una-dinosauria/human-motion-prediction

Camera intrinsics and Extrsinics to convert to image coordinates

alecda573 opened this issue · 1 comments

I have downloaded the euler angle format of h3.6m and have applied forward kinematics to get the joint locations in the cameras reference frame. Is there any code to convert to image coordinates or can you supply the camera intrinsics/extrinsics to perform this conversion?

Hi @alecda573,

The camera parameters are given on the H3.6M website, on a file called metadata.xml.

You can use

https://github.com/una-dinosauria/3d-pose-baseline/blob/666080d86a96666d499300719053cc8af7ef51c8/src/cameras.py#L151-L174

to parse the parameters.