Hi, Thank you for the great dataset! I'm trying to verify I'm doing perspective projection in a pixel-perfect way and for that I want to make sure that if I take LiDAR points and project them to the image this would result in identical depth values. Here is a self-contained script (https://gist.github.com/eldar/902f85e64569d15a4b04925b693a56b8 - relevant code in verify_frame() function) that attempts to do that, and the depth values seem to be close, but they still don't match. I was not able to find documentation on how the camera setup is for this dataset, so my question is - how can I can relate LiDAR point clouds and camera images/depths?
Hi, Thank you for the great dataset! I'm trying to verify I'm doing perspective projection in a pixel-perfect way and for that I want to make sure that if I take LiDAR points and project them to the image this would result in identical depth values. Here is a self-contained script (https://gist.github.com/eldar/902f85e64569d15a4b04925b693a56b8 - relevant code in
verify_frame()function) that attempts to do that, and the depth values seem to be close, but they still don't match. I was not able to find documentation on how the camera setup is for this dataset, so my question is - how can I can relate LiDAR point clouds and camera images/depths?