Kamera Pintar Untuk Pengawasan Penggunaan Masker Di Rumah Sakit

Abstract

 A Hospitals have a high potential as sources for the spread of infectious diseases, including viruses and bacteria. The use of masks by medical staff, patients, and visitors is a crucial measure to minimize the risk of infection. However, manually monitoring mask compliance in hospitals is challenging. With the development of artificial intelligence, automated monitoring systems can be implemented to more efficiently and effectively monitor mask compliance. This study employs YOLOv8 and IP Cameras for mask detection in hospitals. The dataset used consists of 2130 training images, 34 validation images, and 27 test images. The model was trained with parameters of 300 epochs, a batch size of 8, and a patience of 128 to prevent overfitting. Experimental results indicate that the model achieved a precision and recall of 98.3%, with an overall accuracy of 97%. The Precision-Recall and F1-Confidence curves demonstrate that the model is highly effective in detecting mask usage with minimal errors. The confusion matrix indicates that 97% of all mask usage cases were correctly detected, while only 3% were missed. This YOLOv8 and IP Camera-based mask detection system shows great potential for application in hospitals, enhancing mask compliance and aiding in the prevention of disease spread