Vol. 2, Issue 4, Part C (2016)

Abstract
Initial search center (ISC) prediction or motion vector (MV) prediction helps in estimating an initial search center for finding a refined motion vector in order to speed up the entire process of motion estimation (ME). Finding an appropriate initial MV leads to an early detection of actual motion vector. Various methods have been listed in literature for finding the best possible ISCs. In the current work we have presented a comparative analysis of various methods for estimating the initial search centers. Apart from detailing all these methods we have given the experimental results to find the effect of each of these methods on the video quality, time and accuracy of prediction. Video quality is calculated in terms of peak signal to noise ratio (PSNR) and accuracy of prediction is calculated by finding the difference between the predicted MV and the motion vectors obtained from full search (FS) algorithm as FS is considered as a standard algorithm which estimates the best possible MVs. The comparison of all these methods have been made by applying them on various video test sequences.