how to find the distance of all objects in given image

Answer 1

[DGM]

This will give an array mapping the distance from every object to every other object. You could reduce this with triu() if you want, due to the symmetry.

inpict = rgb2gray(imread('dots.jpeg'))>128;

L = bwlabel(inpict); % this identfies all the objects
C = regionprops(inpict,'centroid');
C = vertcat(C.Centroid);

D = sqrt((C(:,1)-C(:,1).').^2 + (C(:,2)-C(:,2).').^2);

If you wanted to find the distance to the nearest object, you could use this (there are probably other ways).

D(abs(D)<1E-6) = NaN; % remove zeros
[Dn Nn] = min(D,[],2); % minimize
% Dn is distance to nearest neighbor
% Nn is nearest neighbor

.

[Rahul punk]

??? Error using ==> minus

Matrix dimensions must agree.

Error in ==> te4st at 9

D = sqrt((C(:,1)-C(:,1).').^2 + (C(:,2)-C(:,2).').^2);

[DGM]

You must be using something prior to R2016b. If so, just use bsxfun().

D = sqrt(bsxfun(@minus,C(:,1),C(:,1).').^2 + bsxfun(@minus,C(:,2),C(:,2).').^2)

[Rahul punk]

for object = 1:length(stats)

bb = stats(object).BoundingBox;

bc = stats(object).Centroid;

tt= plot(bc(1),bc(2), '.');

aa=text(bc(1),bc(2), strcat('X: ', num2str(round(bc(1)))));

aa=text(bc(1),bc(2), strcat('X: ', num2str(round(bc(1)))));

how to extract aa values table on matlab array and ,get these x cordinates points to subtract each other to accurate distance measure

above solution work but not accurate in my case

[DGM]

You're going to have to better explain what you want or explain what exactly is wrong with accuracy. None of the code posted on this page mentions a table or any variable called aa. The code you posted just plots points. You're not going to have enough room to put labels on every point.

FWIW

inpict = rgb2gray(imread('dots.jpeg'))>128;

L = bwlabel(inpict);
C0 = regionprops(inpict,'centroid');
C = vertcat(C0.Centroid);

% distance from every object to every other object
%D = sqrt((C(:,1)-C(:,1).').^2 + (C(:,2)-C(:,2).').^2)
D = sqrt(bsxfun(@minus,C(:,1),C(:,1).').^2 + bsxfun(@minus,C(:,2),C(:,2).').^2)

D(abs(D)<1E-6) = NaN; % remove zeros
[Dn Nn] = min(D,[],2); % minimize
% Dn is distance to nearest neighbor
% Nn is nearest neighbor

% plot lines between points and the calculated nearest neighbor
imshow(inpict); hold on
for p = 1:numel(Dn)
	pts = vertcat(C0([p Nn(p)]).Centroid);
	plot(pts(:,1),pts(:,2))
end

Of course, I imagine a lot of these points have multiple neighbors which are at the same minimal distance.

You can still try to cram the labels in there

text(pts(1,1),pts(1,2),sprintf('X: %d\nY: %d',pts(1,1),pts(1,2)),'fontsize',8)

[Image Analyst]

If you have the stats toolbox you could also use pdist2() to find the distance of every point to every other point.

[DGM]

I didn't even know about that.

D = pdist2(C,C);

It's certainly a lot more concise, and having the extra distance types is a nice feature.

[Rahul punk]

[DGM]

C0 = regionprops(inpict,'centroid','equivdiameter');
C = vertcat(C0.Centroid);
R = vertcat(C0.EquivDiameter)/2;

% distance from every object to every other object (centers)
%D = sqrt((C(:,1)-C(:,1).').^2 + (C(:,2)-C(:,2).').^2)
D = sqrt(bsxfun(@minus,C(:,1),C(:,1).').^2 + bsxfun(@minus,C(:,2),C(:,2).').^2);
% distance between edges of objects (center distance minus each radius)
D = D-R-R.';

D(D<1E-6) = NaN; % remove self-distances
[Dn Nn] = min(D,[],2); % minimize

This will find (and minimize) the distance based on the distance between centroids minus the equivalent radius of each dot. I imagine this could also be done using improfile(), but I don't see any big advantage to doing it that way.

[Rahul punk]

getting error

??? Error using ==> minus

Matrix dimensions must agree.

Error in ==> test at 123

D = D-R-R.';

[DGM]

Auugh. I forgot you're using an older version. Same story:

D = bsxfun(@minus,bsxfun(@minus,D,R),R.');

Anytime you see something that looks like an implicit array expansion during an elementwise operation, bsxfun() can do it.

[Rahul punk]

[Rahul punk]

thanks for giving me your precious time to helping my question please tell can i get the ony horizontal distances.??

[DGM]

If this is the euclidean distance

D = sqrt(bsxfun(@minus,C(:,1),C(:,1).').^2 + bsxfun(@minus,C(:,2),C(:,2).').^2);

Then these are the components

Dx = bsxfun(@minus,C(:,1),C(:,1).';
Dy = bsxfun(@minus,C(:,2),C(:,2).';

[Rahul punk]

[Rahul punk]

could you find this distances in all blob mention in image?? i have try many method but dont get result?.if possible then all blob area find??

[Image Analyst]

Distance(s) of what from what? Did you get the centroids and then use pdist2()? I can't understand what your sentences are saying. Do you want the distances, areas, both? I have almost no idea. Please explain in much more detail.

% Find centroids and areas of all blobs.
props = regionprops(mask, 'Centroid', 'Area');
% Extract the area of all blobs.
allAreas = [props.Area];
% Get the centroids in an N-by-2 list of (x,y) coordinates.
xy = vertcat(props.Centroid);
% Get the distance of every blob's centroid to every other blob's centroid.
distances = pdist2(xy, xy);

[DGM]

I don't know what that means.

[Rahul punk]

i d'nt required center distance i have only required corner to corner distances. i agree with DGm Answer.