High-throughput HW-SW implementation for MV-HEVC decoder

The multi-view HEVC extension was completed in July 2014 by the Motion Picture Experts Group and the Video Coding Experts Group. Multi-view video based on stereo representation has become more and more popular. In addition, a variety of multimedia content can now be provided for mobile devices. Therefore, there is a need for a real-time multi-view video decoder. Obviously, because of the complexity of software-based methods, high-resolution MV-HEVC video cannot be decompressed in real time. Especially in mobile devices, there is no hardware that can support MV-HEVC decoding today. Hence, in this paper, we present a new parallel architecture for MV-HEVC and use FPGA to accelerate complex operations. The experimental results show that the proposed multi-view video coding can decompress 3 view video of 1920 × 1080 resolution in real time on the ZYNQ platform.