We used the cv2.findFundamentalMat function to estimate the fundamental matrix from left-right image point correspondences. This function supports a few different algorithms for fundamental matrix parameters estimation, such as cv2.FM_7POINT (7-point algorithm), cv2.FM_8POINT (8-point algorithm), cv2.FM_LMEDS (least-median approach), and cv2.FM_RANSAC (RANSAC-based approach). Two optional parameters specify the error threshold for RANSAC-based estimation algorithms and confidence levels for the least-median and RANSAC-based approaches.
The following are the expected results:
Fundamental matrix: [[ 1.60938825e-08 -2.23906409e-06 -2.53850603e-04] [ 2.97226703e-06 -2.38236386e-07 -7.70276666e-02] [ -2.55190056e-04 7.69760820e-02 ...
