[pdal] Is PDAL capable of producing Lidar grid metrics?

[Howard Butler]

As I mentioned in the Stack Exchange thread, and Brad mentioned below, this isn't currently possible out of the box with PDAL. I would ask what you are hoping to get by having this capability in PDAL rather than building a tool chain that includes FUSION + PDAL + other tools. Maybe it would be helpful to take a step back and describe how PDAL statistics computation would fit within a larger context.

Most of PDAL's current "characterization" filters are about generating primitive computations for algorithms such as slope, ground, etc. There's things like coplanarity, projection, regression, sampling, and normal computation [1]. These are "per-point" type statistics rather than summaries, however, and are focused on the positional geometry of the point in relation(s) to its neighbors. Brad described the derivative writer, which does some typical surface summaries, and it would be a good place to start on an implementation if you were sufficiently motivated.

PDAL's statistics filter [2] was recently updated to add kurtosis, skewness, variance, etc (no percentiles). You *might* be able to use that in combination with some sort moving window bounding geometry along with the crop filter [3], but this would be quite laborious and probably quite slow. Rémi's pgpointcloud approach seems viable and might be useful for other reasons, depending on what you want to do.

So, plenty of options but nothing that just does it.

Howard

[1] http://www.pdal.io/stages/filters.html
[2] http://www.pdal.io/stages/filters.stats.html
[3] http://www.pdal.io/stages/filters.crop.html

> On Sep 1, 2016, at 8:37 AM, Bradley Chambers <brad.chambers at gmail.com> wrote:
> 
> Aaron,
> 
> This is absolutely possible; it just doesn't exist out of the box today. We do something similar in our DerivativeWriter today (https://github.com/PDAL/PDAL/blob/master/io/derivative/DerivativeWriter.cpp). That may be a good place to look for inspiration. I think in the end you are looking for a GridMetricsWriter though. Are you needing a full replacement of the FUSION function? Or just a few select metrics?
> 
> Brad
> 
> On Thu, Sep 1, 2016 at 9:22 AM Aaron Poznanovic <aaron at redfoxgis.com> wrote:
> Greetings,
> 
> This is a follow-up to a question posted over at GIS Stack Exchange (http://gis.stackexchange.com/questions/175454/are-pdal-or-pcl-capable-of-producing-lidar-grid-metrics). I'm curious what functionality or approach one would take with PDAL to calculate Lidar grid metrics?
> 
> FUSION is currently the best available set of tools that I know of for calculating Lidar grid metrics, which are useful for describing forest structure (http://forsys.cfr.washington.edu/fusion/fusionlatest.html). From the FUSION documentation (page 70 from the documentation http://forsys.cfr.washington.edu/fusion/FUSION_manual.pdf):
> 
> GridMetrics computes a series of descriptive statistics for a LIDAR data set. Output is a raster (grid) represented in database form with each record corresponding to a single grid cell.
> 
> Essentially, I'm interested in producing a raster dataset where each cell represents summarized descriptive statistics for the Lidar point cloud such as cover, skewness, 75th percentile, etc (all of the grid metrics I'm after are available on the FUSION documentation. Is this currently possible, or recommended, with PDAL, or would this require other libraries? Many thanks in advance.
> 
> Best Regards,
> 
> -Aaron
> 
> Currently
> 
> 
> Aaron J. Poznanovic, GISP
> Principal Consultant
> Red Fox GIS & Remote Sensing
> (218) 387-4050
> www.redfoxgis.com
> 
> 
> _______________________________________________
> pdal mailing list
> pdal at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pdal
> _______________________________________________
> pdal mailing list
> pdal at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pdal

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.osgeo.org/pipermail/pdal/attachments/20160901/59b8356a/attachment-0001.sig>