Effective computation-aware algorithm by inter-layer motion analysis for scalable video coding

• We propose a strategy to control the computational complexity in the EL for SVC.
• The MVDs are analyzed to observe their relationship between BL and EL.
• The predicted search points in the EL and the RD-costs of the BL are considered.
• The computations can be efficiently allocated to MBs where need more computations.
• The proposal outperforms the other researches in terms of RD performance.

Scalable video coding incorporated with computation-aware ability achieves quality as well as being computation scalable. This paper presents a computation-aware algorithm for scalable video coding with spatial/quality scalability aiming for the best trade-off between rate distortion performance and computational consumption. We first observe and analyze and then establish a model for the motion vector difference relationship between the scalable base and enhancement layers. By using the modeling results, a linear algorithm for computation distribution is thus proposed to allocate the computation for each macroblock in the enhancement layer. In addition, the rate distortion costs of the base layer are also taken into account for the computation allocation process in order to further improve the coding performance. The simulation results demonstrate that our proposed computation-aware algorithm not only accomplishes better rate distortion performance than other works under the same computational constraints, but also achieves less computation necessities.