Abstract
Video stitching has become popular due to recent advancements in technology to provide broad views and high-resolution displays. Comprehensive view or panoramic videos and high-resolution displays are created by stitching videos captured by multiple cameras or by a single camera at different points of time. This paper proposes a video stitching technique with stabilization for moving multi-camera videos adopting the wavelet decomposition technique. This method uses only those feature points that reduce the mismatching and increase the precision in estimating the transformation from among the feature points identified by the Speed-Up Robust Features detector. This work differs from the similar work of others in two directions. Instead of using all selected feature points for the matching purpose, only significant among them are used. Unlike others, the frames are stabilized before they are stitched.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
