General Structure of the Proposed Method
The proposed filter is an adaptive fuzzy filter for
eliminating Gaussian noise. The overall structure of the
proposed filter is shown in Figure 1. The working
procedure is as follows: first the noiseless image
histogram is estimated from the input image, for this we
have provided a new method. Afterwards the image
undergoes a fuzzy conversion stage. Fuzzy inference
rules are applied according to Takagi Sugeno and the
output is calculated. The calculated output function and
its difference with the estimated histogram of the input
image are sent as error parameters to the routine used to
change parameters of membership functions (as in Figure