This paper presents a probabilistic approach for robust foreground segmentation that distinguishes moving objects from their moving cast shadows in indoor image sequences. Both foreground and shadow can be detected even in monocular grayscale sequences. To handle nonstationariness, the background, shadow, and edge models are set up and adaptively updated. A Bayesian framework is proposed to unify the various information including the segmentation label, background, intensity, and edge. The notion of Markov random field is used to encourage the spatial connectivity of the segmented regions. The solution is obtained by maximizing the posterior probability density of the segmentation field. Experiments on the test data show that our technique greatly improves the accuracy of segmentation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.