LSD-SLAM is a novel approach to real-time monocular SLAM. It is fully direct (i.e. does not use keypoints / features) and creates large-scale, semi-dense maps in real-time on a laptop. For more information see http://vision.in.tum.de/lsdslam where you can also find the corresponding publications and Youtube videos, as well as some example-input datasets, and the generated output as rosbag or .ply point cloud.
We highly recommend to use docker to build the source, because it is very difficult to reproduce the execution environment on general Linux distributions such as Ubuntu etc.
$git clone https://github.com/IshitaTakeshi/lsd_slam_noros.git
$cd lsd_slam_noros
Run the following command in a directory where Dockerfile is placed.
$docker build -t lsdslam_noros:latest .
This part is usually environment dependent because we need to share a window with the host.
If you are on Mac, make sure the latest XQuartz is installed, and run $./scripts/launch-container-mac.sh.
On Linux, run $./scripts/launch-container-linux.sh.
Download the TUM dataset sequence by running
$export SEQUENCE=30 # sequence number of the TUM-mono dataset
$./scripts/download_tum_mono.sh
The debug window should appear by executing $./bin/main_on_images data/sequence_$SEQUENCE.
If segmentation fault or Gtk-WARNING happens, executing the command above multiple times may solve it.
-
LSD-SLAM: Large-Scale Direct Monocular SLAM, J. Engel, T. Schöps, D. Cremers, ECCV '14
-
Semi-Dense Visual Odometry for a Monocular Camera, J. Engel, J. Sturm, D. Cremers, ICCV '13
LSD-SLAM is licensed under the GNU General Public License Version 3 (GPLv3), see http://www.gnu.org/licenses/gpl.html.
For commercial purposes, the original lsd slam authors also offer a professional version under different licencing terms.