<![CDATA[The skin is a vital human organ located on the outermost part of the body and is vulnerable to various external stimuli and diseases. The high prevalence of skin diseases in Indonesia indicates a lack of public awareness regarding skin health. This study aims to develop a web-based application capable of detecting 10 types of skin diseases quickly and accurately using the ResNet50 architecture and computer vision technology. The research stages include the creation of a Haar Cascade Classifier, developing an image classification model, model evaluation, chatbot development, system design, implementation, and application testing. The results show that the model achieved an accuracy of 90.10% on the training data and 89.06% on the validation data. The integrated chatbot also provided additional information with a response accuracy of 87.50%. System testing demonstrated good performance based on black-box testing and scored 77.625 on the System Usability Scale (SUS), which falls into the "Good" category. This application can detect early skin disease without requiring direct consultation with a doctor.]]>
Copyrights © 2025
