How to distinguish different color cells in the image?

I have an image with two different color cells like below shows

and I need to claasify it and count how much cells like below shows red and yellow

how can I do that?

I have tried edge detection but it seems doesnt work.

Answer 1

[Image Analyst]

They don't look very different to me. I'd try the Color Thresholder on the Apps tab of the tool ribbon. Try HSV and RGB color spaces and see if you can get differentiation.

[larry liu]

if I got the image like this, then how can I count the number of circles?

[Image Analyst]

You can first get rid of the grid with imclearborder(). That might get rid of some blobs that are connected to it though. If you don't want to lose them, then don't use a grid.

Then you can find round particles by looking at their circularity.

mask = imclearborder(mask); % Get rid of grid and blobs touching it.
mask = imfill(mask, 'holes');
props = regionprops(mask, 'BoundingBox', 'Area', 'Perimeter');
bb = vertcat(props.BoundingBox);
widths = bb(:, 3);
heights = bb(:, 4);
aspectRatios = widths ./ heights;
allAreas = [props.Area];
allPerimeters = [props.Perimeter];
circularities = allPerimeters .^ 2 ./ (4 * pi * allAreas);
% Get indexes of blobs where aspect ratio is less than about 3 and
% circularities less than about 5.  Adjsut as needed.
keeperIndexes = (circularities < 5) & ...
    (aspectRatios < 3) & ...
    (1./aspectRatios < 3)
[labeledImage, numInitialBlobs] = bwlabel(mask);
fprintf('Before shape filtering, found %d blobs.\n', numInitialBlobs);
mask = ismember(labeledImage, find(keeperIndexes));
imshow(mask);
% Remeasure with the filtered set of blobs.
props = regionprops(mask, 'BoundingBox', 'Area', 'Perimeter');
numFinalBlobs = length(props);
fprintf('After filtering, found %d blobs.\n', numFinalBlobs);