I want to measure the diameter of the circle using image.

I want to measure the diameter of the circle. Please let me know the respective coding for the below attached image.

I used the below code for finding the diameter:

a  = imread('I:\testing\abrupt\60.jpg');
c = rgb2gray(a);
b = imbinarize(c);
imshow(b);
%measuring
h = imdistline(gca);
api =iptgetapi(h);
%
pause();
%
dist =api.getDistance();
u = menu('Choose measuring unit','Pixels','Centim','Millimeters');
if (u==1)
    fprintf('The length of the object is: %0.2f Pixels\n',dist);
elseif (u==2)
    dist_cm = dist*0.2645;
    fprintf('The length of the object is: %0.2f Centim\n',dist_cm);
else
    dist_m = (dist)*0.2645*10;
    fprintf('The length of the object is: %0.2f Millimeters\n',dist_mm);
end

[Ramya Anandhan]

Thank you.

[DGM]

Did my answer do what you need?

If (or when) you find that it satisfies your question, you can click "accept". That way this thread will be moved to the "answered" queue, where it might better help others looking for help with a similar problem.

[Ramya Anandhan]

Hello,

Do you know how to calculate the diameter for the attached picture accurately using Matlab coding.

The actual average spread diameter was (12.4+12.25)/2=12.325.

I should get the average diameter of 12.325.

[DGM]

Note that the blob appears to be 12.xx cm in radius. The diameter is somewhere around 25 cm or so. There's a bit of a perspective distortion, but it's small.

a  = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/911220/751.jpg');
c = rgb2gray(a);
b = imbinarize(c);
b = imfill(bwareafilt(~b,1),'holes'); % clean up the image
b = imopen(b,strel('disk',10,0)); % get rid of graticule lines

imshow(b);

% these are the (manually) measured sizes of the graticule
% 963px/40cm x
% 986px/40cm y

% in a very crude attempt to compensate for perspective
% i'm just going to tweak the image size 
b = imresize(b,[size(b,1) size(b,2)*986/963]);
% now both axes have the same scale of 986px/40cm
objscale = 986/40; % px/cm

% when you have an irregular shape, it depends what you want when you say "diameter"
S = regionprops(b,'minferetproperties','maxferetproperties','equivdiameter'); 

% let's say we pick one
dist = S.EquivDiameter;
%dist = (S.MaxFeretDiameter + S.MinFeretDiameter)/2; 

% i can't demo the menu in-browser, so i'm just bypassing it
%u = menu('Choose measuring unit','Pixels','Centim','Millimeters');
u = 2;

if (u==1)
    fprintf('The length of the object is: %0.2f Pixels\n',dist);
elseif (u==2)
    dist_cm = dist/objscale;
    fprintf('The length of the object is: %0.2f Centim\n',dist_cm);
else
    dist_mm = dist/objscale*10;
    fprintf('The length of the object is: %0.2f Millimeters\n',dist_mm);
end

The length of the object is: 26.26 Centim

% original estimate (in cm)
origest = mean([12.4 12.25])*2

origest = 24.6500

It's off by a bit, but considering the eccentricity of the blob, I think it's plausible that the original estimate of 24.6cm is an underestimate.

[Ramya Anandhan]

Thanks for the code.

Answer 1

[DGM]

This should be a start

a  = imread('https://www.mathworks.com/matlabcentral/answers/uploaded_files/910160/21.jpg');
c = rgb2gray(a);
b = imbinarize(c);
b = imfill(bwareafilt(~b,1),'holes'); % clean up the image

imshow(b);

% when you have an irregular shape, it depends what you want when you say "diameter"
S = regionprops(b,'minferetproperties','maxferetproperties','equivdiameter'); 

% show what the options look like
S.EquivDiameter % the equivalent diameter of a circle with the same area

ans = 961.0452

S.MinFeretDiameter % the minimum diameter

ans = 948

S.MaxFeretDiameter % the maximum diameter

ans = 987.3196

% let's say we pick one
dist = S.EquivDiameter;

% i can't demo the menu in-browser, so i'm just bypassing it
%u = menu('Choose measuring unit','Pixels','Centim','Millimeters');
u = 3;

if (u==1)
    fprintf('The length of the object is: %0.2f Pixels\n',dist);
elseif (u==2)
    dist_cm = dist*2.5/1395;
    fprintf('The length of the object is: %0.2f Centim\n',dist_cm);
else
    dist_mm = dist*25/1395;
    fprintf('The length of the object is: %0.2f Millimeters\n',dist_mm);
end

The length of the object is: 17.22 Millimeters

Note that I changed the scaling factor. Since I doubt the blob is 2.5m in diameter, I just approximated the above scaling factor on the assumption that the glass is a standard 25mm microscope slide. The accuracy of that approximation can stand to be improved.

Once you decide which diameter metric you want, the other cases can be omitted from the call to regionprops().

[Ramya Anandhan]

Yes. Thanks for the help.

[Agata Agnieszka]

Why do we divide the distance by 1395?

[DGM]

As mentioned, the scaling factor was an uninformed assumption that the glass was a standard 25mm microscope slide. The glass is approximately 1395px wide. So we have 25mm/1395px.

As to whether that's actually a 25mm slide, only OP can verify.

[Agata Agnieszka]

thank you!