how to undertstand calculating rays for image
shaoxiang777 opened this issue · comments
Hi, thank you very much for your great work first.
I want to kindly ask how to understand get_rays function in commen.py ? Is there any theoretical support for it? No matter the math foundation or geometry theory about camera calibration. Actually I don't understand "rays" in depth camera. I would appreciate it if you could explain how it works!
I am looking forward to your reply. Thanks a lot in advance!
Hi @shaoxiang777, our system is nerf-based, so we follow how NeRF renders a pixel. For the get_rays we also simply adapted from the original NeRF repo, see here. The "rays" means that the vector from the camera origin to the sampled pixel. To really understand what the get_rays is doing, maybe you can first get yourself familiar with the original NeRF paper.