An Efficient System for Video Stabilization by Differentiating between Intentional and Un-Intentional Jitters

An Efficient System for Video Stabilization by Differentiating between Intentional and Un-Intentional Jitters

Kamran Manzoor (University of Engineering and Technology, Pakistan), Umar Manzoor (The University of Salford, UK) and Samia Nefti (The University of Salford, UK)
DOI: 10.4018/978-1-4666-1788-9.ch020
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Video stabilization is one of the most important enhancements where jittering caused by un-intentional movements is removed. Existing video stabilizer software and tools cannot differentiate between intentional and un-intentional jitters in the video and treats both equally. In this paper, the authors propose an efficient and practical approach of video stabilization by differentiating between an intentional and un-intentional jitter. Their method takes jittered video as input, and differentiates between intentional and an un-intentional jitter without affecting its visual quality while producing stabilized video only if jitter is found to be un-intentional. While most previous methods produce stabilized videos with low resolution, this reduces quality. The proposed system has been evaluated on a large number of real life videos and results promise to support the implementation of the solution.
Chapter Preview
Top

Background

In the last decade, many methods for video stabilization have been proposed. There are three types of image stabilizers currently available (Robert, n.d.): Digital Image Stabilization (DIS), Optical Image Stabilization (OIS), and Mechanical Image Stabilization (MIS). Digital Image Stabilization (DIS) systems controls image stability by using electronic processing. But this system is not efficient because if there is a large motion of object in any frame then the system becomes fool in believing it as camera vibration so attempt to stabilize it and thus cause blurring and distorts picture quality.

The Optical Image Stabilization (OIS) system, manipulates the image before it gets to the Charge Coupled Device (CCD), unlike to the above explained Digital Image Stabilization (DIS) system, where image first hits the CCD. But this system makes the lens very complex and very susceptible to damage (Canon Digisuper 100xs, n.d.).

Mechanical Image Stabilization (MIS) not just stabilizing the image but involves stabilizing the entire camera. A device called “Gyros” is used in this type of stabilization (iMultimedia, n.d.; Kimura, 1985). But they are not suitable for energy sensitive imaging applications because they are heavy and consume more power.

Complete Chapter List

Search this Book:
Reset