An　approach　to　detect　and　track　moving　objects　in　indoor　environment　is　presented　in　this　paper.　HSI　model　is　first　adjusted　for　better　motion　detection;　then　based　on　the　improved　HSI　model,　the　background　subtraction　and　the　frame　difference　are　effectively　used　to　overcome　light　changing　and　to　detect　moving　objects　in　indoor　environment;　finally　based　on　Camshift　and　kalman　filter,　the　detected　objects　are　tracked　in　real　time.　With　video　recorded　ahead　experiments　are　realized　and　the　results　validate　that　the　approach　presented　in　this　paper　is　feasible　and　reliable.　©　2011　IEEE.