DocumentCode
2401279
Title
Video falsifying by motion interpolation and inpainting
Author
Shih, Timothy K. ; Tan, Nick C. ; Tsai, Joseph C. ; Zhong, Hsing-Ying
Author_Institution
Dept. of Comput. Sci. & Inf. Eng., Tamkang Univ., Tamkang
fYear
2008
fDate
23-28 June 2008
Firstpage
1
Lastpage
8
Abstract
We change the behavior of actors in a video. For instance, the outcome of a 100-meter race in the Olympic game can be falsified. We track objects and segment motions using a modified mean shift mechanism. The resulting video layers can be played in different speeds and at different reference points with respect to the original video. In order to obtain a smooth movement of target objects, a motion interpolation mechanism is proposed based on continuous stick figures (i.e., a video of human skeleton) and video inpainting. The video inpainting mechanism is performed in a quasi-3D space via guided 3D patch matching for filling. Interpolated target objects and background layers are fused by using graph cut. It is hard to tell whether a falsified video is the original. We demonstrate the original and the falsified videos in our website at http://www.mine.tku.edu.tw/video_demo/). The proposed technique can be used to create special effects in movie industry.
Keywords
graph theory; image fusion; image matching; image motion analysis; image segmentation; interpolation; object detection; target tracking; video signal processing; graph cut; guided 3D patch matching; modified mean shift mechanism; motion interpolation; motion segmentation; movie industry; object target tracking; quasi 3D space; video falsifying; video inpainting; Computer graphics; Filling; Games; Humans; Image edge detection; Image restoration; Interpolation; Motion pictures; Skeleton; Target tracking; graph cut; image completion; mean shift; motion interpolation; video inpainting; video special effect;
fLanguage
English
Publisher
ieee
Conference_Titel
Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on
Conference_Location
Anchorage, AK
ISSN
1063-6919
Print_ISBN
978-1-4244-2242-5
Electronic_ISBN
1063-6919
Type
conf
DOI
10.1109/CVPR.2008.4587701
Filename
4587701
Link To Document