Welcome to Francis Academic Press

Academic Journal of Computing & Information Science, 2024, 7(2); doi: 10.25236/AJCIS.2024.070203.

A scene-text feature enhanced halftone method based on invertible neural network

Author(s)

Zhang Huajian, Mu Dazhong

Corresponding Author:
Mu Dazhong
Affiliation(s)

Beijing Key Laboratory of Signal and Information Processing for High-end Printing Equipment, Beijing Institution of Graphic Communication, Beijing, China

Abstract

In this paper, we present a halftone method based on invertible neural network (INN). We use the spectral features that halftone image dithering should have as constraints during training to allow the network to learn how to generate satisfactory halftone images. We use the original grayscale image as input during training, and use the halftone image generated through traditional error diffusion methods as the reference target. Our method outputs halftone images that preserve textual information that was not preserved in the error diffusion halftone method, enhancing the direct readability of the textual content that appears in the halftone images.

Keywords

Halftone, Dithering, Invertible Neural Network, Deep Learning

Cite This Paper

Zhang Huajian, Mu Dazhong. A scene-text feature enhanced halftone method based on invertible neural network. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 2: 19-25. https://doi.org/10.25236/AJCIS.2024.070203.

References

[1] Ulichney, Robert. Digital halftoning[M]. MIT Press, 1987.

[2] Knuth DE. Digital halftones by dot diffusion[J]. ACM Transactions on Graphics (TOG). 1987, 6(4): 245-73.

[3] Chang J, Alain B, Ostromoukhov V. Structure-aware error diffusion[C]. InACM SIGGRAPH Asia 2009 papers (pp. 1-8), 2009.

[4] Floyd, R. W. An adaptive algorithm for spatial gray-scale[C]. In Proc. Soc. Inf. Disp. (Vol. 17, pp. 75-77), 1976.

[5] Pang W M, Qu Y, Wong T, et al. Structure-aware halftoning[C]. In ACM SIGGRAPH 2008 papers (pp. 1-8), 2008.

[6] Laurent Dinh, David Krueger, Yoshua Bengio, NICE: Non-linear Independent Components Estimation[C]. In ICLR workshop 2015

[7] Laurent Dinh, Jascha Sohl-Dickstein, Samy Bengio, Density estimation using Real NVP[C]. in ICLR 2017.

[8] Xiao Mingqing, et al. Invertible image rescaling[C]. In Computer Vision–ECCV 2020: 16th European Conference 2020, Proceedings, Part I 16.

[9] R Zhao, T Liu, J Xiao, et al. Invertible Image Decolorization[J]. IEEE Transactions on Image Processing, 2021, vol. 30, pp. 6081-6095.

[10] Zhang Huajian, Mu Dazhong and Cao Peng, An Inverse Halftoning Method Using Invertible Neural Network[C]. 2023 8th International Conference on Intelligent Computing and Signal Processing, 2023.

[11] Y. Xing, Z. Qian and Q. Chen, “Invertible Image Signal Processing[C]. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021.

[12] Xia Menghan, et al. Deep halftoning with reversible binary pattern[C]. Proceedings of the IEEE/CVF International Conference on Computer Vision. 2021.

[13] E Agustsson and R Timofte. Ntire 2017 challenge on single image super-resolution: Dataset and study[C]. 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2017.