The fact that most printing devices, such as laser printers, are restricted to very few colors has led to an increased interest in finding efficient techniques that transform a continuoustone image into a binary one. Such techniques are referred to as halftoning methods. The quality of the printed images is greatly dependent upon the characteristics of the halftoning method. Many classes of halftoning methods have been proposed over the years. One class represents the simple halftoning techniques, which regrettably produce halftoned images of low quality. Another class of halftoning methods, known as iterative halftoning methods, represents the more complicated and time-consuming methods that produce high quality halftoned images. The methods presented in this thesis belong to the latter category. Due to the fast increase of computer power this type of halftoning techniques is becoming more and more popular.

This thesis presents a short background of printing and halftoning and describes a few of the existing and known halftoning algorithms. This is followed by the presentation of a novel iterative halftoning technique for high quality reproduction of monochromatic images. Several criteria with regard to the quality of halftoned images are discussed and examined. The method for monochromatic images is then extended to color halftoning. A new method for color reproduction is presented that halftones the color separations of the original color image dependently. In this method the halftone dots in different separations are prevented from being placed on top of each other as much as possible. The color shifts that might occur due to this dot-off-dot strategy are avoided by transforming the original image before the halftoning process is performed. Dot gain is sometimes considered as an unwanted effect and therefore should be controlled in order to increase the quality of printed halftone images. Optical dot gain is briefly discussed and some simulations for it are carried out.