Automatic Estimation of Camera Fundamental Matrix
• Implemented corner detection and used normalized correlation coefficient for feature extraction and matching with stereo images.
• Implemented the 7-point algorithm with Sampson Error Correction in MSAC for outlier rejection. Implemented the 8-point algorithm in Direct Linear Transformation to estimate the camera’s Fundamental matrix.
• Implemented Levenberg-Marquardt Algorithm to minimize the reprojection error with Triangulation and Optimal Correction.
Original images
Corner detection
Brute-force feature matching
MSAC outlier rejection
Levenberg-Marquardt algorithm optimization
Final result
