作者: J. Rittscher , J. Kato , S. Joga , A. Blake
关键词: Filter (video) 、 Computer science 、 Expectation–maximization algorithm 、 Importance sampling 、 Joint Probabilistic Data Association Filter 、 Artificial intelligence 、 Hidden Markov model 、 Algorithm 、 Computer vision 、 Probabilistic logic 、 Motion estimation 、 Markov random field 、 Markov model 、 Particle filter
摘要: A new probabilistic background model based on a Hidden Markov Model is presented. The hidden states of the enable discrimination between foreground, and shadow. This functions as low level process for car tracker. particle filter employed stochastic use allows incorporation information from via importance sampling. novel observation density which models statistical dependence neighboring pixels random field effectiveness both likelihood are demonstrated.