There is a recursive way of finding threshold in the image using histogram that are biomodal, and is as follows:

An automated procedure for bimodal histograms :

  1. Select an initial estimate of T
  2. Compute the means of the two regions determined by T
  3. Set the new T as the average of the two means
  4. Repeat steps 2,3 until the difference in T in successive iterations is smaller than a predefined parameter

I have an old exam question below and I dont know how to solve it. all I know the T is $114$ and if this also helps total number of pixels are area first triangle+are of second triangle area + 4000 pixels. I am stuck on how to calculate step 2 to get me going.

Old Exam Question

explain why the isodata threshold of image is equal to $114$.

  • $\begingroup$ If this is a question for a real exam you're taking, we can't solve it for you. $\endgroup$ – Phonon Mar 6 '14 at 20:14
  • $\begingroup$ Well this is not a exam question that I am taking , It is old exam paper which is available online on university database. I appreciate more not to give a full solution but to give a hint to get me going. like this I learn. $\endgroup$ – Solix Mar 7 '14 at 15:22
  • $\begingroup$ @Phonon could you please unhold this question? $\endgroup$ – Solix Mar 7 '14 at 15:25
  • $\begingroup$ All good. We're just trying to be careful with content. $\endgroup$ – Phonon Mar 7 '14 at 22:02
  • $\begingroup$ For step 2, split the data into two pieces: samples < 114 and samples >= 114. Compute the means of each piece. $\endgroup$ – John Mar 8 '14 at 0:26

Your Answer

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

Browse other questions tagged or ask your own question.