how to find the median of histogram?

i am trying to implement "Dualistic Sub-Image Histogram Equalisation". where the input image needs to be subdivided into two , based on the median of the image. any suggestion would help me.

Answer 1

[Adam]

Just sort all the image values and take the central one if you want the median value of an image.

If you want the median value of a histogram then just choose an odd number of bins and take the central one.

[Pradeep Gowda]

[x y]=imhist(m);
count=(sum(x(:)))/2;

will his work?

[Image Analyst]

Of course not. The median of the bin counts is not the same as the median of the image gray levels. See my answer.

Answer 2

[Image Analyst]

What's wrong with simply using the median() function?

theMedian = median(grayImage(:))
belowMask = grayImage <= theMedian;
aboveMask = grayImage > theMedian;
belowGray = grayImage .* uint8(belowMask);
aboveGray = grayImage .* uint8(aboveMask);

[Pradeep Gowda]

i guess this what i needed. One of the sub-images is equalized over the range up to the median and the other sub-image is equalized over the range from the median based on the respective histograms. Thus, the resulting equalized sub-images are bounded by each other around the input median? how do i equalzie them?

[Image Analyst]

You can use mat2gray() to scale them each to 0-1.

aboveGray = mat2gray(double(aboveGray));

Multiply by 255 if you need a uint8 image in the range 0-255

aboveGray = uint8(255 * mat2gray(double(aboveGray)));