基于机器学习的HEVC快速帧内预测算法研究进展  

Advances in research of HEVC fast intra-frame prediction algorithm based on machine learning

作　　者：艾达[1] 卢雪磊 高阳 董久军 AI Da;LU Xuelei;GAO Yang;DONG Jiujun(The Key Laboratory of Electronic Information Site-Survey Application Technology of The Ministry of Public Security,Xi’an University of Posts and Telecommunications,Xi’an 710121,China;School of Communications and Information Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)

机构地区：[1]西安邮电大学公安部电子信息勘查应用技术重点实验室,陕西西安710121 [2]西安邮电大学通信与信息工程学院,陕西西安710121

出　　处：《现代电子技术》2018年第18期178-181,186,共5页Modern Electronics Technique

基　　金：国家自然科学基金资助项目(61202183);公安部科技强警专项资助项目(2016GABJC51)~~

摘　　要：高效率视频编码(HEVC)存在帧内预测计算复杂度过高的情况,而机器学习方法在HEVC快速帧内预测算法的研究取得了一定的进展。对近年来常用的机器学习算法在帧内预测快速深度决策和模式决策两方面的应用进行详述,总结比较不同文献提出的机器学习方法在编码性能上的优劣,最后分析机器学习算法在帧内编码应用中存在的问题,为下一步研究提供参考性建议。In view of the high computational complexity of intra-frame prediction in high efficiency video coding(HEVC),and a certain progress that the machine learning method has made in research of the HEVC fast intra-frame prediction algo-rithm,the application of common machine learning algorithms in recent years to the depth decision and mode decision of fast in-tra-frame prediction is discussed in detail.The coding performance advantages and disadvantages of the machine learning meth-ods proposed in different literatures are summarized and compared.The intra-frame coding application problems of machine learning algorithm are analyzed,which can provide reference suggestions for further research.

关 键 词：高效视频编码 帧内预测 深度决策 模式决策 机器学习 计算复杂度 

分 类 号：TN919.81-34[电子电信—通信与信息系统]