facebookresearch / consistent_depth

We estimate dense, flicker-free, geometrically consistent depth from monocular video, for example hand-held cell phone video.
MIT License
1.62k stars 235 forks source link

for given video with distortion, do you estimate the distortion parameters as well as K(intrinsics)? #17

Open shuluoshu opened 3 years ago

shuluoshu commented 3 years ago

Hi @likethesky @roxanneluo @Celebio I wonder given a video with unknown intrinsics and heavy distortion. Can the ColorMap preprocess estimate the distortion params and do undistortion for further process?

mattbev commented 2 years ago

@shuluoshu I imagine if you modify consistent_depth/utils/load_colmap.py you may be able to get to your desired solution