International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 72 - Number 20 |
Year of Publication: 2013 |
Authors: Paresh Rawat, Jyoti Singhai |
10.5120/12657-9254 |
Paresh Rawat, Jyoti Singhai . Adaptive Motion Smoothening for Video Stabilization. International Journal of Computer Applications. 72, 20 ( June 2013), 14-20. DOI=10.5120/12657-9254
Persons using low coast hand held cameras are untrained, thus captured videos suffers from severe handshakes and jitters. Hence digital video stabilization algorithm is required to acquire video sequences by removing these unwanted handshakes and jitters. This paper proposed an adaptive motion smoothening method that removes high frequency jitters, by filtering the accumulated global motion vectors. A modified method for motion vector validation implemented on global motion vectors using adaptive threshold. Proposed method not only removes the jitters but also preserves the scene information. The proposed method reduces the missing image areas significantly. This allows using a simplified edge completion method to generate the full frame stabilized video, which performs better in the presence of large moving object in the scene.