A Modified Error Diffusion Scheme Based on the Human Visual Model

초록

Various techniques have been proposed to reduce the artifacts resulted from the Floyd-Steinberg's error diffusion algorithm. Recently, an adaptive error diffusion scheme which dynamically adjusts the weights for the error propagation was reported.1 This paper presents a modified adaptive error diffusion scheme to further reduce the undesirable artifacts. In the proposed scheme, the human visual characteristics are considered by adding the photopic modulation transfer function to the error criterion for the minimization. Also, weight updating method during the minimization is modified. Experimental results of the proposed scheme are compared with those obtained by the Floyd-Steinberg's algorithm and the adaptive error diffusion scheme.