I can convert my RGB image to grayscale and then change it to binary format using histogram and if-functions. So, I have a binary image with white background and some separate black points. How can I quickly count the number of black points in white background? Is there any function in MATLAB?
Use im2bw:
http://www.mathworks.com/help/images/ref/im2bw.html
Then use nnz to get nonzero values:
http://cens.ioc.ee/local/man/matlab/techdoc/ref/nnz.html
Look for any blob analysis function. You can adjust the size of blob to find objects of a minimum size. Typically, if you just run the blob analysis function, you will get the number of blobs (black points in your case) alongside the properties (for example, size, location, and bounding box).
Dear Bhatia
Thanks a lot.
Calculating the size and position changes of the aggregates, in addition to their number, are among the main purposes of my case, tracing non-suspended soil particles. So, I'll try to use blob analysis function.
get a look to section 9.4 of the book AK Jain Fundamentals of digital image processing. I would use a template similar to table 9.4.2, although this one is for detecting White dots in a black background. you can invert your image or "invert" the template: [1, 1, 1
1, -8, 1
1, 1, 1]
yes. there is. now you must inverse image that have black background and white foreground. then u can use matlab code as below (your binary image is I)
[l8, num8]=bwlabel (I, 8);
d6=regionprops (l4,'BoundingBox');
for i=1: size (d6, 1)
f {i,1}=imcrop (I, d6 (i, 1).BoundingBox);
end
now u have number of points at size(d6,1) and have these pictures at f cell.
I think the shortest way to go for "ImageJ" it is an easy freeware with various add- ins!
Add the number of ones then minus it with the size of matrix.
Matlab function:
m*n-sum(sum(a))
Dear Hemant
What do you mean?
The number of points in the target
take negative of that image. then count the white pixels. and you get the no of white pixels in the original image.
OR
n1=numel(Original image),
n2=nnz(original image),
n3=n1-n2;% you get no of black pixels.
how to detect separate cells of blood cells of a image any idea using matlab r2017b
I second the ImageJ suggestion - use the Analyze particles to get the number and size of each black object. (Or, if you have a stack of images, use 3d object counter).
- Badges
- Science topic