An Optimization of Digital Halftoning Using Content Sensitive Adjustment

A method to generate optimal halftone images using content sensitive adjustment is proposed. There are two steps to implement the algorithm. First, adjust intensity of each gray level to be equal between initial halftone image and con-tone image. Secondly, use direct binary search (DBS) to minimize the difference between con-tone image and halftone image as perceived by human eyes. After a number of iterations, an optimization halftone image can be achieved. The algorithm differs from that in conventional DBS by exploiting the equality of the gray value locally in the corresponding region between con-tone and halftone image allowing the rendition to preserve the appearance perceived similarly. By local processing, the algorithm can improve the quality of the halftone image and decrease the computation cost in frequency weighted mean square error. Besides, it can also reduce the visual artifact effect of the initial halftone image's quality on the manipulated halftone image.