JiayuYANG / Self-supervised-CVP-MVSNet

Self-supervised Learning of Depth Inference for Multi-view Stereo (CVPR 2021)
86 stars 12 forks source link

How to render the triangle mesh to depthmap in different views? #13

Closed YANG-SOBER closed 3 months ago

YANG-SOBER commented 3 months ago

Hi, thank you for your nice work and answers.

May I know why the depth maps rendered remain black?

I use the ground-truth point cloud (scan114 and scan118) to render the depth maps and adopt 49 ground-truth camera parameters.

Could you please help me to identify the problem ?

Thank you sincerely.

YANG-SOBER commented 3 months ago

Thank you for your nice work.

I think there is a typo in the code pseudo_render.py, line 107, where the keyword argument znear should be znear=znear, instead of znear=0.

After this, the depth map can be successfully rendered.

Once again, thank you for your nice work and contributions.

Sincerely