[pdal] Point clound Normalization

[Howard Butler]

On Jun 27, 2013, at 10:07 AM, "Newcomb, Doug" <doug_newcomb at fws.gov> wrote:

> Hi folks,
> I was wondering if there were any plans to add any utilities for normalizing the z values to the value of either an external raster ( via gdal) or from an internally created TIN of the bare earth points.  ( I'm guessing Paul has some thoughts about doing this inside Postgres/Postgis)

You could do this with a Python filters.programmable filter. Additionally, PDAL does have the filters.colorization, which is really just a specialized "overlay these points on a GDAL raster and fetch the RGB values" operation. You could set it to fetch a single value into the Z dimension from the raster, but not do any math on it. The programmable filter would allow you to do whatever you want in Python land, and then stuff the value back into the Z dimension.

https://github.com/PDAL/PDAL/blob/master/test/data/plang/predicate-keep-specified-returns.xml is an example that creates a "keep" mask for predicate-type operations. In combination with a programmable example like https://github.com/PDAL/PDAL/blob/master/test/unit/plang/ProgrammableFilterTest.cpp you might find enough inspiration to combine it with numpy-fetching GDAL operations.

Hope this helps,

Howard