Multiresolution images based on human visual system characteristic

Abstract

In digitizing, an image is usually sampled with an invariable frequency. For the high frequency part of the image, which plays an important role in the visual effect, the sampling is illogical. To overcome the limitation, the wavelet transform provides a multiresolution representation of an image. Incidentally, the resolution of an image can be decreased according to the human visual system characteristic. By reversion, based on the relativity among the wavelet subbands, the details of the higher resolution image are predicted by the details of the lower resolution image. So the higher resolution image is obtained by wavelet synthesis. Experimental results show that the scheme is feasible.