Taming Reversible Halftoning via Predictive Luminance

TL;DR

This work tackles the challenge of recovering color from binary halftones by designing reversible halftoning via predictive luminance. It introduces a three-component network (encoder, predictor, decoder) augmented with a Noise Incentive Block (NIB) and a binary gate to produce high-quality blue-noise halftones while preserving restoration to the original color image $I_c$. A predictor module recovers luminance $O_c^l$ from the halftone $O_h$, freeing the encoder to focus on chrominance, and a multi-stage training strategy with a guiding loss stabilizes optimization. Quantitative results on VOC2012 show competitive halftone quality and restoration accuracy, a favorable blue-noise profile, and insightful data-embedding analyses indicating reduced encoded information with the predictor-embedded approach. The proposed framework offers a practical path toward reversible halftoning with controllable perceptual quality and robust color restoration in real-world imaging.

Abstract

Traditional halftoning usually drops colors when dithering images with binary dots, which makes it difficult to recover the original color information. We proposed a novel halftoning technique that converts a color image into a binary halftone with full restorability to its original version. Our novel base halftoning technique consists of two convolutional neural networks (CNNs) to produce the reversible halftone patterns, and a noise incentive block (NIB) to mitigate the flatness degradation issue of CNNs. Furthermore, to tackle the conflicts between the blue-noise quality and restoration accuracy in our novel base method, we proposed a predictor-embedded approach to offload predictable information from the network, which in our case is the luminance information resembling from the halftone pattern. Such an approach allows the network to gain more flexibility to produce halftones with better blue-noise quality without compromising the restoration quality. Detailed studies on the multiple-stage training method and loss weightings have been conducted. We have compared our predictor-embedded method and our novel method regarding spectrum analysis on halftone, halftone accuracy, restoration accuracy, and the data embedding studies. Our entropy evaluation evidences our halftone contains less encoding information than our novel base method. The experiments show our predictor-embedded method gains more flexibility to improve the blue-noise quality of halftones and maintains a comparable restoration quality with a higher tolerance for disturbances.

Figures (22)

• Figure 1: Observation: the halftone variants of (a) (b) (c) present similar visual quality but with different binary patterns, as the overlaid RGB image visualized in (d). It shows the possibility of modulating the patterns for additional usage.
• Figure 2: Overview of our network architecture with embedded luminance predictor. $\oplus$ denotes the concatenation operation.
• Figure 3: Overview of the predictor architecture xia2018deep. $\oplus$ denotes the addition operation.
• Figure 4: Halftone generated by models trained with and without guidance loss in different stages. (a) Error diffusion; (b) warm-up training for 130 epochs w/o guidance loss; (c) warm-up training for 28 epochs with guidance loss; (d) our stage two w/o guidance loss; (e) our stage two with guidance loss.
• Figure 5: Qualitative comparison of halftone images with intensive structures. (a) Ostromoukhov; (b) Structure-aware; and (c) Ours.