finding dissimilarity between two binary image in a range between 0 and 10

Hi everybody, I have a problem that requires finding the dissimilarity between 2 binary images I tried to use the euclidean distance but the dissimilarity was large. If anyone can help me to implement such a code with a dissimilarity range between 0 and 10

Answer 1

[Alex Taylor]

If all you care about is an elementwise similarity metric that is normalized to the range [0 10]

similarity = @(A,B) 10*sum(A(:)==B(:))./numel(A);
similarity([false true true],[true true true])

Note that there are much better similarity metrics out there, but seemed like most of what you cared about was the normalization between [0,10], not how well the similarity metric works.

[Saad mohammad]

thank you Alex for your answer but can you please tell me what formula did you use for calculating the similarity

Answer 2

[Image Analyst]

Alex's formula is simply the percentage of pixels in A that match in B.

Here's another one called the Sorenson-Dice Coefficient: http://www.mathworks.com/matlabcentral/answers/68720#answer_80012

See Answers questions on it

Answer 3

[Muhammad Usman Saleem]

Absolute difference of two imagescollapse

Z = imabsdiff(X,Y)

About best selection method for image differences

http://www.mathworks.com/matlabcentral/answers/230382-what-is-the-best-algorithm-to-find-difference-between-two-images

Compare differences between images

h = imshowpair(A,B)

a lot of useful links

http://stackoverflow.com/questions/22296670/difference-between-two-images

http://stackoverflow.com/questions/8938817/image-comparison-using-matlab

https://www.researchgate.net/post/Differences_between_two_images_in_MATLAB

http://stats.stackexchange.com/questions/92303/how-to-calculate-disparity-of-two-images-in-matlab