Im trying to code up a function for finding the best fit affine transform between two curves i and j for checking the similarity between two curves. My main problem at the moment is getting the two curves to fit together as best as they can to make the check.
I have an approach that I want to see if it right.
- Run the curves curveA and curveB through a linear least-square formulation to get lines for each curve that best fits their set of points.
- perform an affine transformation to lineA to match it to lineB.
- Use the affine transformation of lineA on curveA.
Would this approach work? I'm pretty new to affine transformations and linear least-squares to be sure if this is the right approach.