Fix Shadows

When the openni_grabber demo is run with any VLP-16 data set, two (or more) copies of all the points appear above/below the first set of points. The points are not vertical translations but appear to be angular translations from the LiDAR origin.

It seems that the openni_grabber demo is designed for HDL-32 data sets as described in The Velodyne High Definition LiDAR HDL Grabber page by PCL.

All the data sets I’ve collected with the Velodyne VLP-16 have these shadows as do the sample data sets Leo has left behind.
I ran some HDL-32 sample data sets and confirmed no occurence of these shadows, so it seems that the program is calibrated for the HDL-32 and not the VLP-16.

There doesn’t seem to be a calibration file for VLP-16 but PCL does actually have a vlp_viewer.cpp which makes use of vlp_grabber.cpp in the library.
I have found the VLP-16 and HDL-32 calibration files thanks to Tsukasa Sugiura’s comment on his video of Drawing VLP-16 Data using PCL.

vlp_viewer

visualization/tools/vlp_viewer.cpp

vlp_grabber

io/src/vlp_grabber.cpp

Next Step:

• Compile these files. [DONE]
• If too hasslesome, try to make the HDL Grabber compatible with VLP-16. [DONE]
• I am still curious about the other programs, but will leave compiling those as secondary tasks. [DONE]

After troubleshooting the vlp_viewer from PCL, I have realized that compiling this program outside of the PCL repository leads to many build issues. Henceforth I will just make a branch in the PCL repository and make adaptations to the programs and run the programs from the pcl-trunk/x86_64/bin folder.