TFI collaborated with City University of Hong Kong to publish a paper about ‘’Perceptual-based global optimization for CTU-Level Rate Control in HEVC’’
15/01/2019
Funded by University-Industry Collaboration Programme (UICP), TFI is pleased to collaborate with City University of Hong Kong to publish a paper about ‘’Perceptual-based global optimization for CTU-Level Rate Control in HEVC’’ In IEEE Transactions on Multimedia 2019, with high rankings at the top of their fields.
Superior visual quality is particularly desirable in video encoding. In view of this, the paper proposed a coding tree unit (CTU)[1] level rate control scheme to improve coding efficiency based on the perceptual rate-distortion model[2]. As shown in the diagram, each frame is divided into multiple CTUs or macroblocks. The blocks are then encoded with perceptual rate-distortion model, which simulates human visual perception and calculates the required coding bits and QP[3] value. The final image with little difference from original is produced. This algorithm can optimally achieve CTU level bit allocation given the bit rate budget[4].
Results show that our algorithm successfully enhances the coding performance and outperforms the rate control scheme in HEVC reference software in terms of rate perceptual distortion performance.
_____________________________________________
[1] Coding tree unit (CTU) is the basic processing unit of the High Efficiency Video Coding (HEVC) video standard and conceptually corresponds in structure to macroblock units that were used in several previous video standards.
[2] Perceptual rate-distortion model is established based on the divisive normalization scheme, which characterizes the relationship between local visual quality and coding bits.
[3] Quantization parameter (QP) controls the amount of compression for every macroblock in a frame.
[4] Bit rate budget measures the number of bits required to conveyed or processed per unit of time for specific frame or macroblock.
