A near-duplicate 3D video detection algorithm by using hypercomplex representations.

Saved in:

Bibliographic Details
Title:	A near-duplicate 3D video detection algorithm by using hypercomplex representations.
Authors:	Sun, Ziqiang¹ sunzq87@gmail.com, Zhu, Yuesheng¹ zhuys@pkusz.edu.cn, Xing, Xiaomei¹, Luo, Guibo¹, Liu, Xiyao¹
Source:	Multimedia Tools & Applications. Jan2017, Vol. 76 Issue 1, p1055-1071. 17p.
Subjects:	Video recordings industry, 3-D video (Three-dimensional imaging), Digital video, Human fingerprints, Video compression
Abstract:	Copyright protection is still a crucial open issue in the 3D video industry with the development of 3DTV coupled with the increasing 3D video spread over the internet. In this paper a novel video fingerprinting is proposed for near-duplicate 3D video detection. Instead of generating fingerprints from the color frames and the depth maps separately, the proposed algorithm processes them in a holistic manner. The hypercomplex representation developed from the RGB and depth components is used to represent the 3D contents as quaternion frames and a novel quaternion centroid of the spatio-temporal gradient orientations is exploited to generate 3D video fingerprints based on these quaternion frames. Comprehensive experiments are conducted to evaluate the performance of the proposed method, and the results show that the proposed near-duplicate 3D video detection algorithm outperforms the state-of-the-art approaches in terms of robustness and discrimination. And the proposed fingerprints are also more compact than the existing approaches. [ABSTRACT FROM AUTHOR]
	Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Copyright protection is still a crucial open issue in the 3D video industry with the development of 3DTV coupled with the increasing 3D video spread over the internet. In this paper a novel video fingerprinting is proposed for near-duplicate 3D video detection. Instead of generating fingerprints from the color frames and the depth maps separately, the proposed algorithm processes them in a holistic manner. The hypercomplex representation developed from the RGB and depth components is used to represent the 3D contents as quaternion frames and a novel quaternion centroid of the spatio-temporal gradient orientations is exploited to generate 3D video fingerprints based on these quaternion frames. Comprehensive experiments are conducted to evaluate the performance of the proposed method, and the results show that the proposed near-duplicate 3D video detection algorithm outperforms the state-of-the-art approaches in terms of robustness and discrimination. And the proposed fingerprints are also more compact than the existing approaches. [ABSTRACT FROM AUTHOR]
ISSN:	13807501
DOI:	10.1007/s11042-015-3086-3