Perform RANSAC to find the best affine transformation that maps the most points from image 2 to image 1. To do this, perform the following steps: a. Randomly pick three matching points, in the format computed in part 2.

b. Compute the affine transformation, A, that relates them, using part 3.

c. Apply A to all points in image 2.

d. For each point, p, in image 2, that has been matched to point q in image 1, find out whether Ap is close to q (close might mean their Euclidean distance is less than 2).

e. Count the number of points that are mapped by A to be close to their matching point.

f. Repeat steps a-e many times (maybe a thousand?) and choose the A with the highest total.

% This function finds the affine transform that best relates the input % points, p1 and p2 % In: p1, input points % In: p2, input points % Out: A, the 2x3 transformation matrix function [ A ] = affine_transformation( p1, p2 )

% First, create an augmented matrix p1_augmented = [p1; ones(1,3)];

% Check determinant of matrix before inverting it if abs(det(p1_augmented))