Identifikasi Luka Ringan dengan YOLO-CNN pada Citra Digital untuk Manajemen Perawatan Luka

Abstract

Minor wounds are a common issue in everyday life often overlooked, yet proper management of minor wounds is crucial to prevent more serious complications. This research aims to implement the You Only Look Once (YOLO) object detection method utilizing Convolutional Neural Networks (CNN) architecture in identifying minor wounds in digital images, while also providing relevant information regarding wound care management such as initial treatment and suitable medication recommendations. This study involves literature review stages to understand the basic concepts of Machine Learning and YOLOv8, collecting a dataset of digital images covering various types of minor wounds, data preprocessing to normalize and enhance dataset diversity, developing the YOLOv8 model for minor wound identification, evaluating model performance, developing a digital image-based application for minor wound identification, and overall application evaluation. Testing was conducted on the application model utilizing the confusion matrix evaluation method, with a test data set consisting of 49 images of minor wounds. An accuracy of 85% was obtained.