MapReduce Analytics Based Fast Search Algorithms

Abstract

ABSTRACT Using the motion estimation process is still one of the best technologies for video coding standards such as H.26X or MPEG-X. These standards have been supplied to video compression, but this area is still open algorithms. It can provide excellent coding efficiency while maintaining a low distortion. A new motion estimation of a full search algorithm with big data analytics in this research was proposed, which uses MapReduce analysis metrics. The developed method can not only be used to reduce the stored number of motion vectors compared with the full search algorithm but also to perform superior peak signal-to-noise ratio than other recent developed scheme, and achieve the same peak signal-to-noise ratio as the full search algorithm. Simulation exemplifies that our proposed scheme obtains better improvement than the full search algorithm and other well-known methods. The comparison of our proposed algorithms with other methods is evaluated both subjectively and qualitatively and the simulation demonstrated that our proposed method could obtain the desired reconstructed performance subjectively and objectively.