0
$\begingroup$

I have picture with high resolution with 20M memory space and this image has been taken from the standing away from the window in room which has significantly less light as compared to outside. Image has less contrast due to large number of bright pixels in it and also large number of near to dark pixels from room objects. I want to enhance the image contrast having pixels with high intensity values and also with having low intensity values without changing the remaining pixels which are in between these range. Image has been taken with almost half of the areas with having high intensity values and half of area having low intensity values. I have read the quantization techniques with uniform quantization and non-uniform quantization Wiki but i am unable to understand the the concept I mean in what direction i need to understand the concept behind this to improve the contrast and what techniques are available to apply these kind of problems.

$\endgroup$
  • 2
    $\begingroup$ Have you tried histogram equalization. There are a number of different versions. Tools like gimp include these. $\endgroup$ – Stanley Pawlukiewicz Apr 23 '18 at 17:07
  • $\begingroup$ @StanleyPawlukiewicz What it matters is to discretization from continuous space or may be i am not clear about the back ground of my objective. Histogram equalization will work but will this transform on to less storage space (20M to less space). $\endgroup$ – Waseem Ali Apr 24 '18 at 14:42
  • $\begingroup$ Your image has pixels that are saturated and pixels that are underexposed. Nothing you can do will fix that. You can try to reduce brightness in one area and increase it in the other, but it's not going to solve anything. You should improve how you take the image. $\endgroup$ – Cris Luengo Apr 24 '18 at 18:18
  • $\begingroup$ histogram equalization doesn’t necessarily change the size of the file. You have a 20M image. It is already quantized. Each pixel is already represented by integers. $\endgroup$ – Stanley Pawlukiewicz Apr 24 '18 at 18:57

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.