Digital image stabilization with sub-image phase correlation based global motion estimation.
Guanghui Ren,Peng Li,Gangyi Wang,C Morimoto,R Chellappa,M Tico,S Alenius,M Vehvilainen,C Wang,JH Kim,KY Byun,J Ni,SJ Ko,Y Luo,M Celenk,R Korah,JRP Perinbam,SJ Ko,SH Lee,KH Lee,SC Hsu,SF Liang,KW Fan,CT Lin,C Feng-Dong,BR Hong,GD Liu,Y Zhang,F Chen,J Wang,W Zhang,Q Zhou,CT Lin,CT Hong,CT Yang,Y Liu,D Fu,B Li,Q Xu,S Erturk
1998-01-01
Information Technology Journal
Abstract:The gray-scale projection algorithm can detect the motion vector of continuous frames rapidly in digital image stabilization system, but the interference of local moving objects may influence the estimation accuracy. In this study, we present a novel coarse-to-fine image stabilization approach which combined gray-scale projection algorithm with block matching algorithm by analyzing two classical motion estimation algorithms. The proposed method considers both accuracy and efficiency, since it makes good use of low computational complexity of gray-scale projection algorithm and high accuracy of block matching algorithm. The experimental simulations reveal that, expect for its low computational complexity, the proposed algorithm achieves high motion estimation accuracy. Moreover, it can efficiently avoid the interference caused by local moving objects.