Rate control for transcoding in video transmission between heterogeneous networks

By inserting a video transcoding gateway between heterogeneous networks, it is very convenient to transmit video in different rates, formats and resolutions. In this paper, we focus on video transmission over the Internet or wireless channels, utilizing transcoding technology. We mainly investigate the rate control scheme for video transcoding, more specifically, for frame-skipped transcoding and spatial resolution reduction transcoding. Our first proposal is frame-layer rate control algorithm. We analyze the buffer constraints and estimate the channel rate for video transmission utilizing transcoding. As to channel rate estimation for the Internet, we adopt a model-based method to obtain the channel bandwidth. In the case of wireless channels, the effective channel bandwidth is estimated basing on the Markov Model. According to the estimated channel bandwidths, we adapt the bit allocation for each frame. In our second proposal, we modify the traditional macroblock-layer rate control schemes in TMN8 which adopts a raster scan order for encoding macroblocks. We utilize a new scan order based on macroblock activity which can be measured by the standard deviation of the macroblock. That is to say in our algorithm, the macroblocks are transcoded according to activities orders. The results indicate that this new scheme achieves a better average PSNR than that of TMN8.