An Efficient Multiview Video Coding Plan Based on Heterogeneous Computing System (IEEE)
Starting from brief introduction to fast developing Multi-view Video Coding technology and Compute Unified Device Architecture (CUDA) of General-purpose computing on graphics processing units (GPGPU), we designed an Heterogeneous computing system plan with cooperation of CPU and parallel GPU to accelerate the MVC coding process. To take advantage of GPGPU, parallel MVC motion estimation approach on CUDA was interpreted. For further optimize the system efficiency, a cooperating mode between CPU and GPU are designed.
Paper available at IEEE.