A ROBUST VIDEO STABILIZATION ALGORITHMS FOR GLOBAL MOTION ESTIMATION USING BLOCK MATCHING

In a camera dictionary, shakiness jitters are normally used. Jitteriness problem is encountered due to frequent use of a phone camera. Shooting video from moving vehicles or by using a handheld phone camera causes Shakiness. Block matching algorithms can be used for implementing the video stabilization process. The proposed algorithm in this paper avoids shakiness from a raw video of YUV and results in stabilizing in YUV video. The proposed method works on extracting motion among consecutive frames of any raw video. With respect to every scene of the video, premeditated camera motions like panning and translation motion are smooth with slow variations in the time realm from frame to frame. Fast motion variation with respect to frame to frame is increased while sponging the camera motion. A low pass filter can be used to find the recovery of the calculated motion parameters. The corrective shifting of noisy components of the motion vector stabilized the video images. The accurate result can be achieved by a successful run on shaky video without hampering its quality. The parallel processing techniques have been applied for the estimation of motion vectors, the results improved by applying to modify the proposed algorithm for dropping the operation time of the program. (C) 2020 INT TRANS J ENG MANAG SCI TECH.