Camera Calibration and Fundamental Matrix Estimation with RANSAC

Computer Vision | Matlab

This project involves computing the projection matrix to find the camera calibration and estimating the fundamental matrix. Projection matrix helps in mapping 3D world coordinates to 2D image coordinates. From the given set of 2D and 3D points, a homogeneous set of equations is set up. Projection matrix is a 3×4 matrix and to solve it, least squares method is used. Fundamental Matrix helps to estimate the mapping of points in one scene to the epipolar lines in the second scene.

Epipolar lines for two images with different angles of view:

capture

 

 

Finding corresponding points in images with different angles of view using Epipolar lines:

capture3

 

Links: Project Report