1

I'm trying to determine the ratio of building to vegetation change (i.e. how has the volume of built structures changed in proportion to the volume of vegetation) in a city over a period of five years. The LAS files for the first time period only have ground, noise, and unassigned class codes; the second period has building, ground, noise, rail, water, and unassigned.

My main difficulty has been creating a building raster for the first time period, which does not have class codes for buildings. Secondly, I need to determine the amount of vegetation in the region for both time periods.

I thought it would be a pretty simple process (perhaps it is), but it has proven rather challenging as I am somewhat new to GIS and Arcmap (10.6). This is for a LiDAR class and the area of interest is only a few sq. km.

Does anyone have any suggestions for how to proceed?

  • There are some processes that will approximate buildings/vegetation but the result needs to be manually finished, nearly half the price of classified LiDAR is in the operator time needed for classification. ArcGIS is not good at this, ENVI may be better but in the end you will invest a lot of time into building classification manual editing. How large is your area of interest? You might find it quicker to heads-up digitize the buildings then use them as a mask for height calculations. – Michael Stimson Dec 4 '18 at 6:31
  • Hey, thanks for the response. I'm afraid I'm stuck using ArcMap - this is for a LiDAR class. The area of interest is only a few sq. km. I'll look into your suggested methods. – user132986 Dec 4 '18 at 6:35
  • There are some tools like pro.arcgis.com/en/pro-app/tool-reference/3d-analyst/… but the editing interface is horribly slow.. if you can tweak the tool and get something that's sort-of right it might be good enough to make a start on. You can export a raster from building only class, polygonize, delete the really small polygons that you'll get for bins, street furniture, cars etc.. by minimum size. Have a go and see if it's near enough for what you want. 3d analyst license required. – Michael Stimson Dec 4 '18 at 7:30
1

Building

As suggested by Michael Stimson use the Classify LAS Building tool (3D Analyst required):

Classifies building rooftop points in aerial lidar data.

To have an idea about how precise and accurate the tool is (before running it with the first dataset, i.e. the unclassified one), run it with the second dataset (the one which is already classified) and compare results (ArcGIS versus vendor classification). First, create a copy of this second dataset and alter the code from building points to unassigned, because the Classify LAS Building tool takes into account building points if they exist in the data:

... Existing building classified points with a class code value of 6 will also be used to determine if adjacent points fit along a plane connecting the same structure ....

Make sure you can consider the classified dataset (the one classified by the vendor) as a benchmark, so the comparison is valid.

Vegetation

Vegetation is more tricky, it depends on what you consider to be vegetation, for example a continuous fragment, a small cluster of trees (like the ones in a urban park), any tree? Some options are then, to manually digitize the point cloud since it is a small area you are working with or try identifying continuous vegetation areas (with auxiliary data) and classify them as such (see for example, the Set LAS Class Codes Using Features tool).


Also, take a look at related posts:

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.