Home

Documentation

Also see the other pages documenting the individual features.

Preparing the dataset

When loading a dataset, the data must be organized as follows:

point cloud folder
├── velodyne/             -- directory containing ".bin" files with Velodyne point clouds.   
├── labels/   [optional]  -- label directory, will be generated if not present.  
├── image_2/  [optional]  -- directory containing ".png" files from the color camera.
├── image_3/  [optional]  -- directory containing ".png" files from the other color camera (stereo).
├── calib.txt [optional]  -- calibration of velodyne vs. camera. needed for projection of point cloud into camera.  
└── poses.txt             -- file containing the poses of every scan, see below on how to obtain them.

To obtain the poses.txt go to http://jbehley.github.io/projects/surfel_mapping/ and download the "Frame-to-Model (with loop closure)" poses. The rss2018_frame2model_loop folder contains text files of the form 00.txt, 01.txt, 02.txt, ..., choose the correct one for your sequence, move it to your point cloud folder and rename it to poses.txt.

Navigation in the Point Cloud

There are three colored vectors indicating the orientation of the coordinate system:

• red vector for the x-direction
• green vector for the y-direction
• blue vector for the z-direction

In the console output you will see the coordinates of the currently selected tile. For example, 200, -100, 50 denotes the tile at x = 200 (red direction), y = -100 (green direction), z = 50

Moving around

• Left Mouse Button + Drag: Move in the x-y-direction (stay in z=constant plane)
• Middle Mouse Button + Drag: Move in the y-z-direction (stay in x=constant plane)
• Right Mouse Button + Drag: Change viewing angle of the camera