Reduction of the blocking effect in block-coded images using a model of early human vision.

We propose a method, based on lowpass and bandpass components of Granrath's early human vision model, which reduces the blocking effect in block-coded images. The method consists of a lowpass filtering stage to smooth out the blocking effect and a contrast-enhancement stage to increase contrast in the restored image. Experimental results of the method show better performance than that of conventional methods for reducing blocking effects.