.png){kind=icon}
.png){kind=icon}
Return
Secure HEVC video steganography using IPMs spatial distribution and transfer probability
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-026-21249-y.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
High Efficiency Video Coding (H.265/HEVC) offers a wide range of Intra-Prediction Modes (IPMs), presenting a valuable opportunity to increase embedding capacity in IPM-based video steganography. A central challenge in this domain lies in establishing a robust and secure mapping between secret data and the available IPMs. However, existing IPM-based steganographic methods for HEVC often overlook two fundamental characteristics of HEVC streams: the spatial correlation of IPMs and the transfer probabilities of IPMs during re-compression. Neglecting these aspects compromises the methods’ resilience, rendering them vulnerable to both statistical and calibration-based steganalysis. In this study, we propose a secure and efficient steganographic scheme that explicitly incorporates both the spatial correlation and transfer dynamics of IPMs to enhance resistance against steganalytic attacks. We conduct a comprehensive analysis of the distribution patterns and spatial correlations of IPMs in HEVC streams, which informs a cover elements selection strategy designed to minimize detectable anomalies. Furthermore, we introduce a mapping mechanism that anticipates IPMs alterations caused by re-compression, thereby preserving the statistical integrity of the IPMs distribution. Extensive experimental evaluations demonstrate that the proposed approach outperforms several state-of-the-art techniques across key performance dimensions, including anti-steganalysis resilience, coding efficiency, perceptual quality, and computational complexity.
Keywords:
HEVC steganography
IPMs
Spatial correlation
Transfer probability
Journal
IF:
3
Papers:
1.9W
Citations:
3.2W