【作者】 王金德；

【导师】 袁占亭； 张秋余；

【作者基本信息】 兰州理工大学 ， 通信与信息系统， 2008， 硕士

【摘要】 随着计算机和通信技术的不断发展,在网络中实时传输视频等媒体信息成为一种必然趋势。然而当前网络的带宽有限且带宽是实时变化的,不同用户的网络情况也存在很大差异,这些都为网络实时传输视频流制造了困难。解决这个问题最合适的方法就是可伸缩性的编码视频序列,可伸缩性编码通常将视频序列编码为基本层和增强层两个层,基本层码流是必须传输的,包含最基本的视频信息,码率较低。增强层是作为可选传输的,可以任意截断,带宽不足时甚至可以不传输,传输的增强层越多,解码出的图像质量就越好。由于这种方法生成的视频流可以在一个很大的码率范围内自我调整,因而能够适应复杂的网络带宽波动。其次,由于增强层的数据包丢失不会影响到其它帧,对视频质量影响也不大,因而可伸缩性视频编码生成的视频流具有较好的鲁棒性。但是这种方法的编码效率较低,和传统的非分级编码方法相比要低3db左右。为了提高它的编码效率,本文重点分析了分级编码和基于对象编码的研究现状,并在此基础上提出了一种基于对象的可分级视频编码方法。该方法在增强层运用了运动补偿技术,提高了编码效率;引入高空间分辨率增强层,获得较好的空间可分级性,可以应用在需要改变空间分辨率的视频传输中;在位平面编码时采用选择性增强技术,提高了感兴趣区域的视频质量。通过NS-2网络模拟器进行了仿真试验,结果表明该方法提高了编码效率和图像质量,并且可以提供视频的空间可分级性。更多还原

【Abstract】 With the development of computer and communications technology,real-time video and media transmission becomes an inevitable trend.However,the bandwidth of network is limited and real-time changed and the networks of different users are significantly different at the same time,which are the reasons for the difficulties of real-time video and media transmission.The scalable video coding sequence is the most suitable method to solve above problem. The video sequence is usually separated into two layers by scalable video coding,in which the basic layer and enhancement layer are included.The basic layer is must be transmitted,in which basic information is contained and bitrate is lower.The enhancement Layer is an option transmission,which can be cut off even non-transmission when the bandwidth is insufficient. So the enhancement layer transmission is more,the image quality is better.The scalable video streams are able to adapt to complicated fluctuations of network bandwidth because the video streams can be self-regulated within the scope of a large bitrate.Secondly,for the loss of enhancement layer data packet will not affect the other frame and the quality of video sequence,the video sequence generated by the scalable video coding has better robustness.Whereas,the encoding rate of the scalable video is less efficient,which is low 3 db compared with the traditional non-hierarchical coding method.In order to improve coding efficiency,a new fine granular scalable video coding method based on the image object is introduced based on the analysis of the fine-granularity-scalable and the coding based on the image object,in which motion compensation technology in enhancement layer is employed to improve the coding efficiency.The high spatial resolution enhancement layer is induced to obtain a better classification,which can be applied in the transmission of the spatial resolution video that needed to be changed.The selectively enhance is utilized in while plane coding to improve the video quality of the interested region.The simulation is carried on the NS-2 network simulator,which shows that the coding efficiency and the image quality can be improved and the video sequence space scalability can be provided,either.更多还原