Automatic video mosaic imaging based on improved SIFT algorithm
Fuhua Song · Cehui kexue · 2013
As the existing methods for video image mosaic take high computational costs and huge computation,this paper proposed a method for making panoramas by getting an adaptive frame and matching method of limiting behavior region.According to inter-frame rate and frame overlap interval of linear model,all frames were matched to their latest neighbor key frames.In the process of mosaic,at the aspect of feature extracting,this paper used the improved SIFT feature operator,and the improved RANSAC algorithm to reduce image registration error.This article used the linear weighting fusion algorithm to achieve the desired gradually-enter gradually-leave effect in the overlap image region.Experiments on video sequences showed that this algorithm could extract mosaic key frames with satisfying quality.