Output: Intrinsic and extrinsic parameters of input cameras
Notes:
Metrics need to be tuned according to usage.
Order of cameras in output need to be paid attention to by verifying the name of input images.
Approach 1 Agisoft Metashape
Download compatible version of Agisoft software (now is called Agisoft Metashape). 30 days free trial will be granted.
Create a chunk or use the default one.
Add input images into the same chunk.
Click Align images in tool tab.
Under File, export camera parameters.
Open the generated XML file. Intrinsic parameters can be easily found under cameras. Extrinsic parameters are written in one line called transformation at the second half of XML file.
For more reference, you can read the official user manual from Agisoft.
Approach 2 Matlab Camera Calibration Tool
Download and install the basic Matlab package.
Under Tools, open camera calibration.
Add input images from different cameras.
Measure and enter the size of square on the checkerboard used in images in the real world.
Select distortion factors from the top bar. Typically, frame cameras use default; fisheye cameras need distortion factors in more dimensions.
Drag outlier factor in histogram in lower right corner.
Export camera parameter to main Matlab window by clicking export on the top bar. Name of the variable depends on your choice.
Feel free to close the camera calibration window now or just minimize and go back to the main window.
Enter Variable_Name.Intrinsics to print out intrinsic parameters of camera. Because Matlab assumes input cameras are the same, there is only one set of intrinsic parameters generated.
Similarly, enter Variable_Name.Translations to print out translation matrices for cameras, respectively. Enter Variable_Name.Rotations to print out rotation matrices. You need to edit them in the same way for default 3x4 extrinsic matrix. See documentation.