This model kindly provided by our Partner T-Kartor
The 2D View on the left shows a typical point cloud product (in LAS format) produced in many countries. Some of the points have had a basic classification assigned to them (such as Ground and Low Vegetation), but many of the points remain Unclassified.
The classified points have already had work performed to make those class assignments, so we do not wish to alter them. However we do want to take the Unclassified points and apply a classification to them. The model shown below performs that function.
Certain points (generally the points with an assignment of Unclassified) are selected, just those points are then taken and run through the Point Cloud Classification process and then those newly classified points are recombined with just the points from the original LAS file that had existing classifications. This produces the point cloud shown in the 2D View on the right above.
There was initially some concern that by only classifying a limited set of points you would not achieve good results. However tests by the original author produced acceptable results using their data. So please take the model in the spirit it is intended - as an example of using some of the Point Cloud operators available in Spatial Modeler.
Input Pointcloud: Name of the input pointcloud file which has only partial classification of points.
Range of classes to be classified: Enter an integer value to determine the range of classes that are selected for new classification. Classes 0 and 1 typically represent new points without class and previously unclassiified points. So a good default value is 1 but it can be any other range valid for .las file classes. enter a value of 1 would define a range of classes to be processed of 0 to 1.
Output pointcloud: Name of the output pointcloud file where the selected Classes have been re-classified.