ADAPTIVE FUZZY FILTERING METHOD

摘要

PURPOSE: An adaptive fuzzy filtering method is provided to prevent an edge portion of a video from decreasing due to a filtering by adjusting a fuzzy filtering at the edge portion and enhance the efficiency of the filtering by adaptively using an intermediate value and a mean value. CONSTITUTION: A size of a window is set for a fuzzy filtering(301). A dispersion value is calculated for obtaining a distribution degree of pixel values within the set window(302). The calculated distribution value is compared with a threshold value(303). If the calculated distribution value is less than the threshold value, a mean value of pixels within the window is set as an initial central value(prototype) of the window(305). If the calculated distribution value is more than the threshold value, an intermediate value of pixels within the window is set as the prototype(304). All pixel values within the window are smoothed with reference to an edge component(305). A membership function is calculated(306). A level of the calculated membership function is compared with a threshold value(307). If the membership function is more than the threshold value, the level of the membership function is set as '0'(308). If the membership function is less than the threshold value, parameters are renewed(309). It is judged whether the parameter is converged(310). If the parameter is converged, an output value of a corresponding pixel is determined as the parameter(311).