How does Matlab produce thin edges?

Im using sobel edge detection.

Here is the original image file if you have Matlab so you can take a look

How does Matlab, by itself manage to achieve very thin edges like this one (using just the Matlab's edge function with 'sobel' as parameter)

matlabsobel = edge(originalImage,'sobel')
imshow(matlabsobel)

but when I try to do sobel algorithm my own way, assuming the process is just the same..

originalImage = gaussianizedimage;
threshold = 60.5;
k = [1 2 1; 0 0 0; -1 -2 -1];
H = conv2(double(originalImage),k, 'same');
V = conv2(double(originalImage),k','same');
E = sqrt(H.*H + V.*V);
edgeImage = uint8((E > threshold) * 255);
imshow(edgeImage);
title('sobel algorithm')

Why is that its different? What is still lacking in my implementation to achieve Matlab's thin edges? Can anyone provide code to make it look/similar to Matlab native sobel edge detection?

Thanks

Answer 1

[Image Analyst]

edge() takes a grayscale edge image like you'd get with imgradient(), and then thresholds it and skeletonizes the thresholded image, like you'd get with bwmorph(BW, 'Skel', inf).

[Ivan Matala]

@Image Analyst

thanks for you answer,,, Ive searched bwmorph and if by chance,, do you what is the best suited morphological operation to make it look like the 1st image above?

Thanks

[Image Analyst]

I told you that already. Here is is again, formatted as code this time:

thinEdgeImage = bwmorph(thresholdedImage, 'Skel', inf);

[Miroslav Hagara]

according this, if I use this code:

BW1 = edge(I,'Prewitt');

BW2NT = edge(I,'Prewitt','nothinning');

BW2 = bwmorph(BW2NT, 'Skel',Inf);

I should get the same pictures. But this is not true.

I have get better result with

BW2 = bwmorph(BW2NT, 'thin', Inf)

but still without same pictures.

[Royi Avital]

If you look inside edge() you will see:

if thinning
    e = computeedge(b,bx,by,kx,ky,int8(offset),100*eps,cutoff);
else

Which does the magic.

MathWorks doesn't specify what's done there.