Use of Particle Filter in Visual Surveillance for Tracking Multiple People |
Author(s): |
| Rupali Manoj Komatwar , Shree Guru Gobind Singhji Institute of Engineering & technology, Nanded; Chetana Pradeep Dodke, Atharva College of Engineering University of Mumbai, India |
Keywords: |
| Visual tracking, Particle filter, Sequential Bayesian estimation |
Abstract |
|
The particle filter (PF) has been presented for tracking multiple people in a visual surveillance application. The particle filter has proven to be an efficient, simple and robust tracking algorithm to detect and track colour objects in video. Detection is based on automated background modelling rather than a manually generated object colour model. In this research a methodical comparison between the new PF tracking method and one other multi-object trackers is presented on the PETS 2004 benchmark data set. Reliable visual tracking is indispensable in many emerging vision applications such as automatic video surveillance, human computer interfaces and robotics. The PF tracker gives significantly fewer false alarms owing to explicit modelling of object birth and death processes, while maintaining a good detection rate. |
Other Details |
|
Paper ID: NCTAAP073 Published in: Conference 4 : NCTAA 2016 Publication Date: 29/01/2016 Page(s): 305-311 |
Article Preview |
|
Download Article |
|
|
