New Steganography Framework Hides Data in Image Wavelet Bands

Researchers Juhi Singh and Arun Kumar Singh have introduced an adaptive hybrid framework designed to conceal information within digital images while preserving their visual quality. Published in Discover Applied Sciences, the study presents a transform-domain steganography approach that integrates discrete wavelet decomposition, block-based frequency analysis, and quantization-driven embedding. The primary objective of this work is to address the persistent challenge in data hiding systems of maintaining high visual fidelity in host images while ensuring the reliable recovery of embedded payloads.

The proposed method, referred to as the Triple-Transform Keyed Steganography or TTKS framework, embeds payload information in singular values extracted from the mid-band region of Discrete Cosine Transform blocks. These blocks are obtained from the LH subband of the wavelet domain.

The embedding process is guided by a quantization index modulation scheme, which serves as the mechanism for encoding the secret data into the image coefficients. To enhance the stability of the extraction process, the framework employs redundancy mechanisms and multi-carrier embedding.

Evaluation of the framework was conducted using the BOSSBase image dataset, which provides a diverse collection of natural images suitable for steganographic experiments. The experimental results indicate that the method achieves high Peak Signal-to-Noise Ratio values, ranging from 50 to 56 dB. Additionally, the Structural Similarity Index Measure values reached 0.998533, which suggests that the stego images exhibit high imperceptibility, meaning the embedded data is difficult to detect visually.

The robustness of the TTKS framework was further examined under JPEG compression, a common operation that can introduce perturbations in image coefficients. Under no attack conditions, the method achieved a Bit Error Rate of zero, indicating perfect extraction of the hidden payload. In scenarios involving compression-induced perturbations, the framework maintained moderate extraction reliability.

An ablation analysis was performed to study the influence of specific parameters on system performance. The researchers investigated the impact of the quantization step, repetition factor, and carrier selection on the overall effectiveness of the steganography framework. The observed trends from this analysis confirm that the proposed embedding strategy provides a balanced compromise between visual imperceptibility and extraction robustness, allowing for high-quality image preservation alongside secure data hiding.

The study concludes that the TTKS framework offers an effective solution for high-fidelity image steganography in transform-domain environments.

Subject of Research: Transform-domain steganography framework for secure image hiding

Article Title: An adaptive hybrid framework for secure image hiding using triple-transform keyed steganography

Article References: Singh, J., & Singh, A. K. (2026). An adaptive hybrid framework for secure image hiding using triple-transform keyed steganography. Discover Applied Sciences. https://doi.org/10.1007/s42452-026-09620-y

Image Credits: AI Generated

DOI: 10.1007/s42452-026-09620-y

Keywords: steganography, image hiding, discrete wavelet transform, discrete cosine transform, quantization index modulation, BOSSBase dataset, peak signal-to-noise ratio, structural similarity index, JPEG compression, bit error rate, singular values, mid-band region, LH subband, multi-carrier embedding, ablation analysis, transform-domain, data hiding, visual fidelity, extraction robustness, Discover Applied Sciences