pass camera location as input for pose estimation

[hcwinsemius]

Currently pose estimation can be done with only control points, or control points combined with camera matrix and distortion coefficient. Extend so that also the following cases are possible:

• control points with camera location (optimizes intrinsics and rvec)
• control points with dist coeffs (optimizes focal length, tvec and rvec)
• control points with camera location and dist coeffs (optimizes rvec and focal length)
• control points with camera location and all intrinsic parameters (optimizes rvec)

How to disclose in CameraConfig:

• use CameraConfig.calibrate, create a cv.calibrate function, and pass all info (rvec, tvec, dist_coeffs, camera_matrix), also if None.
• Deprecate CameraConfig.optimize_intrinsic
• cv.calibrate, create cases with decision model which case applies, with required inputs and per case, what required amount of input points is. There should be 5 cases.
• create a function per case (+unit test)