A comparison of adaptive order statistic filters using different images

Histogram windowing, a local statistic-based edge detector, an adaptive window size, and a median filter are compared to the median filter, the SAM filter, and the moving average filter. The five filters are compared using the peak-to-peak signal-to-noise ratio (PSNR). Four different images are used for comparison: (1) Lenna, (2) geometrical, (3) harbour, and (4) face. Three noise types are used: (1) additive, (2) impulse, and (3) a combination of the two. The results of the experiment show that the PSNR depends very much on what kinds of image are used.<
>