一种基于颜色特征的快速视频时域配准算法

摘要

Peak Signal to Noise Ratio (PSNR) is the common used automatic temporal alignment algorithm in the full reference objective video quality assessment system. However, it costs too much time to meet the need of applications in real-time condition. To solve this problem, a fast video temporal alignment algorithm based on chroma signal histogram is proposed. This algorithm, which based on color characteristic-based image retrieval and PSNR algorithm, not only improves the computational speed, but also ensures the computational accuracy. The proposed algorithm can basically meet the demand of applications in full reference objective video quality assessment system in real-time condition.%PSNR算法是全参考型客观视频质量评价系统中常用的自动时域配准算法,但其计算时间较长,难以满足视频质量评价系统在实时条件下的应用.针对这一问题,提出基于色度信号直方图的快速视频时域配准算法,该算法结合基于颜色特征的图像检索技术和PSNR算法,既提高了计算速度,又保证了计算精度,基本满足全参考型客观视频质量评价系统在实时条件下的应用.