结合编码信息与超分辨率重建的视频编码方法  

A Video Coding Method Combining Encoding Information and Super-resolution Reconstruction

作　　者：贾奇才 林宏伟[1] 田子溢 王创 JIA Qicai;LIN Hongwei;TIAN Ziyi;WANG Chuang(School of Electrical Engineering,Northwest Minzu University,Lanzhou,China,730000)

机构地区：[1]西北民族大学电气工程学院,兰州730000

出　　处：《福建电脑》2023年第7期21-24,共4页Journal of Fujian Computer

基　　金：中央高校基本科研业务费专项资金《面向“双碳”布式光伏发电功率预测系统》(No.31920230141);西北民族大学一流本科课程项目2021(No.YLKC-74);西北民族大学教育教学改革研究项目(No.2021XJYBJG-12);西北民族大学创新创业教育教学改革研究项目《面向专创融合的高校创新创业校企协同育人学程制教育改革研究》、甘肃省自然科学基金(No.21JR1RA206)资助。

摘　　要：如何在现有的视频压缩编码框架下保证视频质量的同时,大大降低压缩后视频的码率,仍是一个亟待解决的问题。本文根据现有的视频编码框架高效视频编码,提出了一种结合高效视频编码编码信息与时/空域超分辨率重建的视频编码方法。本系统预先通过时间域自适应抽帧和空间域下采样分别降低了原视频的帧率和空间分辨率,然后通过标准化编解码器进行编码。在解码端,利用基于视频压缩码流信息的时域超分辨算法,将解码后的视频帧率增加到原始视频帧率。最后通过空域超分辨算法将视频重建恢复到原始分辨率。实验结果表明,相较于标准传统HEVC视频编码框架,本文所提出的算法在低码率段极大地降低了压缩视频的码率,同时降低了解码端超分辨重建过程的复杂度。How to greatly reduce the bit rate of the compressed video while ensuring the quality of the video is still an urgent problem to be solved.Based on the existing video coding framework HEVC,a video coding method combining HEVC coding information and time/spatial super resolution reconstruction is proposed in this paper.In this system,the frame rate and spatial resolution of the original video are reduced respectively by time domain adaptive frame extraction and space domain subsampling,and then encoded by standardized codec.At the decoding end,the time-domain super-resolution algorithm based on the compressed video stream information is used to increase the decoded video frame rate to the original video frame rate,and finally the video is reconstructed to the original resolution through the spatial super-resolution algorithm.The experimental results show that compared with the standard traditional HEVC video coding framework,the proposed algorithm greatly reduces the bit rate of the compressed video in the low bit rate segment,and also reduces the complexity of the decoding terminal's super resolution reconstruction process.

关 键 词：时空下采样 码流信息 新一代高效视频编码 超分辨重建 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]