Bit Allocation And Rate Control Research For Depth Video Coding

In recent years,the multimedia technology has developed rapidly.As the demand of user experiment is increasing,3D video has attracted great attention since it can provide immersive vision for viewers.Great demands are being placed on the transmission and storage of 3D video by the large amount of 3D video data.Multiview video plus depth format(MVD)is one of the most popular 3D video formats.Except for texture video,MVD also includes depth video which provides the vertical distance from an object in the 3D space to the camera plane.Depth video coding methods need further researched.Bit allocation and rate control are important components of video coding.Many effective rate control algorithms for texture video have been proposed for different video coding standards and practical applications.Due to the differences of texture video and depth video,it is inefficient to use the rate control methods for texture video to depth video.Therefore,considering the unique characteristics of depth video,bit allocation and rate control for depth video coding are researched deeply in this paper,mainly in the following two aspects.1)A novel rate control algorithm based on adaptive R-? model is proposed for depth video coding.First,in order to obtain an accurate rate control for depth video coding,a modified frame level bit allocation method based on coding bits statistical distribution of depth maps is proposed.Second,considering that different areas in a depth map have imparity effect on virtual view rendering,the blocks of the depth map are divided into two types,namely,interested blocks for virtual view rending(IBV)and non-interested blocks for virtual view rending(NIBV).Then,two different R-? models are derived for IBV and NIBV,respectively.The optimal bitrates for IBV and NIBV are determined by solving an optimization problem.Finally,LCU level rate control is performed by adaptively adjusting the Lagrange multiplier so as to avoid blocking artifacts and smooth the quality of depth map.Experiment results demonstrate that the proposed method can improve the quality of rendered virtual views efficiently.2)An inter-view frame level bit allocation is proposed for depth video coding based on MV-HEVC.Since inter-view prediction has been adopted in MV-HEVC as an important coding tool,an independent view picture could be predicted from the corresponding basic view picture.The distortion of the basic view could be propagated to the independent view,inter-view dependency is inevitable.In order to improve coding efficiency,the inter-view dependency must be taken into account for the rate-distortion model in independent view.Therefore,a more reasonable bit allocation method for multiview depth video is proposed.In this paper,the inter-view dependency is analyzed in detail.,A RD model for independent view is proposed based on the analysis.Based on the RD model,the inter-view bit allocation is determined by solving an optimization problem by Lagrange Multiplier method.Experiment results demonstrate that the proposed method can improve the coding efficiency of multiview depth videos.