Skip to main content
. Author manuscript; available in PMC: 2010 Jul 8.
Published in final edited form as: Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2008 Jun 23;2008:1–8. doi: 10.1109/CVPR.2008.4587808

Algorithm 1.

Our convex quadratic curve fitting method.

Input: learned patch experts, source image (Y),
Jacobian matrix (V)
initial warp guess (p),
index to the template (z), threshold (ε)
Output: final warp (p)

  1. Warp the source image Y with the current similarity transform from p.

  2. Compute the local responses E based on the learned patch experts and the source image Y.

  3. Estimate the convex quadratic curve fitting parameters Ak, bk and ck from Equation 14 for each patch.

  4. Estimate the warp update Δp using Equation 11.

  5. Update the warp z′ = Inline graphic(z; p) using Inline graphic(z; p) ← Inline graphic (z; p) ○ Inline graphic (z; Δp).

  6. Repeat steps 1–5 until ||Δp|| <= ε or max iterations reached.