Content-Aware Video Transcoding via Visual Attention Model Analysis

In this paper, a frame-drop transcoding algorithm based on visual attention model is proposed for reducing the temporal resolution of a compressed video in order to fit the channel target bitrate. In the proposed method, the visual attention model is employed to measure frame complexity in order to determine whether frames should be skipped or not. Through the model analysis, we can preserve the significant frames to avoid the jerky effect. Experimental results show that the proposed method can achieve higher quality compared to the period frame skipping method.