How to get centroid specific part of an image

I am intended to compute centroid of a part of image. I am getting centroid of all parts of the image which makes me confused of selecting the right one. my code is:

I = imread('Murine_Tibia_Crosssection.png')
Ibw = im2bw(I);
Ibw = imfill(Ibw,'holes');
Ilabel = bwlabel(Ibw);
stat = regionprops(Ilabel,'centroid');
imshow(I); hold on;
for x = 1: numel(stat)
  plot(stat(x).Centroid(1),stat(x).Centroid(2),'ro');
end

please help me to get the right one.

[Andrew Newell]

I formatted your question. Please have a look at Markup Help.

[Ajay Goyal]

Christiaan Sir, Your help is deeply appreciated. You made my day since I was working on this problem for last 3 days. I have made some amendments in your code to get it fit to my need. clc;clear all;close all; J = imread('final1.jpg'); figure(1) imshow(J) I=imcomplement(J); figure(2) imshow(I) stat = regionprops(I,'centroid'); h = fspecial('unsharp'); I_filtered = imfilter(I,h); figure(3) imshow(I_filtered) level = graythresh(I_filtered); BlackWhite = im2bw(I_filtered, level); figure(4) imshow(BlackWhite); original = BlackWhite; filled = imfill(original, 'holes'); holes = filled & ~original; bigholes = bwareaopen(holes, 500); smallholes = holes & ~bigholes; new = original | smallholes; figure(5) imshow(new); stat = regionprops(new,'centroid'); hold on; for x = 1: numel(stat) plot(stat(x).Centroid(1),stat(x).Centroid(2),'b*','linewidth',2); end Thanks Again

Answer 1

[Christiaan]

Dear Ajay,

If an image shows multiple centroids, there are some tricks to still determine the centroid. First you can filter the image using the function medfilt2. Then you can make the image black and white with the function im2bw. Finally you can fill holes when the size of the hole is smaller than a specific size with the function bwareaopen.

See the code below for an example:

clc;clear all;close all;

I = imread('Murine_Tibia_Crosssection.png');
figure(1)
imshow(I)
stat = regionprops(I,'centroid'); 

I_filtered = medfilt2(I, [12 12]);
figure(2)
imshow(I_filtered)

level = graythresh(I_filtered)
BlackWhite = im2bw(I_filtered, level);
figure(3)
imshow(BlackWhite);

original = BlackWhite;
filled = imfill(original, 'holes');
holes = filled & ~original;
bigholes = bwareaopen(holes, 500);
smallholes = holes & ~bigholes;
new = original | smallholes;

figure(4)
imshow(new);

stat = regionprops(new,'centroid'); 
hold on; for x = 1: numel(stat) 
    plot(stat(x).Centroid(1),stat(x).Centroid(2),'ro','linewidth',5); 
end
hold off

Good luck! Christiaan

Answer 2

[Image Analyst]

My Image Segmentation Tutorial shows you how to get centroids.

http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

Answer 3

[Ajay Goyal]

Christiaan Sir, Your help is deeply appreciated. You made my day since I was working on this problem for last 3 days. I have made some amendments in your code to get it fit to my need. clc;clear all;close all; J = imread('final1.jpg'); figure(1) imshow(J) I=imcomplement(J); figure(2) imshow(I) stat = regionprops(I,'centroid'); h = fspecial('unsharp'); I_filtered = imfilter(I,h); figure(3) imshow(I_filtered) level = graythresh(I_filtered); BlackWhite = im2bw(I_filtered, level); figure(4) imshow(BlackWhite); original = BlackWhite; filled = imfill(original, 'holes'); holes = filled & ~original; bigholes = bwareaopen(holes, 500); smallholes = holes & ~bigholes; new = original | smallholes; figure(5) imshow(new); stat = regionprops(new,'centroid'); hold on; for x = 1: numel(stat) plot(stat(x).Centroid(1),stat(x).Centroid(2),'b*','linewidth',2); end Thanks Again