Real Time Video Stabilization using PLK Tracking Algorithm

Sunil Sankol S M | Praveenkumar B S | Vanishree K Rao ^[2]

Abstract: Video Stabilization is a technique to improve video quality. Hence, the unwanted motion fidget removal and enrichment technology which aims in removing troublesome trembling motion from videos. Its impact slopes upward rapidly with increasing popularity of handheld cameras and the cameras mounted on moving platforms comprehensive video stabilization becomes essential. The proposed approach allows for Video Stabilization beyond the conventional filtering of the camera paths and some methods are based on mechanical devices such as gyro sensor to detect the camera motion and then shift the image sensor to compensate vibration. To overcome an imperfection, the algorithm deals with estimation of camera motion path by optical flow method using Pyramidal Lucas Kanade (PLK) Feature Tracking algorithm. Motion estimation is calculated by applying optical flow of common coverage areas. In a practical the proposed techniques has been applied in a various real time objects tracking as a applied in various real time objects tracking as a pre-processing stage. Experimental observations show that the method can perform real time and provide good performance.

Keywords: Stabilization, Pyramidal Lucas Kanade PLK, Descritization

Edition: Volume 7 Issue 6, June 2018,

Pages: 799 - 802