distribution of excess pixels

I am trying to manually equalize the image using standard histogram equalization algorithm in which I am mapping the pixels to 4 bins but when I compare my histogram to the one obtained from histeq(image,4) both are bit different. You can see them here

Here as we can see in the histogram obtained using histeq (first one) all the four red dots are nearly at the same level but not in the second case. When I found the count of pixels at each bin I got following. For the histogram generated as a result of histeq( we can see uniform distribution of pixels here) following is the distribution

 Bin number  No. of pixels
       1             4207
      86             4146
     171             4314
     256             3965

For the histogram generated as a result of manual equalization( we can see non-uniform distribution of pixels here) following is the distribution

Bin number  No. of pixels
     1        2678
    86        5583
   171        5061
   256        3310

Now we can see that in my case there is no uniform distribution of pixels because of which my histogram is different from the one obtained using histeq. Now I want to clip the histogram and distribute the excess pixels to other bins but the question is which pixels should be considered as excess pixels that can be moved to other bins? and to which bin should these pixels be moved? Please explain how this done. Thank you.

Btw I got this link but here i am not able to understand the term priority i.e what is priority of a pixel and how is it calculated?

http://stackoverflow.com/questions/18850508/distributing-equal-number-of-pixels-in-each-intensity-level?answertab=active#tab-top

Answer 1

[Image Analyst]

That is quantization, NOT equalization. You can use histeq() to do a histogram equalization using the standard crude method.

If you want a truly flat histogram, you can adapt my method: http://www.mathworks.com/matlabcentral/fileexchange/28972-custom-shaped-histogram

[Navdeep Sony]

@Image Analyst, thanks for replying. Why do we use quantization? Is it used to increase contrast? and do histeq also used quantization apart from equalization?

[Image Analyst]

I don't know what that application would be. Maybe to achieve a certain artistic look perhaps? Or to save space because they can be compressed more. It might increase contrast around the image right around where the image has those quantization levels, but I don't think in a good way.