微课视频资源适配模型及应用研究

发布时间：2018-08-25 09:15

【摘要】：随着互联网和移动终端技术的发展,多样的终端设备和复杂的异构网络为多媒体技术提出了新的挑战。而随着微课教学的兴起,作为主要的知识承载形式的视频资源大量出现。学习者在进行微课学习时会使用大量不同的终端设备,终端设备的多样性要求视频码流满足不同的分辨率、码率等。视频适配作为解决这一问题的方法,成为了研究的热点。但是现有的视频适配研究主要将重点放在提高视频编码效率上。要想合理安排计算资源和存储资源,在尽量减少浪费的前提下将微课视频资源顺利传输并显示在终端设备上,必须进行相应的适配。经过适配后的视频不仅能够使用户便捷地获取资源,而且可以节省在视频计算和存储时的资源。为满足学习者随时随地获取微课视频资源的需求,本研究提出一种微课视频资源适配模型,设计了针对微课视频资源的描述模型,并对视频资源的适配过程进行了改进。首先,基于MPEG-21框架,提出了新的微课视频资源适配模型。将该模型分为三个部分,适配决策引擎、描述适配引擎、比特流适配引擎。适配决策引擎部分选用较为普遍的基于规则的适配策略,将决策参数分别传送至描述适配引擎和比特流适配引擎。在描述适配引擎中,主要研究了针对微课视频资源的描述模型,设计了基于MPEG-7和CELTS-41标准的微课视频资源描述模型。然后,在比特流描述引擎中设计了转码序列规划算法,将视频资源的适配过程看作是资源由原始状态向最终状态的序列,通过比特流适配引擎中存放的原子性的转码操作,运用搜索法,排序组合所有转码操作后,比较出耗时最短的转码路径,从而确定最优的转码路径。在经过大量的实际测验之后,总结出了一般性的转码组合序列,将这些序列当做启发式规则,重新对原有算法进行改进,改进后的算法即为最终算法。最后,以微课视频资源适配模型为核心,设计并实现了微课视频内容适配系统。此外,本研究还将适配时间作为指标,对提出的转码序列规划算法进行评价。实验证明,该算法能够提高适配操作的复用性和效率,节省适配时间,且微课视频内容适配系统具有较高的用户满意度。以微课视频资源适配模型为研究对象,研究在适配过程中的描述和优化问题,不仅为微课视频的适配提供了新思路,同样可以为适配其他类型的视频资源时提供参考。
[Abstract]:With the development of Internet and mobile terminal technology, various terminal devices and complex heterogeneous networks pose new challenges to multimedia technology. With the rise of microcourse teaching, video resources, which are the main forms of knowledge bearing, appear in large numbers. Learners will use a large number of different terminal devices in micro-learning. The diversity of terminal equipment requires video stream to meet different resolution, bit rate and so on. As a method to solve this problem, video adaptation has become a hot research topic. However, the current research on video adaptation mainly focuses on improving the efficiency of video coding. In order to arrange computing resources and storage resources reasonably, and to transfer and display the micro-lesson video resources on the terminal equipment without wasting as much as possible, the corresponding adaptations must be carried out. The adapted video can not only make it easy for users to obtain resources, but also save the resources in video computing and storage. In order to meet the needs of learners to obtain micro-lesson video resources at any time and anywhere, this paper proposes a micro-lesson video resource adaptation model, designs a description model for micro-lesson video resources, and improves the adaptation process of micro-lesson video resources. Firstly, based on MPEG-21 framework, a new video resource adaptation model is proposed. The model is divided into three parts, the adaptation decision engine, the adaptive engine and the bitstream adaptation engine. In the part of adaptation decision engine, the general rule-based adaptation strategy is adopted, and the decision parameters are transferred to the description adaptation engine and the bitstream adaptation engine respectively. In the description adaptation engine, this paper mainly studies the description model of micro-lesson video resource, and designs a micro-lesson video resource description model based on MPEG-7 and CELTS-41 standard. Then, the transcoding sequence planning algorithm is designed in the bitstream description engine. The adaptation process of video resources is regarded as the sequence of resources from the original state to the final state, and the atomic transcoding operation stored in the bitstream adaptation engine is adopted. The search method is used to sort and combine all transcoding operations, and the shortest transcoding path is compared to determine the optimal transcoding path. After a lot of practical tests, the general transcoding combination sequences are summarized, which are regarded as heuristic rules, and the original algorithm is improved again. The improved algorithm is the final algorithm. Finally, a micro-lesson video content adaptation system is designed and implemented with the micro-lesson video resource adaptation model as the core. In addition, the adaptive time is used as an index to evaluate the proposed transcoding sequence planning algorithm. Experiments show that the algorithm can improve the reusability and efficiency of adaptation operations, save adaptation time, and the micro-course video content adaptation system has a higher user satisfaction. Taking the video resource adaptation model of micro-lesson as the research object, this paper studies the description and optimization problems in the process of adaptation, which not only provides a new idea for the adaptation of micro-lesson video, but also provides a reference for adapting other types of video resources.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G434

【参考文献】