FDR: An Automated System for Finding Missing People

Abstract

According to the National Missing and Unidentified Persons (NamUS) database, on an average 600,000 individuals of all ages go missing every year. With the assistance of surveillance technology like CCTV cameras, our proposed model can help to make the search process easier and faster-using Face Detection and Recognition (FDR) techniques. It can be used for locating missing children, mentally challenged and old-aged individuals with Alzheimer's, and criminals by providing correct datasets. The face recognition model matches the face encodings of the uploaded image with the face encodings that are already available in the database to find a suitable match. Here we tend to use HAAR for face detection, LBP for face recognition, and a cloud system for storing the information. The disadvantage posed here is that generally they are susceptible to false detection and need parameter calibration. Further research will be conducted to overcome these drawbacks because achieving high accuracy is important.