High payload H.266/Versatile Video Coding (VVC) steganography based on affine motion estimation and chaotic maps

Abstract

The widespread data breaches across the internet, along with the vast usage of high-definition and ultra-high-definition video applications, have prompted institutions to focus on developing steganography techniques for the next-generation video codec standard, H.266/Versatile Video Coding (VVC). This focus is attributed to VVC's superior technological capabilities, which offer a 50 % reduction in bitrate while allowing for a greater data embedding capacity. All while preserving comparable visual quality to its predecessor, H.265/High Efficiency Video Coding (HEVC). This paper proposes a video steganographic technique for the VVC standard that utilizes Motion Vectors (MVs) based on Affine Motion Estimation (AME) as confidential information carriers, where four novel chaotic maps are adopted to ensure a high level of system security. The paper’s contributions include a large payload approach based on MVs for the H.266 standard, along with the use of four novel chaotic maps to enhance system security. Simulation results proved that the proposed technique achieves an average hiding capacity of 73.760 Kbits per frame, significantly surpassing state-of-the-art methods accomplished in 1920 × 1080 video resolution with a difference Peak Signal to Noise Ratio (ΔPSNR) of -0.097 dB, a Bit Rate Increase (BRI) of 4.120 %, and at a Quantization Parameter (QP) of 32. Furthermore, the integration of four highly complex chaotic maps, featuring a large key space size of 2⁷⁹⁷, elevates the security of the suggested technique, rendering it nearly unbreakable against all types of brute-force cyberattacks.