MACHINE LEARNING METHODS FOR ROAD EDGE DETECTION ON FUSED AIRBORNE HYPERSPECTRAL AND LIDAR DATA

In the last decades, remote sensing sensors, such as hyperspectral systems or LiDAR scanners, have been used for urban mapping. However, an analysis in the urban environment is very complex in applications, e.g., road detection, city management, and urban planning. One of the important urban features is the detection of the road edges. In this study, an approach on multisensory hyperspectral and LiDAR data fusion (HL-Fusion) is introduced for road edge detection using different machine learning algorithms, such as Support Vector Machines, Random Forests, and Convolutional Neural Networks. The first results show that the Random Forest algorithm outperformed in the experiments on the study area at Oslo's surroundings in Norway. This study opens a window for further investigation on machine learning algorithms and a better understanding of HL-Fusion capabilities.