Monday, January 4, 2016

Normalize a Lidar LAS file with Fusion ClipData

There are occasions when it is necessary to normalise a LiDAR las point cloud file, i.e. removing the terrain heights. This can be useful if you are interested in the heights of objects e.g. vegetation canopy above the ground and not the undulations of the terrain below. An example of a las file point cloud is shown below; the shape of the terrain is very much evident.


The ClipData utility from the open source Fusion software can be used to normalise a LiDAR point cloud. The following steps illustrate how to normalise a point cloud.

  1. In a Command Prompt, type in the following to convert a digital terrain model DTM from ESRI Ascii ArcGrid to Fusion's DTM format, using the ASCII2DTM.exe utility.

    C:\> ascii2dtm.exe output.dtm m m 0 0 0 0 input_dem.asc

    An example run is shown below.

  2. At the prompt, type in the following.

    C:\> clipdata.exe /height /dtm:dakota_dem.dtm dakota.las output.las 606034 4889354 607400 4890580

    Note:
    /height tells the utility to normalise the point cloud
    /dtm:dakota_dem.dtm specifies the input digital terrain in Fusion's DTM format.
    dakota.las is the input las file
    606034 is the minimum bounding x value
    4889354 is the minimum bounding y value
    607400 is the maximum bounding x value
    4890580 is the maximum bounding y value


    An example run is shown below:


    A screenshot showing the normalised point cloud. Notice the point cloud no longer undulates with the terrain. 

No comments: