LIDAR point data is typically attributed with a return number (e.g. return #3) and a total number of returns (e.g. 5 returns). This is standard point attribute information in a .LAS file. From this information the return combination can be easily determined. Return combinations typically used to improve processing filters include “First of Many”, “Last of Many”, “Intermediate (neither first or last)” and “Single” (or sometimes “Only”). Using this return combination information is important in writing effective macros for automatic classification routines. To assist the user, TerraScan has several classification routines (found under the “Classification” menu in the TerraScan Main Window, and all available either interactively or via the macro creation tools. The most common one is “By Echo …” (Note: TerraScan uses the term “echo” instead of “return”). This allows the user to move points of a certain return combination to a specific class. It can also be used to classify based on a specific return number (up to 7).
For example, to create a first return surface (a TIN generated from all first returns) the user can create a custom class, and move all first returns to that class, and then generate a surface model from just that class. This approach can be further refined by using different return combinations and classes to determine the specific surface model that is needed (for example all Water points could be excluded regardless of return combination; first pulse only returns could be used). The example below shows a data set classified by return combinations with blue being first and yellow being last returns.
Note that in TerraScan there are no separate viewing or processing flags based on return combination that can be applied on-the-fly. That is, you can’t set your display to just show you first returns, without first setting up a class to do so. This sometimes limits what can be done with classified data. For example, you can’t extract first return pulses from a ground class while maintaining that ground classification. Other software tools, such as LP360, do not have this limitation.