Video stabilization for an aerial surveillance system using sift and surf

Abstract

Nowadays, an aerial closed-circuit television provides an outsize quantity of information compared with ancient closed-circuit television. But, it suffers from unsought cameras motion that presents new challenges. This paper presents a work on video stabilization and detection of the objects which are in moving position, system supported camera motion estimation. Most effective algorithms SIFT (Scale Invariant Feature Transform) and SURF (Speeded Up Robust Features) used for feature extraction. When estimating the worldwide parameters of camera motion victimization transformation, we have a tendency to find moving object by Kalman filtering. SURF (Speeded up Robust Features) is additionally a property extraction algorithmic rule works nearly like SIFT algorithmic rule, however quicker than SIFT.