DocumentCode :
3043891
Title :
Compressed video copy detection based on edge analysis
Author :
Zhang, Zhijie ; Zou, Jianhua
Author_Institution :
Syst. Eng. Inst., Xi´´an Jiaotong Univ., Xi´´an, China
fYear :
2010
fDate :
20-23 June 2010
Firstpage :
2497
Lastpage :
2501
Abstract :
In this paper, we propose a new compressed video copy detection approach, which is based on edge analysis without decompression. We use AC coefficients of DCT domain to obtain edge information in video frames, then measure the similarity of videos through edge information. The experimental results demonstrate that the proposed approach outperforms the previous approaches in accuracy and efficiency.
Keywords :
data compression; edge detection; video coding; DCT domain coefficient; compressed video copy detection; edge information analysis; video frames; Asset management; Automation; Data mining; Decoding; Discrete cosine transforms; Information analysis; Robustness; Video compression; Videoconference; Watermarking; AC coefficient; compressed domain; edge; video copy detection;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information and Automation (ICIA), 2010 IEEE International Conference on
Conference_Location :
Harbin
Print_ISBN :
978-1-4244-5701-4
Type :
conf
DOI :
10.1109/ICINFA.2010.5512049
Filename :
5512049
Link To Document :
بازگشت