A new method for extracting trees and buildings from sparse LiDAR data in urban areas

In this letter, we propose a new method for extracting trees and buildings from sparse Light Detection and Ranging (LiDAR) point clouds. We use two features, height and normal variation, to classify LiDAR points into buildings and trees. The accuracy obtained from an urban area in New Orleans, Louisiana, is 92% for both buildings and trees. The results indicate that three-dimensional objects such as buildings and trees can be effectively classified using the new method. The results also show that the new method is very effective and suitable for sparse LiDAR data in urban areas.