PISS-IoT: Person Identification and Spotting System in an Internet-of-Things Way

Abstract

Missing persons has become one of the most serious issues in many countries around the world. It has been a fact to worry not only for those persons who lost their relatives but also for the law and order officials to search them and find them. So with this kind of problem statement in mind we have created a person identification and spotting system using the Internet-of-Things concept. Our system uses android mobile phone cameras and a server system coded with LBPH Facial recognition along with email sending functions. All these components are interconnected through Amazon AWS cloud in an Internet-of-Things fashion. This is a unique, novel, scalable and better approach, when compared to the previous works because instead of focusing on face detection and recognition in just a single device, our system involves interconnection of face detecting cameras with the face recognition and verification system which is running elsewhere. Also, our system includes the automatic email sending component interconnected with the identification system through cloud which has not been performed in any of the previous works. Experiments were conducted in real-time with real subjects at various locations around the university campus and the system worked very well where the average timings measured from the detection of face till notifying the user with the spotted location was within the range of 30 seconds to 1 minute and 40 seconds with an average value of accuracy ranging from 92 to 96 percent.