A novel IOT based approach to establish an ultra-low power self security system
Mary L Monisha, Pooja S Mohan · 2017
Rape cases and similar violation are nowadays a leading menace to women and children. An important step to eradicate this problem is to monitor and support them in the case of a crisis. To endure with this situation, an intelligent camera system is introduced to capture and transmit the video of unlawful activities. The system acts as an alert system and also as a location tracker. The embedded system implemented here is adopted to provide the system with fast processing rate, accuracy and low power. The proposed system adopts a hidden camera system that makes smart decisions and becomes active on recognizing human postures and hand gestures maintained in the database. On processing and recognizing gestures by the frontend of the camera system, the system starts recording the video and sends an alert message to the authorized officials. As the self security system is supported with IOT environment, it enables the system to transmit the video to the cloud platform. And the anti-social activities could be monitored through smartphone/PC. Moreover, the hidden Markov and Baum-Welch algorithm is adopted to classify, train and test the system. In order to enable an ultra-low power operation, the system is implemented using ARM cortex A8 processor with optimized algorithm.