Improved Error Diffusion Based on HVS and First-Order Gradient Shaping Filter

Halftone image is a binary image that preserves the visual quality from the original continuous-tone image, and it is used for printing and output devices. In this paper, an improved error-diffusion with human visual system (HVS) model and first-order gradient shaping filter is proposed. The details of local edges are reflected, and it presents better halftone images. First, Floyd-Steinberg error diffusion algorithm is applied, and calculate the visual error based on HVS model, then obtain new continuous-tone images with HVS characteristics. Next, generate the details of information by extracting the first-order gradient shaping filter to regulate threshold, execute error diffusion procedure again on obtained new continuous-tone images. The halftone images from the improved algorithm not only meet the human eyes low-pass filter characteristics, but also enhance the local information of mutative details.