Video compression LSI technologies for ultra-high definition video

本研究的重点是超高分辨率视频的低复杂度视频压缩算法和硬件架构，这是创建下一代高真实度视频通信和广播服务的关键。经过三年的努力，我们提出了许多低复杂度的视频压缩算法和大规模集成电路结构，用于实现视频压缩的基本功能（如运动估计、自适应插值滤波器和平移/量化）。此外，我们提出了低复杂度帧内预测和SAO（采样自适应偏移）算法，它们是2013年1月标准化的HEVC（高效视频编码）中的新功能。我们亦促进与视像系统、通讯及广播业的合作。现在，基于4K和超级HDTV的新视频广播服务即将出现。

This research focused on low-complexity video compression algorithms and hardware architectures for ultra-high-resolution video which is a key for creating the next generation high-reality video communication and broadcasting services. Through three years activities, we proposed many low-complexity algorithms and LSI architectures for basic functions of video compression (e.g. a motion estimation, an adaptive interpolation filter and a translation/quantization). Furthermore, we proposed low-complexity intra prediction and SAO (Sample Adaptive Offset) algorithms which are new functions in HEVC (High Efficiency Video Coding) standardized on Jan. 2013. We also promoted collaboration with video system, communication and broadcasting industries. Now, new video broadcasting services based on 4K and super HDTV are going to appear.