Improved background modeling for real-time spatio-temporal non-parametric moving object detection strategies
作者：
Highlights：
• Real-time dynamic bandwidth estimation that reduces the amount of false detections.
• Selective update mechanism that significantly reduces the number of misdetections.
• The quality provided by previous background modeling strategies is improved.
• The computational cost of the simplest modeling strategy is barely increased.
• The proposed methods can be used by any spatio-temporal non-parametric strategy.
摘要
•Real-time dynamic bandwidth estimation that reduces the amount of false detections.•Selective update mechanism that significantly reduces the number of misdetections.•The quality provided by previous background modeling strategies is improved.•The computational cost of the simplest modeling strategy is barely increased.•The proposed methods can be used by any spatio-temporal non-parametric strategy.
论文关键词：Dynamic bandwidth estimation,Moving object detection,Non-parametric background modeling,Selective update,Spatio-temporal data
论文评审过程：Received 7 September 2012, Revised 11 February 2013, Accepted 3 June 2013, Available online 11 June 2013.
论文官网地址：https://doi.org/10.1016/j.imavis.2013.06.003
