Mask Detection System with Computer Vision-Based on CNN and YOLO Method Using Nvidia Jetson Nano

Abstract

Health is an essential aspect of life. The World Health Organization (WHO) has officially declared the Corona Virus (Covid-19) a global pandemic that has spread to Indonesia. For preventive measures against Covid-19, the Indonesian government is trying to deal with the Covid-19 pandemic with 3M health protocol aimed at community activities, such as Memakai Masker (wearing masks), Mencuci Tangan (washing hands), and Menjaga Jarak (maintaining distance). In this study, software and hardware design was carried out to detect mask users and immediately warn violators who do not use masks automatically and can function automatically offline by utilizing digital image processing using NVIDIA Jetson Nano using the YOLO (You Only Look Once) method. The CNN YOLOv4-tiny model is chosen to obtain measurement results for mask user detection accuracy because it has a relatively minor computational value and is faster. The best camera detection angle is obtained at a vulnerable angle of 45O-90O or in the range of 90O-135O with value confidence that the average is 99.94% and the best accuracy is at a lux value greater than 70, and a minimum camera height of 1 meter and a maximum of 3 meters. Under conditions of lux 96 (bright), the maximum distance for detecting a face object is 12 meters, and the ability of the system to output a warning sound has been successfully integrated with a relay to run the mp3 module separately from the system, so as not to interfere with the Jetson Nano computation process and the model is successfully run on the Jetson Nano with an average computation of 13 frames per second.

A Comprehensive Analysis of Nvidia's Technological Innovations, Market Strategies, and Future Prospects

“…The study provides valuable insights for developers seeking enhanced performance in GPU database applications. Hakim et al (2023) offered a novel Nvidia Jetson Nano-powered system designed to enhance mask enforcement during the pandemic. By combining the YOLO algorithm with mask detection capabilities, this system achieves an impressive 99.94% accuracy.…”

Classification of crown density and foliage transparency scale for broadleaf tree using VGG-16

“…Visual Geometry Group (VGG-16) is an AlexNet that has undergone modifications to the convolution filter, where the filter used is 3x3 smaller than the 5x5 or 11x11 filters [9]. AlexNet is a groundbreaking deep neural network developed in 2012 for image classification and won the ImageNet LSFRC-2010 competition with 62 million trainable parameters [29].…”

Performance Comparison Between LeNet And MobileNet In Convolutional Neural Network for Lampung Batik Image Identification

“…There are several past research addressing the same problem of batik patterns classification, for example Deep Learning approach [18]. Deep Learning(DL)as AI technology (Artificial Intelligence) which became a hot topic, and many creating a detection of objects, faces, and several other types of health innovation development [19]. In this context, Convolutional Neural Network (CNN) has garnered attention for its ability to recognize patterns in visual images such as pictures [20].…”