Automatic threshold selection for automated visual surveillance

被引:3
作者
Çelik, T [1 ]
Kabakli, T [1 ]
Uyguroglu, M [1 ]
Özkaramanli, H [1 ]
Demirel, H [1 ]
机构
[1] Dogu Akdeniz Univ Gazimagusa, Ileri Teknol Arastirma & Gelistirme Enstitusu, KKTC, Gazimagusa, Turkey
来源
PROCEEDINGS OF THE IEEE 12TH SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE | 2004年
关键词
D O I
10.1109/SIU.2004.1338568
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Automated Visual surveillance systems mostly depend upon effective background subtraction technique. Most of the background subtraction techniques mainly suffer from parameter updates for threshold selection. Here a new threshold selection technique which is found while system trains to learn background, is proposed.
引用
收藏
页码:478 / 480
页数:3
相关论文
共 10 条