Eduarda R. Monteiro,  Bruno B. Vizzotto,  Cláudio M. Diniz,  Bruno Zatt,  Sergio Bampi. 

This work presents a parallel GPU-based solution for the Motion Estimation (ME) process in a videoencoding system. We propose a way to partition the steps of Full Search block matching algorithm in the CUDA architecture, and to compare the performance with a theoretical model and two implementations (sequential and parallel using OpenMP library). We obtained a O(n2/log2n) speed-up which fits the theoretical model considering different search areas. It represents up to 600x gain compared to the serial implementation, and 66x compared to the parallel OpenMP implementation.

http://www.lbd.dcc.ufmg.br/colecoes/sim/2011/007.pdf

Processando consulta.