From what I can gather, you are trying to estimate the relative "R" and "T" of a stereo camera system consisting of a fisheye camera and a basic lens camera.
I believe that to estimate the relative R and T of this stereo camera system, "cameraParameters" object's "estimateFisheyeParameters" and "estimateCameraParameters" functions can be leveraged respectively. See:
- https://www.mathworks.com/help/vision/ref/cameraparameters.html
- https://www.mathworks.com/help/vision/ref/estimatefisheyeparameters.html#d126e356605
- https://www.mathworks.com/help/vision/ref/estimatecameraparameters.html
With individual camera intrinsics in hand, their relative "R" and "T" can be determined using a PnP-like (perspective-n-point) approach.
This is achieved by capturing images of a common calibration object (e.g., a checkerboard) visible in both cameras and detecting corresponding 2D image points.
The "estworldpose" function, implementing the perspective-three-point (P3P) algorithm, can then solve the PnP problem. For more information, please refer to the following:
I hope this helps, thanks!