Max Likelihood Classification of UAV LiDAR

cmaloch

Trying to migrate over to Max Likelihood classifications from Grid method for UAV LiDAR Data, approximately 26 pts/sqm density. Anybody else having trouble getting good classification results? I assumed max likelihood would give better results based on newer algorithms, but so far I haven't seen it.

BMG_Amanda

Don't forget to adjust your resolution, the defaults are likely too small for this density. For example, when classifying poles, you want the resolution to be large enough to see that the edges of the pole are curved (0.5ft? depending on pole diameter). Too small, and the sections of pole will just look like a flat surface and will not be properly classified. Too large, and individual features might not be detected.

Also, feel free to kick a section of your data over to technical support to see what settings they recommend.

GeoHelp@BlueMarbleGeo.com