|
V11 N2 Paper 3
|
Annals of the MS in Computer Science and Information Systems at
UNC Wilmington
|
Fall 2017
|
EAmber Scan: Face Recognition for Missing Children in the National Amber Alert System
Harry Atwall
Committee
Abstract
Data published by the National Center for Missing and Exploited Children (NCMEC) shows how hundreds of thousands of children are reported missing every year. The Amber Alert
system was designed and implemented to address this problem but has failed to stay current with technological advancements. By utilizing mobile platforms, e.g. mobile phones, tablets, etc., and breaking technological improvements in computer vision and machine learning via deep artificial neural networks, the biometric modality face, can be leveraged to demonstrably reduce the painstaking manual process of triaging sightings. This work details a framework to implement face recognition into the current sighting process.
The Amber Alert system can be modernized with the use of an intelligent mobile application that leverages ubiquitous mobile platforms to generate actionable leads via biometrically
confirmed sightings. The mobile application will employ a RESTful architecture, which enables cross compatibility with many different front interfaces. Sightings are then analyzed through commercial-off-the-shelf (COTS) face recognition solutions and open source software to generate enhanced leads. An enhanced lead is demonstrably more informative than the current process comprised of emails, SMS, or phone calls reporting a physical sighting. The solution proposed in this work will provide an iOS application that will secure a photography of the suspected missing child complete with geolocation data, time stamp information, etc. and a face recognition match report to law enforcement. The proposed solution will provide officials with tools that will help to quickly and efficiently comb through hundreds of reportings/sightings to find actionable intelligence. Officials will be able to use the reports to view images of suspected sightings and focus location efforts.
This proposed solution is codified in an IOS mobile application with a set of actual missing child cases to demonstrate the efficacy of this solution over the current process.
download
(pdf)
Recommended Citation:
Atwall, H., Ricanek, K, Gebauer, J., Morago, B. (2017) EAmber Scan: Face Recognition for Missing Children in the National Amber Alert System. Annals of the Master of Science in Computer Science and Information Systems at UNC Wilmington, 11(2) paper 3. http://csbapp.uncw.edu/data/mscsis/full.aspx.
|
V11 N2 Paper 3
|
Annals of the MS in Computer Science and Information Systems at
UNC Wilmington
|
Fall 2017
|
