FP7 IQmulus enabled efficient extraction of tree parameters in huge mobile mapping point clouds.
Car based laser mobile mapping has proven itself as an efficient and accurate sensor system to sample our 3D road environment. At an operational basis, tens to hundreds of km of road is sampled on a daily basis. A challenge however is to correctly extract geometric parameters of objects in the resulting point clouds with the same efficiency. The FP7 IQmulus project coupled a cloud computing facility with novel geometry extraction algorithms and state of the art visualization techniques. One of the project showcases considers the problem of identifying all individual trees in mobile mapping point clouds. For this purpose several alternative algorithms have been implemented. The results in terms of quality, robustness and computational efficiency will be demonstrated on examples consisting of 10 to 50 km of road environment at different locations in France and Germany.