Fundamental Matrix Computation for Stereo Vision Cameras

This algorithm was developed in MATLAB 2008b environment specifically for computing the fundamental matrix in stereo vision camera systems. The implementation consists of four core functions: main, sevenPoint, eightpoint, and eightpoint_norm. The main.m function serves as the program entry point, responsible for generating and loading experimental data, sequentially calling the three fundamental matrix computation methods, and finally plotting the results. The sevenPoint.m function implements the 7-point algorithm which uses a minimal set of 7 point correspondences to compute the fundamental matrix through singular value decomposition (SVD). The eightpoint.m function employs the standard 8-point algorithm that utilizes 8 or more point matches to solve the fundamental matrix using linear least squares. The eightpoint_norm.m function enhances the standard 8-point approach by first normalizing the input data points to improve numerical stability and accuracy. During development, each function underwent rigorous debugging and optimization to ensure computational precision. Extensive testing and validation were performed to verify the algorithm's ability to accurately compute the fundamental matrix, thereby providing reliable support for stereo vision applications.