Abstract:

With the return to classes (universities, elementary and high schools), and the concern of contagion by Covid-19, especially these new strains of variants, establishes a context with increase level of concern. With this in mind, a low-cost prototype was developed that could be produced in any region of the planet without implementation complexity. The same involves an alcohol gel dispenser with a temperature gauge and facial recognition, where not only the student's face can be detected, but also if the student is wearing a protective mask as a physical barrier. Typically, hand sanitizer dispensers only make available hand sanitizer, and commercial-use temperature gauges only measure temperature.
In this prototype we tried to integrate three systems:
1) Alcohol gel dispenser,
2) Temperature gauge and
3) Face recognition and surgical mask or cloth identifier through the use of artificial intelligence.
The project is primarily intended for entrances to schools, universities, shopping malls, supermarkets and general commercial use. The solution consists of a rectangular container where the alcohol gel is placed in its upper base with a visible alcoholometer; the temperature measurement is on its lower base, and when you bring your hand closer to dispense the gel alcohol, the temperature measurement is measured and reported on an LCD display and an audible sound also informs the temperature. After collecting the temperature, the system asks the user to look at the camera that analyzes whether the user is wearing a mask or not; if the user is not wearing a mask, the system asks the user to put it on to enter the room, and audible sound communication was added with social inclusion in mind. The main concern of the system is the safety of students, especially in the school environment. Because of the implementation facilities, students will be able to make changes to the system, increasing their knowledge, as we are concerned with using low-cost processors and sensors in an open-source solution (that include a computer vision approach) to allowed students and teachers use them in their school curriculum, improving their knowledge in robotics. The system uses a Raspberry Pi 4 processor with 4 Gigas of RAM. The OpenCV library, Neural Networks and the Python program widely used in computer vision research. The operating system used was the Linux system, chosen for its versatility and the possibility that other researchers could also implement and improve its use and resource. The article details the mechanical way in which the prototype was developed, as well as the electrical parts, sensors and actuators, the software flowchart, the camera, the connections between the various devices and the processor, version of the libraries, the functioning of the Neural Network and the filters used to recognize the mask on the user's face. The research was developed thinking about the technical enrichment of students and teachers. The technical-scientific challenges of the development of this project are also detailed, as well as its possible improvements in the near future. This project presents a viable solution for a low-cost dispenser with built-in artificial intelligence, the mechanical design is simple and adaptable, opening up a range of other possibilities and functionalities, as it was designed and developed to adapt to academic, social or commercial needs.

Keywords:

Covid-19, school, health issues, artificial intelligence.