ABSTRACT
Halftoning via green-noise masks is a process identical to halftoning via blue-noise masks with the exception that given a continuoustone, monochrome image of constant gray-level g, the resulting dither pattern has green-noise characteristics appropriate to g. The physical construction of binary dither patterns for the green-noise mask is done through BIPPCCA, the binary pattern pair correlation construction algorithm. The basic premise of BIPPCCA is to take an empty array and assign, to each element, a probability of that element becoming a minority pixel. One important aspect to designing an optimal green-noise mask is the cluster size required to reduce the spatial variation in tone below a pre-specified tolerance. BIPPCCA’s advantage over error-diffusion with output-dependent feedback is, therefore, that it can create clusters of virtually unlimited size, allowing BIPPCCA to create the 36-pixel cluster capable of maintaining a total variance across the page below 0.002 at any gray-level.
