Study of x265 and Genetic Motion Search algorithm
Vivek Diliprao Indrale, Vidya Nitin More · 2016
The motion estimation and motion compensation techniques are widely used in video compression by different video codecs like MPEG2, MPEG4. Fast and precise algorithms are needed to target the need of real time video processing of high quality videos in emerging applications. Many fast block motion search algorithms have been developed to minimize computational cost required by full search algorithm, which depends upon the assumption that the matching error decreases monotonically as the searched point move closer to global minimum. However these techniques converge to local minimum which in turn increases the residual data and then increases the size and bit rate of output file. So to find optimized minimum, genetic algorithm is used in block motion search algorithm. Time saved for Genetic Motion Search method is 85.07% compared to full search Algorithm (FSA) and bitrate is increased by 5.62% and PSNR is decreased by the 0.1028% compared to FSA.