Smart Anti-Burglary Surveillance System using Image AI

Abstract

In today's world, with an increasing number of robberies, looting and shoplifting, it is indispensable to ensure the safety of our belongings and valuables. When it comes to certain places like bank vaults, uninhabited houses, business establishments like retail shops, jewellery stores, watch stores, etc., surveillance for observing the premise is very essential and has become the unavoidable need of the hour. At such places, it is unlikely to continuously monitor the venue using conventional cameras or a security guard person. To address such type of issues which involve uncertainty of safety, our project proposes a system which will tirelessly monitor the venue at required hours, intimate the intended person (administrator) via mobile communication (Short Message Service), initiate an alert alarm at the premise in order to aware the surroundings and also save the intrusion detected frames in the specified choice of storage. Also, it is developed in such a manner that the system recognizes and does not initiate alert/notification for persons those who are reliable and trustworthy (such as the shopkeeper, habitant of the house, etc.). This system is implemented with specific considerations using Image AI and Tensor flow built upon Convolutional Neural Networks, remotely operatable by users, without compromising the importance of security and protection.