Digital Halftoning addresses the problem of developing algorithms that best match the specific parameters of any target display device. It is the first significant study of the process of producing quality images on practical computer displays. Along with text and graphics, images are fast becoming a generic data type for general-purpose computer systems. This poses new problems for the systems designer, who must be able to preprocess digital image data for a wide variety of video and hard copy displays. Digital halftoning, the method by which the illusion of continuous-tone images are created through the arrangement of binary picture elements, is a key component of any preprocessing. Digital Halftoning addresses the problem of developing algorithms that best match the specific parameters of any target display device. It is the first significant study of the process of producing quality images on practical computer displays. To aid the systems designer, Ulichney devises the concept of blue noise—which has many desirable properties for halftoning—and suggests efficient algorithms for its use. He also introduces new metrics for analyzing the frequency content of periodic and aperiodic patterns for both rectangular and hexagonal grids, and presents a unique "aspect ratio immunity" argument in favor of hexagonal grids. Included are several carefully selected digitally-produced images.

The sections in this article are White Noise Clustered Dot Recursive Tessellation Blue Noise Rendering Systems

No abstract

No abstract

In this paper "wobulation" is presented. Wobulation is a cost-effective method of increasing the resolution of digital projection systems without changing the spatial light modulator (SLM). Multiple sub-frames of data are generated from each frame of image data. An optical image shifting mechanism displaces the projected image of each sub-frame by a fraction of a pixel. The sub-frames are projected in rapid succession, thereby appearing as if they were projected simultaneously and superimposed. The resulting image has significantly higher resolution than images produced by unwobulated systems having the same number of SLM pixels. (a) unwobulated (b) 2-position wobulated, vertical shift Figure 1. Digital photos of projected images.