Abstract:
Video data traffic is moving from Broadcast to Internet (e.g. OTTTV) distribution. Related economic models are different. It requires to review the rate control paradigm ...Show MoreMetadata
Abstract:
Video data traffic is moving from Broadcast to Internet (e.g. OTTTV) distribution. Related economic models are different. It requires to review the rate control paradigm moving from Constant Bitrate (CBR) model to Variable Bitrate (VBR) model using a capped bitrate constraint. In this paper, we propose a Constant Quality Control (CQC) algorithm which minimizes the bitrate under constraints of a target video quality level and a capped bitrate. CQC algorithm relies on a R-D model which considers the temporal distortion propagation along a video sequence. Introducing a D(Q) model, and a capped bitrate constraint, the optimal local quantizers are computed to fulfill a target level of distortion. Consistency and R-D performances of the proposed solution are verified in the context of HEVC. It meets the target level of quality with an average deviation of 6.9% while reducing bitrate of -7.6% in average compared to constant QP mode. CQC algorithm is particularly well adapted to the OTT-TV distribution, where the Adaptive Bitrate (ABR) mechanism is built on top of quality levels.
Date of Conference: 07-10 October 2018
Date Added to IEEE Xplore: 06 September 2018
ISBN Information:
Electronic ISSN: 2381-8549