Real-time objective quality assessment based on coding parameters extracted from H.264/AVC bitstream

In this paper, we propose a new method for objective video quality assessment from coding parameters of an H.264/AVC bitstream as a hybrid/bitstream category. Conventional image and video quality assessment methods evaluate image or video quality using a degraded video or image after fully decoding it. On the other hand, the proposed method assesses video quality not with reconstructed videos but using coding parameters extracted from a bitstream. The parameters can be easily extracted while decoding the H.264/AVC bitstream. To assess video quality, the proposed algorithm is based on statistics of boundary strengths (BS), quantization parameters (QP), and average bitrates. Accuracy of the proposed method is 9.7% higher than those of the conventional algorithms in terms of Pearson correlation. Furthermore, the proposed method has significantly lower computational complexity than the conventional methods.