A Hyper-Graph Embedded Bandlet-Based Facial Emotion Monitoring System for Enhanced Urban Health

The state of health of a person can affect their facial expressions. As a result, a system that recognizes facial expressions can be beneficial for healthcare services. In this study, a Facial-Expression Recognition system has been developed to improve healthcare in smart cities by extracting features from a face image through a bandlet transform and Center-Symmetric Local Binary Pattern (CS-LBP). The most prominent features are selected using a Feature-Selection algorithm and then provided to two classifiers, Gaussian mixture model and support vector machine, to determine the facial expression with a confidence score that is calculated from the combined ratings of the classifiers. The proposed system has been tested with large data sets and found to have an accuracy of 99.5% in identifying facial expressions.