A new algorithm that restores original continuous tone images from digital halftone dot images is proposed. The algorithm enables moiré free color proofing, even when the page image contains halftone pictures with unknown parameters – halftone dot period, direction, etc., or dot patterns are not regularly repeated. The proposed algorithm applies discrete Fourier transformation to each local area of an image, analyzes the halftone dot period, direction and offset, and restores the original continuous tone image. This local image analysis enables restoration of a page that consists of images screened using various directions and periods. Despite the use of local area division, the restored image does not suffer from block noise.
Johji Tajima, Masato Tsukada, "Restoration of Continuous Tone Images from Digital Color Halftone Images Based on Fourier Analysis of Local Areas" in Proc. IS&T CGIV 2004 Second European Conf. on Colour in Graphics, Imaging, and Vision, 2004, pp 525 - 530, https://doi.org/10.2352/CGIV.2004.2.1.art00105