Is it possible to use the trajectory of the joint to calibrate the extrinsics of the cameras

[RustyShackleford73]

I am trying to use the calibrated cameras, and use the joint movement to find the fundamental matrix and extrinsics, but the result seems not as good as yours