StegoStack: a layered hybrid encoding framework for secure and high-fidelity image steganography
Year: 2026
Type: article
Abstract: Image steganography plays a critical role in secure digital communication by concealing sensitive information within multimedia content while maintaining visual imperceptibility. This paper presents StegoStack, a layered hybrid image steganography framework that integrates modern encoding techniques, historical cipher transformations, columnar transposition, and 2-bit RGB least significant bit embedding to enhance information security and concealment. Unlike conventional approaches that employ a limited set of encoding mechanisms, the proposed framework enables systematic evaluation and integration of diverse modern and historical encoding strategies within a unified architecture. A comprehensive experimental study involving 22 encoding methods was conducted across 6 real world application scenarios, including military communication, medical record transmission, digital watermarking, journalism, forensic evidence exchange, and gaming communication. The framework was evaluated using 17 performance metrics related to imperceptibility, payload efficiency, robustness, entropy, and computational performance. Resistance to automated detection was assessed using machine learning and deep learning steganalysis models on a dataset comprising 1,080 cover and stego images. Additional validation was performed through robustness analysis, ablation studies, scalability evaluation, and statistical significance testing. Experimental results demonstrate high visual fidelity, effective information concealment, and promising resistance against the evaluated steganalysis models. The findings further confirm the contribution of individual framework components and the scalability of the proposed approach across varying payload sizes. The proposed StegoStack framework provides a flexible and extensible solution for secure image steganography in diverse application domains.
Source: Scientific Reports
Authors: Aayushi Jeeban Tripathy, C. R. Kavitha
Institution: Amrita Vishwa Vidyapeetham
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7225777712
FWCI:
Cites:
Cited by:
Related to: