DEVELOPMENT OF MASK USE DETECTION SUBS SYSTEM USING AI AND NON-CONTACT TEMPERATURE DETECTION BASED ON EMBEDDED SYSTEM ON THE SMART GATE PROTOTYPE SYSTEM

• Main Author: Dwi Setya, Renzy
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/55583
• Institution: Institut Teknologi Bandung

COVID-19, which attacks respiratory infections, has caused a pandemic and Indonesia is one of the countries affected by the COVID-19 pandemic. The spread of this virus is very fast and can be spread in various ways such as droplets when someone sneezes and surfaces contaminated with the virus. To stop the spread of COVID-19, the Indonesian government enforced the PSBB and Keputusan Menteri regarding health protocols. The policy was made so that the impact on the economy, industry, etc. was over because people could carry out their normal life again. Some important points regarding health protocols are wearing masks, washing hands, maintaining distance and for managers of public places they need to check people body temperature when going into public places because one of the symptoms of someone suffering from COVID-19 is having a high body temperature. The implementation of health protocols in public places needs to be made effective and inexpensive so that a smart gate prototype system is created which features AI mask detection using the MobileNETv2 architecture and non-contact body temperature detection based on embedded system. The system can be accessed via the web and will display a real-time detection of mask use using AI which is equipped with a warning in the form of a sound when someone is not wearing a mask. The web also displays the results of temperature detection which is equipped with distance detection as a warning if the user is too far from the sensor. The smart gate prototype system has a mask detection accuracy of 99% and body temperature detection accuracy of ± 0.2oC with measurement results using the YNA-800 thermogun.