Contextual Classification of Full Waveform Lidar Data in the Wadden Sea

A novel approach combining conditional random fields with random forest algorithms has demonstrated significant potential for accurately classifying airborne lidar data, particularly in challenging coastal environments like the Wadden Sea. This method leverages both geometric and full waveform information to distinguish between different land cover types, even in flat areas where traditional classification techniques struggle. The study highlights the importance of contextual analysis and the use of a high number of classification features, showing promise for improving scientific monitoring of coastal morphology and habitats. By investigating the relevance of various features for lidar points and their interactions with neighboring points, this research offers a robust framework for future applications in remote sensing.

Schmidt, A., Niemeyer, J., Rottensteiner, F., & Soergel, U. (2014). Contextual Classification of Full Waveform Lidar Data in the Wadden Sea. IEEE Geoscience and Remote Sensing Letters, 11(9), 1614–1618. https://doi.org/10.1109/lgrs.2014.2302317