Real-time nonparametric background subtraction with tracking-based foreground update

摘要

Highlights?Combined foreground-background spatio-temporal nonparametric models with tracking-based update of reference data.?Bayesian classifier for combining foreground and background models with different spatial bandwidths.?Selective analysis strategy based on random sampling and regions of interest.?Efficient automatic appearance bandwidth selection switching.?Real-time, GPU-based implementation of the proposed strategy.AbstractA nonparametric real-time and high-quality moving object detection strategy in a GPU is proposed. To improve the quality of the results in sequences where the moving objects and the background have similar appearance, not only the background but also the foreground is modelled. Both models are constructed from spatio-tempo