Please use this identifier to cite or link to this item:
Title: Digital Video Stabilization
Keywords: video stabilization;bundled camera paths;steadyflow;depth camera;rolling shutter effects;kinect
Issue Date: 23-Jul-2014
Source: LIU SHUAICHENG (2014-07-23). Digital Video Stabilization. ScholarBank@NUS Repository.
Abstract: This thesis addresses the problem of digital video stabilization. Videos captured by hand-held devices (e.g., cell-phones or DVs) often appear remarkably shaky. Digital video stabilization improves the video quality by removing unwanted camera motions. In this thesis, three different methods are presented. We first address video stabilization by adopting a depth camera. We show that the depth can facilitate both camera motion estimation and frame warping, thus make the video stabilization a much well posed problem. Then, we present a video stabilization approach named as bundled camera paths, in which multiple 2D camera paths are proposed to represent camera motions. Its mesh-based , spatially-variant motion representation allows us to fundamentally handle parallax issues. Finally, we present a novel motion model, SteadyFlow, which is a specific optical flow by enforcing strong spatial coherence. Our experiments demonstrate the effectiveness of our stabilization methods on real word challenging videos.
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
LiuSC.pdf7.02 MBAdobe PDF



Page view(s)

checked on Feb 16, 2018


checked on Feb 16, 2018

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.