An Online Discriminative Approach to Background Subtraction

We present a simple, principled approach to detecting foreground objects in video sequences in real-time. Our method is based on an on-line discriminative learning technique that is able to cope with illumination changes due to discontinuous switching, or illumination drifts caused by slower processes such as varying time of the day. Starting from a discriminative learning principle, we derive a training algorithm that, for each pixel, computes a weighted linear combination of selected past observations with timedecay. We present experimental results that show the proposed approach outperforms existing methods on both synthetic sequences and real video data.

• Type: Conference
• Year: 2006
• Where: AVSS
• Authors: Li Cheng, Shaojun Wang, Dale Schuurmans, Terry Caelli, S. V. N. Vishwanathan
• Keywords: AVSS 2006, Discriminative Learning, Discriminative Learning Principle, Discriminative Learning Technique, Signal Processing