An efficient rate–distortion optimization method for dependent view in MV-HEVC based on inter-view dependency

• Published in: Signal processing. Image communication 2021-05, Vol.94, p.116166, Article 116166
• Main Authors: Li, Tiansong, Yu, Li, Wang, Hongkui, Kuang, Zhuo
• Format: Article
• Language: English
• Subjects: Coders; Coding; Dependent view; Distortion; Inter-view dependency; Lagrange multiplier; Lagrangian multiplier; Multi-view coding; Optimization; RDO; Scaling factors
• ISSN: 0923-5965; 1879-2677
• DOI: 10.1016/j.image.2021.116166

Rate–distortion optimization (RDO) is utilized to select the optimal coding parameters in multi-view video coding (MVC), which employs a Lagrange multiplier to balance the relationship between the distortion and the bitrate. In this paper, an efficient RDO method for the dependent view (DV) in multi-view video (MVV) is proposed based on inter-view dependency. First of all, by investigating the sources of the distortion in the DV, a new distortion model for the DV is established. In addition, based on the proposed distortion model, an efficient Lagrangian multiplier decision for B frame is proposed by considering the inter-view dependency. Finally, the optimized Lagrangian multiplier for P frame is designed using the scaling factor which is deduced to have a linear relationship with the disparity between I frame and P frame. Experiment results demonstrate that compared with the original HTM-16.0 encoder, the proposed overall method reduces 12.19% BD-rate for the DV on average, bringing 0.40 dB BD-PSNR gain. •First of all, by investigating the sources of the distortion in the DV, a new distortion model for the DV is established.•In addition, based on the proposed distortion model, an efficient Lagrangian multiplier decision for B frame is proposed by considering the inter-view dependency.•Finally, the optimized Lagrangian multiplier for P frame is designed using the scaling factor which is deduced to have a linear relationship with the disparity between I frame and P frame.