BIT-based weighted mean filter

Linear-Nonlinear hybrid filters that have appeared in literature suffer from some severe disadvantages. They smear edges and are very hardware intensive. These shortcomings can be overcome by having a Bit-based Weighted Mean filtering scheme. This filter starts by calculating the median of a set of sample values. The sample values are then scaled. Those values which lie in the proximity of the median, are granted more weightage. The weighted sample values are then averaged to yield the filter output. Results show that these filters perform much better than their earlier counterparts with respect to edge preservation and minimizing the minimum absolute error criterion when applied to images corrupted by both impulsive and nonimpulsive noise. These filters are also much more hardware efficient than the L, ATM and M filters[5]-[6].