Algorithm 3 Connected Domain Filtering Method
Part I	1:	Give a gray binary image $G_d$ or a gradient binary image $G_{\mathrm{w}}$.
	2:	Process and get $G_d$’s anti-color image $G_{dc}$.
	3:	Get the connected domain set $Q$ and the number k of the connected domain.
	4:	fori = 1, …, k do
	5:	Get the area $S_i$ of the connected domain $Q_i\in Q$, the minimum bounding rectangle $T_i$ (with length $H_i$ and width $W_i$).
	6:	Set each value of $\left(T_{\mathrm{wh}},R_{\mathrm{wh}},T_{\mathrm{h}},T_{\mathrm{w}},T_s\right)$.
	7:	if $condition$ is satisfied
	8:	$Q_i$ is judged to be a defect.
	9:	end
	10:	end
	11:	Output the results.
Part II	1:	Given a gradient binary image $G_{\mathrm{w}}$.
	2:	Get the width $W$ of $G_{\mathrm{w}}$.
	3:	Set the step length $step$ and a rectangular search box $Sbox$ with a width of $w_b$ and a length of $h_b$. Set number of connected domains $N_i=0$.
	4:	fori = 1, …$w_b+i\ast step$, …, $W$ do
	5:	Use $Sbox$ to traverse $G_{\mathrm{w}}$ from top to bottom in steps of $step$.
	6:	For the $i-th$ $Sbox$, get k connected domains.
	7:	for j = 1, …, k do
	8:	Get the area $S_j$ of the connected domain $Q_{\mathrm{j}}\in Q$, the minimum bounding rectangle $T_j$ (with length $H_i$ and width $W_j$).
	9:	Set each value of $\left(T_{\mathrm{wh}},R_{\mathrm{wh}},T_{\mathrm{h}},T_{\mathrm{w}},T_{hs},T_{ss}\right)$.
	10:	if $condition1$ is satisfied.
	11:	$N_i+=1$ and save the coordinate of the connected domain $(x_j,y_j)$.
	12:	end
	13:	end
	14:	Set $T_n$ and $T_{hs}$.
	15:	if $condition2$ is satisfied.
	16:	The $i-th$ $Sbox$ contains defects.
	17:	end
	18:	end
	19:	Output the results.