In many image processing papers, I've seen that they used fuzzy logic for segmentation I wonder how fuzzification impact the result that made Fuzzy-C-Means better than ordinary K-Means.

PS. If possible could you provide me the sample data sets for case study

  • 1
    $\begingroup$ I don't think it "requires" fuzzy c-means. But as there is "noise" in such data, it may still work better than ordinary k-means. $\endgroup$ – Anony-Mousse Feb 20 '16 at 9:53

If you take aerial images for example, you might need to identify overlaps.

Lets say your goal is to segment water, vegetation and rock areas. More to be assigned to the closest centroid, there are hidden memberships to other centroids that can be identified by fuzzy c-means image segmentation.

On this thesis at https://zone.biblio.laurentian.ca/handle/10219/2616, at section (4.5 Analysis - Task 1) you can find detailed analysis using both (k-means & fuzzy c-means).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.