سال انتشار: ۱۳۹۰

محل انتشار: هفتمین کنفرانس ماشین بینایی و پردازش تصویر

تعداد صفحات: ۵

نویسنده(ها):

Alireza Asvadi – Department of ECE, DSP Lab.Babol University of TechnologyBabol, Iran
MohammadReza Karami-Mollaie – Department of ECE, DSP Lab.Babol University of TechnologyBabol, Iran
Yasser Baleghi – Department of ECE, DSP Lab.Babol University of TechnologyBabol, Iran
Hosein Seyyedi-Andi – Department of ECE, DSP Lab.Babol University of TechnologyBabol, Iran

چکیده:

In the present paper, an improved method for objecttracking is proposed using Radial Basis Function NeuralNetworks. Here, the Pixel-based color features of object are usedto develop an extended background model. The object andextended background color features are then used to train RBFNeural Network. The trained RBFNN will detect and track objectin subsequent frames. The performance of the proposed trackeris tested with many video sequences. The proposed tracker isillustrated to be suitable for real-time object tracking due to itslow computational complexity