Abstract

Halftoning techniques are applied to convert graytone images into binary distributions. The optimization problem consists of achieving an acceptable compromise between graytone and detail resolution. By using iterative techniques it is possible to observe constraints in both the image and its spectrum. The possibility of controlling the spectrum during the binarization is used to influence the quantization noise. It will be shown how manipulation of the amplitude and phase in the spectrum can be used to enhance details and edges of the image, i.e., to conform the binarization noise to characteristics of the graytone image.

© 1992 Optical Society of America