Abstract: An iterative filter algorithm is proposed to filter high density salt-pepper noise in images.It detects the noise pixel with minimum-maximum inspection, and then filters iteratively until all noise pixels are suppressed.To overcome the drawback of enlarged neighborhood, the algorithm takes a uniform 3×3 filtering window all the time, which ensures that all noise pixels be properly restored with enough information because of the trait of iterative manner.The output value is calculated using weighted mean based on each pixel's correlation with the filtering pixel.Compared with other methods, simulation results show that the proposed algorithm gives better signal-to-noise ratios (SNR) and more satisfactory images.