 |
|
Velodyne SLAM - Dataset |
|
 |
Here, you can find two challenging datasets recorded with the Velodyne HDL64E-S2 scanner in the city of Karlsruhe, Germany.
 |
 |
| scenario 1 |
scenario 2 |
Dataset
The following zip files include everything you need (including a description of the data format).
Tools for viewing / converting the data can be found below.
This data can be freely used with one restriction: In case it is used for scientific publication
you are required to cite the article "Velodyne SLAM" of Moosmann et al. (see here).
Results
The 3D maps and trajectories calculated by the proposed algorithm (without loop closure!):
Data-Viewer
In theory, the recorded png-files can be viewed by any imaging software.
In practice, there's not a lot to see due to low contrast.
This is why we provide a tiny image viewer in C++/Qt:
Source Code
To ease accessing the data, here are some C++ classes:
- PngDistanceImage (.zip) (.tar.gz), a class for loading/saving distance data from/to png-files.
- ImageProjector (.zip) (.tar.gz), a class for converting distance images into 3D point clouds and back.
If you write any converters (e.g. to P3D files), I'd be happy if you send me the code to publish on this website.
Contact
If you have any questions or comments, feel free to contact me
http://www.mrt.kit.edu/mitarbeiter_moosmann.php