Abstract

Information hiding in videos plays a critical function in several areas, such as video authentication and content augmentation. Although information hiding in video has been extensively studied, research on combining it with the latest video compression standard, H.266/Versatile Video Coding (VVC), has received limited attention. This study proposes new techniques for embedding information in compressed video streams using H.266/VVC. Recent advances in transform coding techniques have led to the adoption of advanced tools in VVC. This research aims at identifying special permutations within the VVC framework tailored towards hiding information by focussing on Selective Quantisation Techniques (SQT), Selective Sub-Block Transformation Techniques (SSBTT), and Conjoined Techniques (CT) for embedding different types of information in video streams. The merit of each technique is analysed while considering the context of files of various sizes, from small image and audio files to large, combined data formats. Our results are summarised in comprehensive performance metrics, including achieved bit rate, Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Bjøntegaard Delta PSNR (BD-PSNR), and Bjøntegaard Delta Bit Rate (BD-BR), which indicate that these techniques are effective in preserving video quality while embedding data. SQT are particularly effective for smaller files, while SSBTT and CT are effective in handling larger merged files. This study supports the applicability of VVC-based information hiding techniques in secure communications and digital rights management, as well as its capability of inserting several data formats into the video stream while maintaining video integrity. 

Research Objectives

Research Scopes

• The research determines how data can be embedded in the video using a VVC encoder and how the hidden information can be decoded using the VVC decoder without affecting the video’s decodability.

• The research analyses the effect of the alteration of the indices in the transform stage of the VVC encoder on video quality and compression efficiency. Several measures assess the impact of changes in the two factors, which are video quality loss and embedding effectiveness, measured as MSE, PSNR, SSIM, BD-PSNR, BD-BR, hiding capacity, and average hiding capacity per frame.

• The research develops an application with the information hiding techniques that are specifically used to embed different types of data into the video streams using VVC codec. The study assessed the practicability and flexibility of the proposed information hiding techniques in VVC codec.

Information Hiding Application Flow

Information Hiding Encoding / Decoding Design Architecture

Test Video Sequences

Resources

User Application

Publications

1. Joan, H., Tan, L. P., & Tew, Y. (2023a). Data Embedding with Dynamic Slice Quantisation in Versatile Video Coding. 2023 IEEE International Conference on Computing, ICOCO 2023, 41–46. https://doi.org/10.1109/ICOCO59262.2023.10397877

2. Joan, H., Tan, L. P., & Tew, Y. (2023b). Quick Response (QR) codes embedding in VVC using Quantisation Parameter Manipulation. 2023 Asia Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2023, 1697–1704. https://doi.org/10.1109/APSIPAASC58517.2023.10317417

3. Joan, H., Tew, Y., & Tan, L. P. (2024a). A Novel Information Hiding Approach using Selective Quantization Technique in Video Coding. 2024 The 4th International Conference on Computer, Information Technology and Intelligent Computing, CITIC 2024. (Accepted in International Journal on Informatics Visualization Journal (JOIV))

4. Joan, H., Tew, Y., & Tan, L. P. (2024b). Innovative Information Hiding in H.266/VVC Using Sub-Block Transform Technique. 2024 Asia Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2024 (Accepted on 28 Sept 2024)

Project Team

Acknowledgement

We would like to extend our sincere gratitude to everyone who has supported me throughout this research. We are deeply appreciative of the assistance provided by Tunku Abdul Rahman University of Management and Technology (TAR UMT). The university's valuable resources, including its library, online databases, and laboratory facilities, were instrumental in the completion of this project.