How to morphologically open a binary image ?

Hello , i am trying to morphologically open(to decrease the retrieved area) a binary image . However i am getting a black image (all pixels are 0) .What is wrong ?

se=strel('disk',3);
Im_open=imopen(original_binary,se);
figure;
imshow(Im_open);

[KALYAN ACHARJYA]

Please attach original_binary image.

Answer 1

[Image Analyst]

This works just fine:

grayImage = imread('moon.tif');
subplot(2, 2, 1);
imshow(grayImage);
title('Original Gray Scale Image', 'FontSize', 15);

original_binary = imbinarize(grayImage);
subplot(2, 2, 2);
imshow(original_binary);
title('Binarized Image', 'FontSize', 15);

se=strel('disk',3, 0);
Im_open=imopen(original_binary,se);
subplot(2, 2, 3:4);
imshow(Im_open);
title('Opened Image', 'FontSize', 15);

If it doesn't for you then it's possible all your white blobs in your original image are less than 7 pixels in diameter (3 in radius). Or else your image is so big that the small things just get subsampled away for display (zoom in, in that case to see them).

[Gn Gnk]

Yeah all of my white blobs are less than 7 pixels .What can i do?

[Gn Gnk]

Check the image.

[Image Analyst]

Then don't call imopen() of course. What imopen() does is an erosion of the given radius, followed by a dilation of the given radius. But after the erosion, all your blobs are gone. There is nothing left to dilate. What made you think you should do an opening anyway? Did someone tell you to do that?

In my experience, the usual use of imopen() is to get rid of little tendrils and spurious/random hairs that are sticking out of a blob. You certainly don't have that if all your blobs are so tiny that they're only a few pixels big.

[Gn Gnk]

In the project i am doing right now i have to take this binarized image and then perform the opening technique by using imopen() . I was asked to do this by the professor.

[Image Analyst]

OK, but then what that will do is to remove any blobs smaller than 3 pixels in radius, or 7 pixels in diameter.

[Gn Gnk]

Got it ! Do you know any other function that i can use to do what i want ?For example check this binary image before and after using imopen() . This is what i want to achieve in the end.

Answer 2

[asd asd]

Good luck with ΨΕΕ file!

[Gn Gnk]

Thank you Ψηλε!