TY - JOUR
T1 - Full-Waveform Airborne LiDAR Data Classification Using Convolutional Neural Networks
AU - Zorzi, Stefano
AU - Maset, Eleonora
AU - Fusiello, Andrea
AU - Crosilla, Fabio
PY - 2019
Y1 - 2019
N2 - Point cloud classification is one of the most important and time-consuming stages of airborne LiDAR (Light Detection and Ranging) data processing, playing a key role in the generation of cartographic products. This paper describes an innovative algorithm to perform LiDAR point-cloud classification, which relies on Convolutional Neural Networks (CNNs) and takes advantage of full-waveform data registered by modern laser scanners. The proposed method consists of two steps. First, a simple CNN is used to preprocess each waveform, providing a compact representation of the data. By exploiting the coordinates of the points associated with the waveforms, output vectors generated by the first CNN are then mapped into an image that is subsequently segmented by a Fully Convolutional Network (FCN): a label is assigned to each pixel and, consequently, to the point falling in the pixel. In this way, spatial positions and geometrical relationships between neighboring data are taken into account. These particular architectures allow to accurately identify even challenging classes such as power line and transmission tower.
AB - Point cloud classification is one of the most important and time-consuming stages of airborne LiDAR (Light Detection and Ranging) data processing, playing a key role in the generation of cartographic products. This paper describes an innovative algorithm to perform LiDAR point-cloud classification, which relies on Convolutional Neural Networks (CNNs) and takes advantage of full-waveform data registered by modern laser scanners. The proposed method consists of two steps. First, a simple CNN is used to preprocess each waveform, providing a compact representation of the data. By exploiting the coordinates of the points associated with the waveforms, output vectors generated by the first CNN are then mapped into an image that is subsequently segmented by a Fully Convolutional Network (FCN): a label is assigned to each pixel and, consequently, to the point falling in the pixel. In this way, spatial positions and geometrical relationships between neighboring data are taken into account. These particular architectures allow to accurately identify even challenging classes such as power line and transmission tower.
U2 - 10.1109/TGRS.2019.2919472
DO - 10.1109/TGRS.2019.2919472
M3 - Article
SN - 0196-2892
VL - 57
SP - 8255
EP - 8261
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
IS - 10
ER -