LiDAR to Numpy/Pandas in ArcGIS Pro

[Anonymous User]

Data Scientists and GIS Users can now take advantage of LiDAR processing in Numpy & Pandas, two python libraries used for processing and making meanings from "Big Data".

The following script serves as an example of how LiDAR tiles can iteratively be read into Numpy arrays or Pandas DataFrames.

https://github.com/geoffrt/esri_lidar_3d_management/blob/master/scripts/las_to_numpy.py 

It is up to the user to determine what python machine learning libraries you want to feed the data into. 

ex: SciKit Learn

Once features have been detected data can be converted back to Points, lines or polygons in ArcGIS Pro using one of the following libraries:

• ArcPy: Numpy Array to Feature Class
• ArcGIS API for Python Spatially Enabled Dataframe

**Expect an update in the near future supporting reading from LAS Datasets also**

[DanPatterson]

the following script link is 404

[JoeBorgione]

@DanPatterson it worked for me and the link below it  where it sends me.

https://github.com/geoffrt/esri_lidar_3d_management/blob/master/scripts/las_to_numpy.py

[DanPatterson]

It was fixed, it directs to the same github page as the other link.

I am just too quick 😉

[JoeBorgione]

Does the SciKit and ArcPy: numpy array to feature class work for you ?  I get a 404 and  redirect to the Pro 2.7 respectively...

[Anonymous User]

Resolved. Seems like the url editor was adding trailing spaces... %20% one would assume a community platform software would have the logic to automatically remove such errors at the end of urls... Fun.

[JoeBorgione]

 ...assume a community platform software would...   😁

[Anonymous User]

Updates:

1.  Added operation to directly parse the lidar to a pandas dataframe instead of just numpy array.
2. Notated user parameters under __name__ == "__main__" 
3. Still need following update if anyone wants to take it on:
   • code for converting the 1D array formatting in the numpy array to a 2D dataframe. Without this it's impossible to plot the data etc....

[DanPatterson]

@Anonymous User do you have a small dataset? I do numpy and graphing extensively

[Anonymous User]

@DanPatterson 

Here is a sample las dataset that has been classified. Thinned evenly at 1m point spacing.

LiDAR Sample tile: https://esriis-my.sharepoint.com/:u:/g/personal/geof7015_esri_com/EfdIBFdKCVxLmr2BbOGfvMYBtr28RHYZ0x... 

The Projected Coordinate System WKID = 6549

The Vertical GCS WKID = 6318

Thanks in advance for taking a look at this.