How to find local minima in a image histogram?
2 views (last 30 days)
Show older comments
Hello, I have few search about it in google. but i'm confused :( I would like to search for the first (leftmost) local minimum in the image histogram. For example the attached image is histogram of sample image. and we need local minima. I need just one number. Thanks
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/168473/image.jpeg)
0 Comments
Answers (1)
Image Analyst
on 20 Oct 2017
Edited: Image Analyst
on 20 Oct 2017
Use diff() to get the difference between the n+1st bin and the nth bin. Then use find to get the index
[counts, grayLevels] = imhist(grayImage);
index = find(diff(counts) >= 0, 1, 'first');
minGL = grayLevels(index);
5 Comments
Walter Roberson
on 20 Oct 2017
That would not be uncommon. Suppose an image had no true black, then the first histogram entry would be a count of 0. The very first shade that did occur would have a non-zero entry, which is an increase. You would have a local minimum of 0 from the first bin to the first bin that had some data, at which point you would have an increase.
Image Analyst
on 20 Oct 2017
Edited: Image Analyst
on 20 Oct 2017
Yes, that's the first place where the histogram is flat. Perhaps you want findpeaks() where you get more control over how big a peak or valley you detect.
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!