Color halftoning with M-lattice

This paper presents new results in the halftoning of color images. The first automatic generation of the Wall Street Journal type halftones of gray-scale images was done using the M-lattice system and reported in Sherstinsky and Picard [1994]. The M-lattice system was derived from the reaction-diffusion model, first proposed by Turing [1952] in order to explain mammal coat patterns. The M-lattice is a non-linear dynamical system that is well-suited for a variety of applications formulated as constrained non-linear optimization. In particular, it can perform image processing operations that emphasize oriented patterns. The present study uses this property to extend the special-effects halftoning of gray-scale images to that of color images. The duality metric comes from the directionality information extracted by steerable filters from the gray-scale version of the original color image. The binary requirement is stated as an explicit constraint, and all three (red, green, and blue) halftone components are synthesized simultaneously by the M-lattice