What techniques can we use in image processing to reduce the entropy of an image?
From what i understand entropy is basically the amount of different "information" contained inside of an image. Going from there the general understanding would be to reduce said information in order to reduce entropy. I haven't been able to find reliable resources of how to reduce entropy while still allowing us to make out what is in the picture with our eye. Some suggest that:
- Playing around with the contrast can reduce the entropy.
- Having salt and pepper noise reduces the entropy.
Is there another more concrete paper or technique on how to reduce entropy ?
Here are the following entropies after adjusting contrast and adding salt and pepper noise:
Original image
Original image with contrast adjusted
Original image with Salt and Pepper noise added
The easiest approach is, to play with the compression factor associated with formats as jpg. You have to reduce entropy for a significant picture compression, as soon primitive mechanisms as run length encoding are exhausted. (They don’t work well for multi-color pictures anyway.)