2D filter design for the restoration of visually sharp images based on human perception characteristics

A type of 2D filter is proposed which can display visually high-contrast restored images by simulating the human-eye's perception characteristics. It is widely known that, in the spatial frequency domain, human-eyes have a kind of lowpass characteristic with emphasis at some middle frequency range. This may be interpreted as to make percepted images clearer by band-pass and reducing noise by high-cut characteristics. A trial is made so that, by designing 2D filters with simulated characteristics of human-eyes, much clearer images may be obtained from low-duality original ones. This paper shows that, under some conditions, the expected results are obtained for many image examples. Some modifications are desirable in filter coefficients in order to obtain much more satisfactory results.