论文部分内容阅读
针对视频流在通过易错信道传输时丢包和延时导致的降低接收端的视频质量问题,为了提升视频流的容错性能,提出了一种快速容错编码策略,对每个主要帧自适应地编码多个冗余帧,并且进一步提出了快速多冗余帧数量计算的方法.利用实验数据得到失真和码流消耗的函数,使用编码参数来快速估计率失真代价值,进而根据失真模型对具有不同冗余帧计数的冗余帧的编码参数进行限制.由于序列相邻帧的特征具有相似性,因此使用临近帧的编码参数及失真和码流消耗对统计模型进行更新.最终通过利用失真和码流消耗的统计模型,分析冗余帧数量对总体端对端率失真代价的影响,以此选择优化的冗余帧数目和编码参数.实验结果表明:该算法与自适应调整编码参数的冗余帧编码的算法相比,峰值信噪比平均提高约1dB,具有更好的端对端视频传输质量.
Aiming at the problem of reducing video quality at the receiving end caused by packet loss and delay when the video stream is transmitted through the error-prone channel, a fast error-tolerant coding strategy is proposed to adaptively encode each main frame in order to improve the fault tolerance performance of the video stream Multiple redundant frames, and further proposes a method for calculating the number of fast and redundant frames.Furthermore, the function of distortion and bit-stream consumption is obtained by using the experimental data, and the coding parameters are used to estimate the value of rate-distortion cost quickly. Based on the distortion model, The coding parameters of redundant frames in redundant frame count are limited.As the characteristics of the adjacent frames in the sequence are similar, the statistical model is updated by using the encoding parameters and distortion of adjacent frames and the code stream consumption.Finally, by using the distortion and code The statistical model of stream consumption is analyzed and the effect of the number of redundant frames on the overall end-to-end rate distortion cost is analyzed to select the optimal number of redundant frames and coding parameters.The experimental results show that the proposed algorithm can adaptively adjust the redundancy of coding parameters Compared with the frame coding algorithm, the peak signal-noise ratio is improved by about 1dB on average, which has better end-to-end video transmission quality.