SFM: Known intrinsic parameters and camera pose

[HelliceSaouli]

Hi
Is it possible for your sofware to create a bundler for a ground truth dataset such as Middlebury. where you have all information about the camera

[snavely]

I am not 100% sure what you would like to do -- do you mean that you just want a bundle.out file with the known Middlebury camera geometry, or do you want to run bundler but constrain the intrinsic or extrinsics in some way? Perhaps you could clarify by saying a but more about what you are trying to do. Noah

…

On Sun, Oct 29, 2017 at 9:31 AM, Saouli ***@***.***> wrote: Hi Is it possible for your sofware to create a bundler for a ground truth dataset such as Middlebury. where you have all information about the camera — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#57>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABt6q_7ojHFaXS_C36RoVsyh4u7e6hd2ks5sxH4hgaJpZM4QKUIW> .

[HelliceSaouli]

Thank you for your reply sir;
Well here is my issue : I'm working on depth map reconstruction algorithm similar to the one available in MVE lib (dmrecon) such algorithm neede sfm as input (bundle.out basically) and use the features as seeds. everything was fine until I submitted my results to Middelburry webpage for evaluation and they said that my reconstructed object is not in the scene coordinate system. it turned out that because I didn't use their ground truth camera parameters and I used the one of sfm my object is not in the correct coordinate system. for this, I need a structure from motion algorithm that fixes intrinsic and extrinsic (given) and output bundler.out which seems weird and counter-intuitive. but this what i need. a more detailed discussion with Mr Simmon on the subject is here

Thank you