Subscribe Now Subscribe Today
Science Alert
 
FOLLOW US:     Facebook     Twitter
Blue
   
Curve Top
Journal of Software Engineering
  Year: 2014 | Volume: 8 | Issue: 4 | Page No.: 239-251
DOI: 10.3923/jse.2014.239.251
 
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail
Shot Boundary Detection Based on Non-negative Matrix Factorization in DT-CWT Domain
Jia-Xing Wang, Lie-jun Wang and Ting Yan

Abstract:
An efficient algorithm for shot boundary detection is proposed in this study. In order to eliminate the disturbances caused by illumination and camera motion, Dual Tree Complex Wavelet Transform (DTCWT) is used in the process of raw feature generation. Then, the enhanced texture features were extracted from the six detailed subbands by calculate the mean and standard deviation of the high-frequency coefficients. Non-negative Matrix Factorization (NMF) is used for extracting structure feature from the low-frequency sub-bands image and dynamic threshold is used for shot boundaries detection. The experiment result conducted on TRECVID 2001 test data and other test videos shows that the proposed scheme can not only overcome illumination and motion effects efficiently, but also achieve a high detection speed.
PDF Fulltext XML References Citation Report Citation
How to cite this article:

Jia-Xing Wang, Lie-jun Wang and Ting Yan, 2014. Shot Boundary Detection Based on Non-negative Matrix Factorization in DT-CWT Domain. Journal of Software Engineering, 8: 239-251.

DOI: 10.3923/jse.2014.239.251

URL: https://scialert.net/abstract/?doi=jse.2014.239.251

 
COMMENT ON THIS PAPER
 
 
 

 

 
 
 
 
 
 
 
 
 

 
 
 
 
 

       

       

Curve Bottom