I made a similar post sci.image.processing a while back. I'm trying to understand why a my natural images are higher entropy than uniform noise. The order from lowest to highest is constant (0 obviously), uniform noise, LPF'd image, same image not LPF'd, gaussian noise My theoretical calculation for uniform noise agrees with what I'm getting -- about 1/256*log2(256)*241^2. Now there are nice smooth areas in my image. Should the entropy really be higher? Or am I just not doing this right.
which should have higher entropy?
Started by ●January 17, 2008