Algorithm 1 Layer registration utilizing multiple-layer RANSAC

Input: Initial pair set ${\mathcal{F}}_1=\mathcal{F}$, threshold $N_{min}$ and iteration index $k=0$;

Output: Each layer’s matching pair set ${\mathcal{L}}_k$ and its corresponding homography ${\mathbf{H}}_k$;

repeat

$k\leftarrow k+1$

RANSAC in pair set ${\mathcal{F}}_k$ for model ${\mathbf{p}}_{k_n}^1\times \mathbf{H}{\mathbf{p}}_{k_n}^2=0$, where $({\mathbf{p}}_{k_n}^1,{\mathbf{p}}_{k_n}^2)\in {\mathcal{F}}_k$;

Divide outliers ${\mathbf{V}}_{out}$ and inliers ${\mathbf{V}}_{in}$ according to $\mathbf{H}$;

if $|{\mathbf{V}}_{in}|\ge N_{min}$ then

Set matching pair set of the k-th layer as ${\mathcal{L}}_k={\mathbf{V}}_{in}$;

Set homography of the k-th layer as ${\mathbf{H}}_k=\mathbf{H}$;

end if

Set the pair set of next iteration as ${\mathcal{F}}_{k+1}={\mathbf{V}}_{out}$;

until $|{\mathcal{F}}_{k+1}|<N_{min}$