Generally fuzzy adaptor filters provide better results in
eliminating impulse noise as compared to ordinary
methods, specially in high NP that ordinary methods are
not efficient and don’t work at all. We proposed a new
interesting method for estimating the histogram of the
original image from the image containing gaussian noise.
We used the difference of this estimated image histogram
with the histogram of the filtered image as an evaluation
function through which we tried to find the best output
result by changing parameters of membership functions.
Obtained results prove the higher performance and
efficiency of our proposed method.