workflow.txt

## Section 4.1: Solve for camera intrinsic matrix

# Get world coordinates of checkerboard corners # These will remain fixed throughtout all the images
# Get image coordinates of corner for all images. # Use cv2.findChessboardCorners
# Get Homography between world and image coordinates
# Define Vij and b to setup the system of homogenous linear equations: Vb=0
# Use eigenvector of V_t*V corresponding to smallest eigenvalue as b
# Get A matrix from b vector # Use Appendix


## Section 4.2: Estimate approximate camera extrinsics

# Get R, t from camera intrinsics and homographies # Use section 3.1


## Section 5: Non-linear Geometric Error Minimization

# Use minimize function from the scipy.optimize module
# Define objective function as the product of cam intrisincs, extrinsics, world coords
# Use objective function in the scipy.optimize module to reduce the reprojection error