首页> 外文会议>International Conference on Emerging Trends in Engineering, Technology and Science >A high-performance and high-precision sub-pixel motion estimator-interpolator for real-time HDTV(8K) in MPEGH/HEVC coding
【24h】

A high-performance and high-precision sub-pixel motion estimator-interpolator for real-time HDTV(8K) in MPEGH/HEVC coding

机译:用于MPEGH / HEVC编码的实时HDTV(8K)的高性能和高精度子像素运动估计器 - 内插器

获取原文

摘要

Despite increased complexity sub-pixel motion estimation and compensation significantly outperforms integer motion estimation and compensation in HEVC/MPEGH/H.265 Video Codec's, since moving objects do not necessarily move by integer pixel locations between successive video frames. Typically, fractional pixel accuracy is obtained by means of bilinear interpolation producing a spatially blurred predicted signal. The motion estimation and compensation is improved in this paper by means of the filtering effect using a very effective spatial digital low pass FIR filter, which allows the motion vectors to be determined with higher levels of precision and accuracy than existing algorithmic implementations. The fractional pixel accuracy was achieved using a total of 112 8-tap digital FIR filter for one-eighth pixel precision, which includes half and quarter pixel accuracy. The design has been implemented on a 28nm foundry process, with a speed of 1.101 GHz and it has achieved 2262 GOPS at this speed, outputting data at the rate of 1.8 Tera bits per second, for one-eighth pixel accuracy. Computational complexity, Memory & I/O Bandwidth has been reduced by inputting the Mean Square Error Map of the pixels to the Fractional Pixel Estimator and then searching in the sub-pixel grid. This design is targeted for 8K Ultra High Definition Television (UHDTV). 8K HDTV format is 8192 × 4320 pixels. This amounts to 35.4 Million Pixels per Frame. The incoming Video Pixel Rate for 8K HDTV at 60 Frames per Second (fps) is 2.12 Billion Pixels per Second or 2.12 Giga Pixels / second. This amounts to an incoming Video Data Rate of 51 Billion bits per second or 51 Gbps. At 120 frames per second the Incoming Video Pixel rate is 4.24 Billion Pixels per Second or 4.24 Giga Pixels per Second. This amounts to incoming Video Data Rate of 102 Gbps. For Quarter Pixel Motion Estimation, we are adding 3 Sub-Pixels for every Integer Pixel. Pixel Count increases by a factor of 16. For 4K HDTV this becomes 142 Million Pixels per Frame. At 60fps, the Pixel rate is 8.5 Billion Pixels per Second or 8.5 Giga Pixels per Second, with a Video Processing Data Rate of 195 Gbps. At 120fps the Pixel rate 17 Billion Pixels per Second or 17 Giga Pixels / Second, with a Video Processing Data Rate of 390 Gbps. For 1/8th Pixel Motion Estimation, we are adding 7 Sub-Pixels for every Integer Pixel. Pixel Count increases by a factor of 64. For 8K HDTV this becomes 570 Million Pixels per Frame. At 60fps, the Pixel rate is 34 Billion Pixels per Second or 34 Giga Pixels / Second, Video Processing Data Rate of 780 Gbps. At 120fps the Pixel rate 70 Billion Pixels per Second or 70 Giga Pixels / Second, Video Processing Data Rate of 1560 Gbps / 1.56 Tbps.
机译:尽管复杂性较高的子像素运动估计和补偿显着优于HEVC / MPEGH / H.265视频编解码器中的整数运动估计和补偿,因为移动对象不一定由连续的视频帧之间的整数像素位置移动。通常,通过产生空间模糊的预测信号的双线性插值获得分数像素精度。本文通过使用非常有效的空间数字低通冷滤波器的滤波效果在本文中提高了运动估计和补偿,这允许运动矢量以比现有算法实现更高的精度和精度确定。使用总共112个8抽头数字FIR滤波器实现分数像素精度,用于一八个像素精度,包括一半和四分之一像素精度。该设计已经在28nm铸造过程中实现,速度为1.101 GHz,它已经实现了2262个速度,以每秒1.8 Tera位的速率输出数据,以实现一八个像素精度。通过将像素的平方误差映射输入到分数像素估计器,然后在子像素网格中搜索,通过计算计算复杂性,存储器和I / O带宽已经减少。该设计是针对8k超高清电视(UHDTV)的目标。 8K HDTV格式为8192×4320像素。该框架数量为3540万像素。 8K HDTV的传入视频像素速率为每秒60帧(FPS)为每秒21.12亿像素或2.12千兆像素/秒。这增加了每秒51亿比特或51 Gbps的传入视频数据速率。在每秒120帧时,进入的视频像素率为4.24亿像素每秒或每秒4.24千兆像素。这增加了102 Gbps的视频数据速率。对于四分之一像素运动估计,我们正在为每个整数像素添加3个子像素。像素计数增加了16倍。对于4K HDTV,每帧变为1.42亿像素。在60fps时,像素速率为每秒85亿像素,每秒8.5个千兆像素,视频处理数据速率为195 Gbps。在120fps,像素率为每秒17亿像素或17个千兆像素/秒,视频处理数据速率为390 Gbps。对于1/8像素运动估计,我们为每个整数像素添加7个子像素。像素计数增加了64倍。对于8K HDTV,每帧变为5.7亿像素。在60fps时,像素率为每秒34亿像素或34个千兆像素/秒,视频处理数据速率为780 Gbps。在120fps时,像素率每秒700亿像素或70千兆像素/秒,视频处理数据速率为1560 Gbps / 1.56 Tbps。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号