Design and analysis methods for visual-model based digital halftoning

Digital halftoning techniques are used to render continuous-tone images on binary display devices. An important class of digital halftoning techniques are the ones that convert a gray level in an image into a binary pattern. The important aspects of the design of such patterns are to reduce visible artifacts as well as to render the image with high fidelity. An optimization approach to the design of dither matrices used in the dispersed-dot digital halftoning method is described. The design approach is based on shaping the spectrum of the dithering signal according to a model of the human visual system. Next, the dither pattern ensemble (DPE) algorithm, an approach developed for binary image rendering using the designed patterns, is described. By incorporating a data preprocessing step, the moment preserving quantization, into the original DPE algorithm, the grainy appearance of the output image due to pattern-to-pattern transitions of the individually optimized dither patterns is reduced.; Besides the conventional Fourier domain analysis, a new analysis method in the spatial domain of halftone images based on the Voronoi tessellation is presented. A statistical analysis of the geometric structure of dot patterns is used to evaluate their performance. The proposed analysis method has the advantage over the Fourier domain analysis in that it can be used to evaluate halftone patterns representing more than one gray level. The inverse halftoning problem is formulated, and an algorithm for solving such a problem by means of the Markov random field model-based surface reconstruction is presented. With the availability of approximately 37% of the original data, the proposed algorithm is capable of reconstructing the image intensity surface which correlates well with the original data.