Merge Mode for Deformable Block Motion Information Derivation
Na Zhang,Xiaopeng Fan,Debin Zhao,Wen Gao
DOI: https://doi.org/10.1109/tcsvt.2016.2589818
IF: 5.859
2016-01-01
IEEE Transactions on Circuits and Systems for Video Technology
Abstract:Motion estimation (ME) plays an important role in most video coding standards since it can improve the coding performance significantly. Block matching ME (BMME) based on translation motion model is employed in both H.264/Advanced Video Coding and High Efficiency Video Coding (HEVC), which assumes that all the pixels in a block undergo the same motion. However, BMME cannot represent the complex motion accurately, such as rotation, zoom, and deformation motion. To resolve the problem, a merge mode for deformable block motion information derivation is proposed, which considers bilinear interpolation model, six-parameter affine model, and four-parameter affine model simultaneously. The sum of absolute transformed differences criterion is used to select the best model for a block. In the proposed merge mode, the motion information of the current block is derived from the motion information of its neighbors. It does not need to perform ME and transmit motion parameters. The proposed merge mode is integrated into HEVC test model (HM) 14.0. The simulation results show that compared with HM14.0, the proposed merge mode can bring 1.3%, 1.7%, and 2.7% Bjøntegaard delta (BD) bitrate savings on average for test sequences of HEVC, and 11.0%, 15.2%, and 17.4% BD bitrate savings for deformation sequences in random access main, low-delay B main, and low-delay P main test conditions, respectively, while the increased encoding and decoding complexities are moderate.