I have a binary image, how can I count the balck pixels as a group??

I process a rgb image to gray to binary and use filters to remove noise and get some specific black areas. Now I want to count the areas. How many are there? How can I do that? can anyone help me?

Answer 1

[Walter Roberson]

Add a white border around the entire image. Now use regionprops() and ask for EulerNumber

"Returns a scalar that specifies the number of objects in the region minus the number of holes in those objects. "

The number of objects in the region would be 1 (all of the white part together as you have no isolated white areas), so the number you get out would be negative; take 1 minus the number to get the number of black regions.

Or you could invert the binary matrix. After that there would be a number of different approaches, including, for example, asking regionprops for any simple statistic such as Area, and then just counting the number of outputs you get

NumberOfBlackAreas = length( regionprops( ~YourBinaryMatrix, 'Area' ) );

[REZA]

Hi, thank you for your reply. Can you please explain a little with some code example...As I am new with matlab..

[Walter Roberson]

cman = imread('cameraman.tif');
YourBinaryImage = cman > 30;
NumberOfBlackAreas = length( regionprops( ~YourBinaryImage, 'Area' ) )

Answer 2

[Image Analyst]

You can get it by inverting your image and calling bwlabel:

[labeledImage, numRegions] = bwlabel(~binaryImage);

No need to call regionprops() just to get the count, though you might want to if you want to measure other things like areas.