A Error Diffusion Halftone Algorithm Based On Visual Error Feedback

Digital halftone is a screen technology. It bases on human visual system and imagecharacteristics, using mathematical tools and computer tools in binary or polychromaticbinary device to realizes image representation. Now, it plays a significant role inprinting,display fields. With improving performance of digital hard-copy device,halftone algorithm has become one of the key factors that affect the quality of theoutput.To solve the problems of fuzzy edges, poor contrast and directional hysteresisexisting in the improved error diffusion algorithms, this paper presents an improvedhalftoning algorithm based on human visual system model. Firstly, Floyd-Steinbergerror diffusion algorithm is applied and the visual error is calculated based on HVSmodel. Secondly, new continuous-tone images with HVS characteristics by adding thefeedback of visual error to original images are obtained. Finally, error diffusionprocedure is executed again to get the final halftoning images. Our method is testedwith various types of images. Simulation results show that the proposed algorithmpreserves fine visually sensitive texture details with a high processing speed. In addition,mean structural similarity index measurement(MSSIM) for the proposed method islarger than that for the existed algorithms, revealing that halftoning images producedwith our method are more similar to the original images.