论文部分内容阅读
针对IP网络(如因特网)带宽随时变化的特性,传输视频时,为了有效地利用现有带宽,尽可能提高视觉质量,该文在MPEG-4精细可伸缩性视频编码(FGS)的选择性增强技术基础上提出一种基于运动矢量的自适应选择性增强算法。通过提取、分析基本层运动矢量,自适应地确定目标区域,并将结果在增强层编码用于选择性增强。使FGS编码实现基于内容的增强和优先传输感兴趣区域(目标区域)。实验表明该算法使目标区域在不同的码率下都能获得相对最佳的视频质量,适合网上视频监控之类背景静止而人们仅对场景中运动的人或物感兴趣的应用。
In view of the characteristic that the bandwidth of the IP network (such as the Internet) changes from time to time, in order to effectively utilize the existing bandwidth and improve the visual quality when the video is transmitted, the selectivity of the MPEG-4 Fine Scalability Video Coding (FGS) Based on the technique, an adaptive selective enhancement algorithm based on motion vector is proposed. By extracting and analyzing the base layer motion vector, the target area is adaptively determined and the result is encoded in the enhancement layer for selective enhancement. Enables FGS encoding for content-based enhancement and preferential transmission of regions of interest (target regions). Experimental results show that the proposed algorithm can achieve the best video quality in different target rates and is suitable for applications such as online video surveillance where the background is stationary and people are only interested in moving objects or objects in the scene.