Improved Error Diffusion Algorithm Incorporating a Visual Model

This paper describes an improved error diffusion a rithm for the purpose of digitally halftoning images. one variation of the algorithm an error signal is cal lated by the difference between a visually perceived put value and a visually perceived output value. Thi accomplished by applying a causal visual blur func to both the input and output images. This approach the advantage that it minimizes the appearance of w artifacts in the output image while simultaneously eli nating the edge artifacts associated with a previou sual error diffusion algorithm. In a second variation the improved error diffusion algorithm, a local ima activity detector is used to adaptively modify the in and output blur filters. This allows the error diffusi parameters to be optimized for different types of im content.