How can I count the black particles without counting the small particles in the background?

I want to automatically measure the particle size and number, but I can't because of the out-of-round shape and the contrast with the background.

rgb_im = imread('bild_1.jpg'); 
gray_im = adapthisteq(rgb2gray(rgb_im));
%  bw_im = imbinarize(gray_im,'adaptive','ForegroundPolarity','dark','Sensitivity',1); 
bw_im2 = ~imbinarize(gray_im,'adaptive','ForegroundPolarity','bright','Sensitivity',1); 
bw_im4 = imdilate(bw_im2,strel('disk',1)); 
BW1 = imfill(bw_im2,'holes'); 
BW2=255-BW1; 
figure, imshow(BW2)
imshowpair(BW1,BW2,'montage')

I have inverted the image and now I want to remove the small black dots from the background. However, I can't get that to work.

Answer 1

[Antoni Garcia-Herreros]

Hello loris

Regionprops does exactly what you are looking for.

You can try something like this:

clear all
close all
rgb_im = imread('bild_1.jpg');  %Read RGB image

Error using imread>get_full_filename
File "bild_1.jpg" does not exist.

Error in imread (line 372)
fullname = get_full_filename(filename);

gray_im=rgb2gray(rgb_im);   % RGB to grayscale

BW=~imbinarize(gray_im,'adaptive','ForegroundPolarity','bright','Sensitivity',1); %Binarize image
BWao = bwareaopen(BW,20); % Remove particles smaller than 20 pixels
BWaod = imdilate(BWao,strel('disk',2)); 
BWaodf=imfill(BWaod,'holes');

r=regionprops('table',BWaodf,'Centroid','Area','Circularity','EquivDiameter'); 
m=table2array(r); %Transform table to array

%Filter by cicularity
thr=0.6; % You may change this parameter
Points=m(m(:,4)>thr,:); % Array containing all the Particles 
% Points=[Area [pixels], Centroid X, Centroid Y, Circularity, Equiv Diameter ]

imshow(rgb_im)
hold on
viscircles([Points(:,2),Points(:,3)],Points(:,5)*0.5,'LineWidth',0.5);

Nparticles=size(Points,1); %Number of Particles
MeanD=mean(Points(:,5)); % Mean Diameter

Hope this helps