Real-time face mask detection on smart thermal body screening with Haar cascade classifier

During the pandemic of Covid-19, wearing face mask is required as part of standard operational procedure (SOP) in many countries including Malaysia. However, face mask compliance in premise entrances is normally checked manually which requires staff attendance. Therefore, it is of high interest to develop automated real-time face mask detection system integrated with IoT in a smart thermal body screening module. This paper focuses on the development of a simple-yet-efficient implementation of a real-time face mask detector based on the Haar cascade classifier for object detection. This mechanism uses two detector features, i.e. eye and face, to scan for the presence of masks. The system can alarm when detecting a person without wearing a face mask. The result shows a maximum of 6% false positive detection rate for eye and positive detection accuracy of various mask types while lacking in detecting improper mask usage.