You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@tianf-code
Yes we solved it, and mesh extraction was done successfully.
The problem was from the output of colmap, it showed us that it was successfully done but actually the output from comlap alot of things were missed.
U can check undistorted images if there’s many images or not
To begin, I processed the video using Colmap with the following commands:
Convert the video to individual frames:
ffmpeg -i /project/gaussian-splatting/Museum.mp4 -vf fps=2 -vsync vfr -q:v 2 /project/gaussian-splatting/data/input/%06d.jpg
Convert frames to a suitable format:
python convert.py -s /project/gaussian-splatting/data
Next, I utilized Gaussian Splatting to generate the Gaussian representation of the scene:
Perform Gaussian Splatting:
python SuGaR/gaussian_splatting/train.py -s data/ --iterations 5000 -m gaussian_output/
Then, I executed SuGaR:
Run SuGaR with specified parameters:
python SuGaR/train.py -s data/ -c gaussian_output/ -r "density"
Please review this workflow and identify any potential issues that make the background and foreground meshes empty.
The text was updated successfully, but these errors were encountered: