search

jzhangbs / Vis-MVSNet

Visibility-aware Multi-view Stereo Network
MIT License
235 stars 27 forks source link

About the camera intrinsics in Tank and Template #37

Closed zhao-yiqun closed 1 year ago

zhao-yiqun commented 1 year ago

Hi

Thanks for your work and code releasing.

I checked your preprocessed tank and template dataset with known cameras like follows:

extrinsic 0.999749 0.0222156 -0.00306238 0.0954387 -0.0223105 0.999119 -0.0355459 -0.0991731 0.00227 0.0356053 0.999363 1.04493 0.0 0.0 0.0 1.0

intrinsic 1162.86 0 963.831 0 1162.86 541.647 0 0 1

0.367867 0.00391897 730 3.22871

I understand the extrinsic and intrisic means. But I wonder what's the meaning of the last lines. I checked the code, However still don't know what does it mean.

Thanks.

jzhangbs commented 1 year ago

Hi @zhao-yiqun, they describe the depth range for this view. They are depth start, depth interval, depth numbers, depth end respectively.

Note that we will adjust the depth number by cmd args. We will keep the start and the end, and set the depth interval accordingly. And in coarse-to-fine structure, the actual depth number and interval are scaled based on this default depth grids.

zhao-yiqun commented 1 year ago

Thanks for your kindly reply. Now I understand it.

Bests.