论文部分内容阅读
科学技术与生产力的发展带来了数据量的高速增长,其中视频图像等多媒体数据占了很大的比重。如何高效处理这些海量数据并从中快速挖掘有价值的信息是当前的研究热点。通常大数据具有四个特点,即数据量大、需要快速响应、数据类型多样和价值密度低。视频大数据同样具有以上特点,但其特殊性在于数据冗余更大,需要进行高效的压缩编码与分析处理。总的来说,视频大数据的研究内容包括了视频数据表示、智能视频分析、视频压缩与传输、视频显示与评价等方面。在发展趋势上,视频数据的表示将向真实感与智能化两个方向发展;智能视频分析技术将会借助深度神经网络获得更准确的识别分类结果;视频压缩技术在提升压缩效率的同时也会探索降低编码复杂度的方法,并通过结合人眼视觉感知特性的编码算法来减少视频大数据的视觉冗余;视频显示设备将伴随着视频数据表示形式的改变而进行相应的升级换代;视频质量的评价准则将由单一的图像质量评价向更加综合全面的用户体验质量评价发展。
The development of science and technology and productivity have brought the rapid growth of data volume, in which multimedia data such as video images account for a large proportion. How to efficiently deal with these huge amounts of data and quickly extract valuable information from them is the current research hotspot. Often, big data has four characteristics: large amount of data, fast response, diverse data types and low value density. Video big data also has the above characteristics, but its particularity lies in the fact that data redundancy is greater and requires efficient compression coding and analysis processing. In general, the study of video big data includes video data representation, intelligent video analysis, video compression and transmission, video display and evaluation. In the development trend, the representation of video data will develop in realistic and intelligent directions; the intelligent video analysis technology will use the deep neural network to obtain more accurate recognition classification results; and the video compression technology will improve the compression efficiency as well Explore the method of reducing coding complexity and reduce the visual redundancy of video big data by combining the coding algorithm of human visual perception characteristics; video display equipment will be accompanied by the corresponding change of video data representation; video quality Of the evaluation criteria will be a single image quality evaluation to a more comprehensive and comprehensive user experience quality evaluation development.