Home security has been a concern of worldwide. The need of security systems is considered as one of the important aspects of our modern life, people are using such systems to be aware about anything could be happen in their places like their houses when they are away, these system also let them feel more secure and that is what the people really need. As the technology is developing, rich home based security systems are implemented to ensure safety and security of. Video surveillance is an important area of computer vision research, its applications including both outdoor and indoor automated surveillance systems Home security system is an essential mean of protecting our home from illegal invasion. A conventional home security system consists of a Closed Circuit Television, CCTV and burglar alarm which can be replaced by computer based systems with the capability of smarter detection and alert system. In the context of smart home environments, the In-House Video Surveillance systems have as main goal to control the safety and the security of materials and of people living in a domestic environment. This paper provides a comparative approach between tracking learning and detection algorithm with illumination sensitive background model. Further a face recognition module is implemented which has the capability of identify intruder. On finding the intruder, the system sends an email on the owner mail id with images of the intruders so that the further action can be initiated.