MATLAB: How to find the centroid of cells in an image

cell countingcentroid

clear
% Read in image
image = imread('F:\SPRING 2020\LIVING SYSTEM ANALYSIS-BIOSTATS\CELL SEEDING\A1_TR_02122020_1200.jpg');
% Show image you're working with
imshow(image,[]);
% Convert image to a grayscale image
grayImage = rgb2gray(image);
% Show grayscale image
imshow(grayImage,[]);
% Enhance the contrast of the image using imadjust
image_imadjust = imadjust(grayImage, [0 1], [0 1], 0.4);
imshow(image_imadjust,[]);
imshowpair(grayImage,image_imadjust,'montage');
% Additional step of removing noise by adaptive filtering using a small 5×5
% window
noiseRemoval = wiener2(image_imadjust, [10 10]);
imshow(noiseRemoval,[]);
imshowpair(image_imadjust,noiseRemoval,'montage');
% Creating a binary mask that contains the edge detected cells
binary_mask = edge(noiseRemoval,'sobel');
imshow(binary_mask,[]);
imshowpair(grayImage,binary_mask,'montage');
% Fill in the gaps in the lines surrounding the object in the gradient
% mask by dilating the sobel image using linear structuring elements
% Create two perpindicular linear structuring elements using strel function
se90 = strel('line',3,90);
se0 = strel('line',3,0);
% Dialate the binary gradient mask using the vertical structuring element
% followed by the horizontal structuring element
dilated = imdilate(binary_mask,[se90 se0]);
imshow(dilated,[]);
% Filling any holes within the cells
fill = imfill(dilated,'holes');
imshow(fill,[]);
% Perform another morphological opening of the cells using a disc kernel
fillagain = imopen(fill,strel('disk',4));
imshow(fillagain,[]);
imshowpair(fill,fillagain,'montage');
% Objects on the borders can be caused by noise and other artifacts
% Can eliminate objects on the borders using imclearborders
clearBorder = imclearborder(fillagain,8);
imshow(clearBorder,[]);
imshowpair(fillagain, clearBorder,'montage');
% Remove all connected components (cells) that have fewer than 10 pixels
bware = bwareaopen(clearBorder, 100);
imshow(bware,[]);
% Find the perimeter of cells
bw4_perim = bwperim(bware);
% The [1 .3 .3] is just specifying the color
overlay1 = imoverlay(bware, bw4_perim, [1 .3 .3]);
imshow(overlay1,[]);
I = grayImage .* uint8(bware);
imshow(I,[]);
% Some cells will overlap so apply watershed algorithm on the image which
% will be able to partially divide the groups into distinct cells
% Find cell centroids
maxs = imextendedmax(I, 5);
maxs = imclose(maxs, strel('disk',2));
maxs = imfill(maxs, 'holes');
maxs = bwareaopen(maxs, 1);
overlay2 = imoverlay(I, bw4_perim | maxs, [1 .3 .3]);
imshow(overlay2,[]);
% modify the image so that the background pixels and the extended maxima pixels are forced to be the only local minima in the image.
Jc = imcomplement(I);
I_mod = imimposemin(Jc, ~bware | maxs);
% Apply the watershed algorithm
L = watershed(I_mod);
labeledImage = label2rgb(L);
% Count number of cells
[L, num] = bwlabel(L);
% Overlay the detected cells over the original grayscale image to evaluate
% the performance of the algorithm
mask = im2bw(L, 1);
overlay3 = imoverlay(I, mask, [1 .3 .3]);
imshow(overlay3,[]);
I am trying to find the centroid of the cells in this image so that I can count the number of cells despite if any of them overlap, however the code that I have bolded above is not finding the centroid of every cell. I was wondering if I have my numbers wrong or what could be the solution to my problem, because I still want to overlay the centroids onto my image to see that it has correctly found the centroids. I've included the original image below.

Best Answer

Continue your code with this:
% Find the centroids of the mask.
props = regionprops(mask, grayImage, 'Centroid', 'WeightedCentroid');
allCentroids = vertcat(props.Centroid);
% Plot centroids over image in the overlay
hold on;
for k = 1 : size(allCentroids, 2)
    x = allCentroids(k, 1);
    y = allCentroids(k, 2);
    plot(x, y, 'r.', 'MarkerSize', 25)
    % You can plot the Weighted Centroids (weighted by the gray level) instead if you want.
end
Related Question