The objective of this project is to develop a theft detection system based on facial recognition. The system would detect faces though a camera, process the faces and decide whether the face belongs to an authorized entity or not. If the person is not authorized to enter the room/area, the system would send an SMS to the authorized person through GSM module, and the authorized person would be able to surveillance the area using an Android application in real time though internet an internet connection. This report explores different techniques used for the recognition of faces. Different stages involving image processing will be studied and discussed. Finally the algorithm coded in python implemented in the Raspberry Pie module. The report discusses the methods used for the completion of the project and the flow of the system. The report goes into detail about various stages in image processing about what methodology is used for facial detection, and then recognition.