mean, when I load an SLPK into ArcGIS, the software itself seems to be able to figure out the extents for the sake of highlighting it on screen pretty effectively, so I imagine an extent must be stored somewhere.
Any workflows or tips?
Thank you.
Hi,
There is no geoprocessing tool that will create footprints from a scene layer package. There is a fullExtent property in the slpk that clients read to set initial extents. You can see this documented in the specification here https://github.com/Esri/i3s-spec/blob/master/docs/1.9/3DSceneLayer.cmn.md
The only other option I can think of is to use a Catalog layer as there is a footprint layer available as part of that. https://pro.arcgis.com/en/pro-app/3.3/help/mapping/layer-properties/catalog-layers.htm
thanks,
Andrew
Hi Andrew,
Thanks for your response.I did try out the catalog layer workaround, but unfortunately to no avail. By default, SLPKs (or at least integrated mesh profile ones) simply cannot be selected as a valid addition to a catalog dataset, despite 'SLPK' being one of the options in the Input Item Types dropdown menu. I eventually found a workaround to the input validation error by simply pointing to the containing directory and selecting SLPK as the type.
The result was not what I expected though. I just got a bounding rectangle of the dataset (often quite heavily buffered), aligned with the cardinal axes. As best I can tell, this is simply a rectangle drawn from the bbox values in the top level ASCII files if you smashed the SLPK open in something like 7-Zip. Selecting 'Envelope' or 'Convex Hull' Footprint Type didn't seem to impact the output.
Best regards,
Paul