Moving Object Detection Algorithm in Video Sequences
Huawei Wang · Acta Armamentarii · 2006
An efficient algorithm of detecting moving objects in video sequences was proposed.An adaptive background model was built and subsequently the background model is updated by a foreground object classification based background update algorithm which can resolve the deadlock problem efficiently.An improved shadow suppression algorithm which combines HSV color information with first-order image gradient information was exploited to segment shadows.Extensive experiments results on indoor and outdoor image sequences demonstrate that the proposed system can effectively detects moving objects and suppresses their shadows.