Real-Time Video Stabilization for Unmanned Aerial Vehicles
Yue Wang, Zujun Hou, Karianto Leman, Richard Chang · 2011
This paper presents a new real-time video stabilization method for Unmanned Aerial Vehicles (UAV). It mainly consists of three steps. Firstly, the keypoints are located based on FAST corner detection and preliminarily matched. Secondly, the matched keypoints are then involved for estimation of affine transform to reduce false matching keypoints. Finally, motion estimation is performed based on affine transform model and the compensation for vibration is conducted based on Spline smoothing. The experiments show that our method performs well and processes up to 30 fps. 1.