Retraction: AI enabled smart surveillance system

Abstract

The conventional household door locking system has lot of drawbacks and it is still yet not resolved. Most of the security systems so far in our markets includes video surveillance or vigilance system. In order to improve security level facial recognition and object detection technique using CNN algorithms can be used which is also provides remote proctoring facilities to owners. The proposed system detects the object and identifies the anomalous activity near the door by applying Convolutional Neural Network. Electric door lock solenoid is used to unlock the door. An ultrasonic sensor is utilized to measure the distance between a person and door through the facial recognition when it reaches a certain threshold value that has been kept to detect the person reaching the doors and it tries to capture the human image only if it is mismatched from database. When a stranger try to access the door an alert message might be triggered to registered mobile number and he proprietor would be able to control the door locking system and inspect the image of person which has been mailed.