A Probabilistic Framework Based on KDE-GMM Hybrid Model for Moving Object Segmentation in Dynamic Scenes

Overview & Synopsis

In real scenes, dynamic background and moving cast shadow always make accurate moving object detection difficult. In this paper, a probabilistic framework for moving object segmentation in dynamic scenes is proposed. Under this framework, we deal with foreground detection and shadow removal simultaneously by constructing probability density functions (PDFs) of moving objects and non-moving objects. Here, these PDFs are constructed based on KDEGMMhybrid model (KGHM) which has advantages of KDE and GMM. This KGHM models the spatial dependencies of neighboring pixel colors to deal with highly dynamic scenes. Moreover, in this framework, tracking information is used to refine the PDF of moving objects. Experimental results demonstrate the effectiveness of our method.