A Comparison of Different Block Matching Algorithms for Motion Estimation

AbstractIn this paper, four different block matching algorithms using motion estimation are evaluated where the effects of the macro block size used will be reviewed to find the best algorithm among them is scrutinized to determine the most optimal algorithm. Four different block matching algorithms are considered and implemented. Each algorithm is evaluated using different movies from the TRANS database [11] and comparisons are made through the Peak Signal to Noise Ratio (PSNR) and search points per macro block (i.e. computation time) for different sizes of macro blocks and search areas. The results suggest that among all the evaluated algorithms, ARPS has the best PSNR based on computation time