Proper

Abstract

In the present era, security is a primary concern in all facets of life; therefore, there is a strong need of an efficient security system. There are many existing systems, which perform surveillance by way of motion detection only. The proposed intruder detection system is a security system that performs detection through motion as well as integrates a biometric identification technique (i.e. front face recognition for increased accuracy). The use of facial recognition when compared to motion detection is a more efficient method as it can be implemented for larger distances and there is no need for equipment to record or compare the results. In this system, a passive infrared (PIR) sensor is used to detect motion. Once detected, the sensor will send a signal to raspberry pi to activate the web camera, which will capture an image of the activity. The captured image will then be processed and if any facial feature is detected, then facial recognition and detection algorithms will be run in order to identify the face. If no message will be given indicating that activity was detected but no facial features were found and an email notification is also sent to the owner. Regardless of content, all images captured by the camera are stored on a local disk (i.e. raspberry pi) as well as on the cloud (i.e. Google drive).