Abstract
In the Scalable Video Coding (SVC) standard, a multi-layer based structure is utilised to support scalability. However in the latest Joint Scalable Video Model (JSVM) reference software, the rate control algorithm is implemented only in the base layer, and the enhancement layers are not equipped with a rate control scheme. In this work, a novel rate control algorithm is proposed for when inter-layer prediction is employed. Firstly, a Rate-Quantisation (R-Q) model, which considers the coding properties of different prediction modes, is described. Secondly, an improved Mean Absolute Difference (MAD) prediction model for the spatial enhancement layers is proposed, in which the encoding results from the base layer are used to assist the linear MAD prediction in the spatial/CGS enhancement layers. Simulation results show that, on average, rate control accuracy is maintained to within 0.07%. Compared with the default JVT-G012 rate control scheme employed in SVC, the proposed rate control algorithm achieves higher coding efficiency, namely an improvement of up to 0.26dB in PSNR and a saving of 4.66% in bitrate.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
