High Efficiency Video Coding (HEVC) Multiview Video Streaming Mechanism

Pang, Yik Siang (2022) High Efficiency Video Coding (HEVC) Multiview Video Streaming Mechanism. Masters thesis, Tunku Abdul Rahman University of Management and Technology.

Text 47 Pang Yik Siang (MCS, FT).pdf Restricted to Registered users only Download (2MB)

Abstract

Video as a tool for humans to transmit information and irreplaceable, the usage of video growth rapidly in these twenty years. In order to improve the authenticity and viewing experience of video, the development of video technology is focusing on increasing the resolution and multiple view. The latest video codec is HEVC (high efficiency video coding) and also named H.265 which are developed by Joint Collaborative Team on Video Coding (JCT-VC) and finalised in 2013. HEVC codec shows a 50% bitrate reduction for UHD video compared to previous video codec (AVC/H.264). The future demand of multi-view video (MV) and 3D video increases rapidly, they formed a development group (JCT-VC) specially researching new video coding extensions for multi-view and 3D video which named MV-HEVC and 3D-HEVC extension. In this research mainly focus on multi-view video coding Multi-view High Efficiency Video Coding (MV-HEVC) which is a stereoscopic video compression standard that encodes video sequences efficiently from simultaneously capturing multiple camera views to a single video stream. The presentation will base on two significant tasks for the research on Multiview video coding extension of HEVC standard, referred to as MV-HEVC. A novel method that performs multi-view video compression and decompression processes, tested on a Linux-based operating system to achieve a near real-time video streaming mechanism. The processes of MV-HEVC with two views in real-time streaming are evaluated and compared with pre-recorded video streaming. The experimental results indicate a minor change in Bjøntegaard Delta (BD) Bitrate and Peak Signal-to-Noise Ratio (PSNR) scales with an average of 0.475% bandwidth reduction and 0.018% improvement in video quality. Besides, the proposed method is compared to multi-view simulcast HEVC and shows 1.444% BD-rate saving and 0.078 BD-PSNR gain. Then, a data embedding solution is introduced as an application on MV-HEVC code using four video test sequences with different image resolution and camera arrangement utilised in this project. The performance of data embedded results is validated and compared to original HTM software encoded results. The results show that no significant changes of bitrate and video quality on data embedded video sequences measured by using Bjøntegaard metric.