Deep Steganography with U-Net: Hiding and Revealing Text in Video

Yıl 2026, Cilt: 5 Sayı: 1, 345 - 362, 28.02.2026

Mahmut Sinecen

https://doi.org/10.62520/fujece.1780558

https://izlik.org/JA84MG47PB

Öz

Video-based steganography has attracted increasing attention due to its high payload capacity and improved imperceptibility compared to image-based approaches. In this study, a deep learning–based steganographic framework is proposed to embed and recover textual information within video content using the U-Net architecture. Unlike traditional least significant bit (LSB)–based techniques, the proposed method utilizes region-of-interest (ROI) selection and patch-based embedding to enhance robustness and visual quality. Textual data are first encoded into image patches and embedded into selected regions of video frames via a trained hiding network. A corresponding revealing network is employed to recover the hidden information, followed by an optical character recognition (OCR) pipeline for text extraction. Experimental results demonstrate character recovery accuracies between 81% and 88% while preserving high visual fidelity in the stego videos. This ROI-guided U-Net framework provides an effective and scalable solution for secure and imperceptible text hiding in video streams.

Anahtar Kelimeler

Deep steganography , U-Net , Video hiding , Text recovery , OCR

Etik Beyan

Ethics committee approval is not required for this study. “There is no conflict of interest with any person/institution in the prepared article.”

U-Net ile Derin Steganografi: Video içerisine Metni Gizleme ve Yeniden Elde Etme

https://izlik.org/JA84MG47PB

Öz

Video tabanlı steganografi, görüntü tabanlı yaklaşımlara kıyasla sunduğu yüksek veri taşıma kapasitesi ve gelişmiş görsel algılanamazlık özellikleri nedeniyle son yıllarda artan bir ilgi görmektedir. Bu çalışmada, U-Net mimarisi kullanılarak video içeriği içerisine metinsel bilgilerin gizlenmesi ve geri elde edilmesi amacıyla derin öğrenmeye dayalı bir steganografik yöntem önerilmektedir. Geleneksel en düşük anlamlı bit (LSB) tabanlı tekniklerden farklı olarak, önerilen yaklaşım; dayanıklılığı ve görsel kaliteyi artırmak amacıyla ilgi alanı (ROI) seçimi ve yama (patch) tabanlı yerleştirme stratejilerini kullanmaktadır. Metinsel veriler öncelikle görüntü tabanlı yamalara dönüştürülmekte ve eğitilmiş bir gizleme ağı aracılığıyla video karelerinin seçilen bölgelerine gömülmektedir. Gizlenen bilginin geri elde edilmesi için karşılık gelen bir çıkarım ağı kullanılmakta ve ardından metnin çıkarımı optik karakter tanıma (OCR) yöntemi ile gerçekleştirilmektedir. Deneysel sonuçlar, stego videolarda yüksek görsel bütünlük korunurken karakter geri kazanım doğruluğunun %81 ile %88 arasında değiştiğini göstermektedir. Önerilen ROI güdümlü U-Net tabanlı çerçeve, video akışlarında güvenli ve algılanamaz metin gizleme için etkili ve ölçeklenebilir bir çözüm sunmaktadır.

Anahtar Kelimeler

Derin stenografi , U-Net , Video gizleme , Metin çıkarımı , OCR

Etik Beyan

Bu çalışma için etik kurul onayı gerekmemektedir. "Hazırlanan makalede herhangi bir kişi/kurumla çıkar çatışması bulunmamaktadır."

Kaynakça

• M. A. Idakwo, M. B. Muazu, A. E. Adedokun and B. O. Sadiq, “An extensive survey of digital image steganography: State of the art,” arXiv:2404.19548, 2024.
• I. I. Araujo and H. Kazemian, “Vulnerability exploitations using steganography in PDF files,” Int. J. Comput. Netw. Appl., vol. 7, no. 1, p. 10, 2020.
• F. Nabi and M. M. Afzal, “Image steganography: Critical findings through some novel techniques,” Int. J. Innov. Technol. Explor. Eng., vol. 9, no. 5, pp. 878–890, 2020.
• H. Li et al., “Smaller is bigger: Rethinking the embedding rate of deep hiding,” 2023.
• K. Koptyra and M. R. Ogiela, “Distributed steganography in PDF files—Secrets hidden in modified pages,” Entropy, vol. 22, no. 6, p. 600, 2020.
• A. D. Wiranata and R. T. Aldisa, “Aplikasi steganografi menggunakan least significant bit (LSB) dengan enkripsi Caesar Chipper dan Rivest Code 4 (RC4) menggunakan bahasa pemrograman JAVA,” J. JTIK (J. Teknol. Inf. dan Komun.), vol. 5, no. 3, pp. 277–281, 2021.
• M. A. Ali Khodher and T. W. Aldeen Khairi, “Review: A comparison steganography between texts and images,” J. Phys.: Conf. Ser., vol. 1591, no. 1, p. 012024, 2020.
• J. Kose, O. B. Chia and V. Baboolal, “Review and test of steganography techniques,” 2020.
• Q. Li et al., “A novel grayscale image steganography scheme based on chaos encryption and generative adversarial networks,” IEEE Access, vol. 8, pp. 168166–168176, 2020.
• P. Wei, Q. Zhou, Z. Wang, Z. Qian, X. Zhang and S. Li, “Generative steganography diffusion,” 2023.
• A. K. Sahu and M. Sahu, “Digital image steganography and steganalysis: A journey of the past three decades,” Open Comput. Sci., vol. 10, no. 1, pp. 296–342, 2020.
• J. Fridrich, Steganography in Digital Media: Principles, Algorithms, and Applications. Cambridge, U.K.: Cambridge Univ. Press, 2012.
• L. Zeng, W. Lu, W. Liu and J. Chen, “Deep residual network for halftone image steganalysis with stego-signal diffusion,” Signal Process., vol. 172, p. 107576, 2020.
• J. Wang et al., “Fighting malicious media data: A survey on tampering detection and deepfake detection,” 2022.
• S. Rahman et al., “A novel and efficient digital image steganography technique using least significant bit substitution,” Sci. Rep., vol. 15, no. 1, p. 107, 2025.
• B. Chen, W. Luo, P. Zheng and J. Huang, “Universal stego post-processing for enhancing image steganography,” J. Inf. Secur. Appl., vol. 55, p. 102664, 2020.
• D. Volkhonskiy, I. Nazarov and E. Burnaev, “Steganographic generative adversarial networks,” in Proc. 12th Int. Conf. Mach. Vis. (ICMV 2019), SPIE, 2020, p. 97.
• J. Liu et al., “Recent advances of image steganography with generative adversarial networks,” IEEE Access, vol. 8, pp. 60575–60597, 2020.
• S. Rahman et al., “A comprehensive study of digital image steganographic techniques,” IEEE Access, vol. 11, pp. 6770–6791, 2023.
• R. Chaganti, V. Ravi, M. Alazab and T. D. Pham, “Stegomalware: A systematic survey of malware hiding and detection in images, machine learning models and research challenges,” 2021.
• R. Apau, M. Asante, F. Twum, J. Ben Hayfron-Acquah and K. O. Peasah, “Image steganography techniques for resisting statistical steganalysis attacks: A systematic literature review,” PLoS One, vol. 19, no. 9, p. e0308807, 2024.
• Y. Sanjalawe, S. Al-E’mari, S. Fraihat, M. Abualhaj and E. Alzubi, “A deep learning-driven multi-layered steganographic approach for enhanced data security,” Sci. Rep., vol. 15, no. 1, p. 4761, 2025.
• Z. Zhou, Y. Su, Q. M. J. Wu, Z. Fu and Y. Shi, “Secret-to-image reversible transformation for generative steganography,” 2022.
• M. Shukor, B. B. Damodaran, X. Yao and P. Hellier, “Video coding using learned latent GAN compression,” in Proc. 30th ACM Int. Conf. Multimedia, New York, NY, USA: ACM, 2022, pp. 2239–2248.
• R. C. Gonzalez and R. E. Woods, Digital Image Processing, 4th ed. New York, NY, USA: Pearson, 2018.
• I. Sobel and G. Feldman, “A 3×3 isotropic gradient operator for image processing,” 1968.
• scikit-image: Image processing in Python. [Online]. Available: https://scikit-image.org/
• OpenCV documentation index. [Online]. Available: https://docs.opencv.org/
• O. Ronneberger, P. Fischer and T. Brox, “U-Net: Convolutional networks for biomedical image segmentation,” 2015.
• J. Hu, L. Shen and G. Sun, “Squeeze-and-excitation networks,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2018, pp. 7132–7141.
• Y. Wu and K. He, “Group normalization,” in Proc. Eur. Conf. Comput. Vis. (ECCV), 2018, pp. 3–19.
• P. Ramachandran, B. Zoph and Q. V. Le, “Searching for activation functions,” in Proc. Int. Conf. Learn. Represent. (ICLR), Workshop Track, 2018.
• J. Fridrich and J. Kodovsky, “Rich models for steganalysis of digital images,” IEEE Trans. Inf. Forensics Security, vol. 7, no. 3, pp. 868–882, 2012.
• Z. Wang, A. C. Bovik, H. R. Sheikh and E. P. Simoncelli, “Image quality assessment: From error visibility to structural similarity,” IEEE Trans. Image Process., vol. 13, no. 4, pp. 600–612, 2004.
• S. Baluja, “Hiding images in plain sight: Deep steganography,” in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS), 2017, pp. 1–11.
• I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” 2019.
• ReduceLROnPlateau — PyTorch documentation. [Online]. Available: https://docs.pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html
• ReduceLROnPlateau — Keras API documentation. [Online]. Available: https://keras.io/api/callbacks/reduce_lr_on_plateau/
• I. Goodfellow, Y. Bengio and A. Courville, “Optimization for training deep models,” in Deep Learning. Cambridge, MA, USA: MIT Press, 2016, pp. 271–325.