Hungarian MLS point clouds of railroad environment and annotated ground truth data

Published: 4 April 2022| Version 1 | DOI: 10.17632/ccxpzhx9dj.1
Mate Cserep


These sample LiDAR datasets were collected by the Hungarian State Railways with a Riegl VMX-450 high density mobile mapping system (MMS) mounted on a railroad vehicle. The sensor was capable of recording 1.1 million points / sec with an average 3 dimensional range precision of 3 mm and a maximum threshold of 7 mm. Average positional accuracy was 3 cm with a maximum threshold of 5 cm. The acquired point clouds contain the georeferenced spatial information (3D coordinates) with intensity and RGB data attached to the points. The applied reference system is the Hungarian national spatial reference system, EPSG:23700. 3 datasets with different topographical regions of Hungary were selected: 1) mav_szabadszallas_csengod_665500_162600_665900_163200.laz is a curved rail track segment on flat terrain between the city of Szabadszállás and the town of Csengőd. The selected segment is ca. 600 m long, containing 51.8 million points. 2) mav_sztg_szh_439040_183444_440377_183863.laz is a curved rail track segment with varied terrain and slopes between the cities Szentgotthárd and Szombathely. The selected segment is ca. 1500 m long, containing 58.6 million points. 3) mav_szabadszallas_csengod_666285_159100_666436_159200.laz is a curved rail track segment on flat terrain between Szabadszállás and Csengőd, 100 m long, containing 7.3 million points. Manually annotated ground truth data for cable and rail track recognition is also attached.



Eotvos Lorand Tudomanyegyetem Informatikai Kar


Lidar Remote Sensing, Railway, Point Cloud