How to find distance between two curves?
2 views (last 30 days)
Show older comments
Hey guys, I have a set of binary images each consisting of two curves, one directly on top of the other. I want to find the normal distance between them. This is very similar to another question found here: https://www.mathworks.com/matlabcentral/answers/214151-how-to-find-the-distance-between-lines-in-a-image-in-matlab. I'm guessing I need to fit curves onto the image first. Any help would be much appreciated!
3 Comments
Image Analyst
on 13 Jul 2016
Do you want vertical distance or "radial"?
Have you seen the FAQ for fitting a circle: http://matlab.wikia.com/wiki/FAQ#How_can_I_fit_a_circle_to_a_set_of_XY_data.3F
Accepted Answer
Image Analyst
on 13 Jul 2016
I'd invert the image so that the curves are white (foreground). Then I'd make masks for each curve by dilating the image a bunch and using bwareafilt() to extract the 2 largest blobs. Then for each blob, mask the image of the original, undilated image so that you have only that blob. Then I'd use find() to get all the coordinates and then fit it to a circle to get an average radius. Then do the same for the other masked blob. Subtract the two radii to get the average, overall separation.
9 Comments
Image Analyst
on 22 Jul 2016
I gave you the formula for the radius. Here it is again:
meanRadii = mean(sqrt((x - xCenter).^2 + (y - yCenter).^2));
Recall that you said you have a known, specified center, so you know what xCenter and yCenter is. Don't you? I assume you did when you said "given the origin". Because if you just had the origin but not the center, then you'd just use the FAQ like I originally said.
No you don't need to use some circle fitting routine. The formula for radius I gave you is, in essence, the fitting code. It works because you have specified the center already so all we need to determine is the radius, which would just be the average of the distances of each point from your known center.
More Answers (0)
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!