Algorithm 1: Circle detection using Hough transform
	Input: the segmented images, image coordinates $(x,y)$.
	Output: circle center coordinates $(a,b)$, circle radius $\left(r\right)$.
1.	Initialize: $A[a,b,r]=0$.
2.	Detect all edge points in the segmented image using the Canny operator.
3.	for $r=0$ to the diagonal length of the image
4.	for each edge pixel $(x,y)$ in the image
5.	for $\theta$ = 0 to 360
6.	$a=x-r\times cos\theta$
7.	$b=y-r\times sin\theta$
8.	$A[a,b,r]=A[a,b,r]+1$
9.	Find the largest and second largest values of $A[a,b,r]$, whose parameters correspond to the inner and outer circles of the weld.