About cameraintrinsicsextrinsics.pickle file.

[Jordi Pozo]

Hello,

we are currently testing to be able to process locally the experiments using 2 ipads to capture the videos. Starting from the example code that you offer in github, we have managed to create the processed video files with the graphical information related to the joints (colored skeleton). Now, the question is if the "cameraintrinsicsextrinsics.pickle" files must be generated previously to the execution of this process and, if so, when and how they are generated. Is there any of the code scripts that you have published that can be reviewed and adapted or just use one of the .pickle files corresponding to the iPad model that you have published in github, inside opencap-core/CameraIntrinsics?

Thank you very much in advance.

Best regards,

Jordi

[Antoine Falisse]

Hi,

The camera intrinsics are specific to your hardware and to the settings you use when taking the video. We fix them in OpenCap, which allows us to load them from the database. If you use videos not recorded using OpenCap, you will need to compute them somehow. For reference, we use this code to do it https://github.com/stanfordnmbl/opencap ... rinsics.py using videos from a checkerboard. This goes slightly beyond support we can offer, since it is not using the OpenCap system per se. For the extrinsics, you could use our code https://github.com/stanfordnmbl/opencap ... in.py#L163

Hope it helps,
Antoine