| Summary: | The purpose of this project is to develop an affordable loT-based solution that will improve the safety of the COVID-19 Standard Operating System. COVID-19 poses a significant risk to mankind. The COVID-19 virus is now being fought all over the world. Wearing masks is a sensible strategy for successfully managing COVID-19. This project's goal is to create a simple and efficient real-time monitoring paradigm. The proposed model successfully recognizes if an individual is wearing a face mask or not in a crowded place. The NVIDIA Jetson Nano was used to prototype a real-world scenario using the Object Detection Algorithm (YOLOv5) and an edge AI application: a smart camera capable of estimating the percentage of people wearing face masks in its field of view. The object detection algorithm YOLO, which stands for "You Only Look Once," divides images into a grid system. Each grid cell is in charge of detecting objects within itself. Because of its speed and accuracy, YOLO is one of the most well-known object detection algorithms. YOLOv5 need to be install of all the requirements into jetson nano itself. While this algorithm was accurate, it required too much processing power to run on the Jetson Nano in its current form. The algorithm have to simplified to use a singleshot object detection model as part of the Nano transition, recognizing the following object classes: face with mask, face without mask, and face not visible. The training process by using makesense.ai is a free-to-use online tool for labeling photos. It's ideal for small computer vision deep learning projects, as it simplifies and speeds up the dataset preparation process. Labels can be downloaded in one of the many formats that are supported.
|
|---|