<![CDATA[The increasing number of monkeypox cases has become a global health issue that de-mands rapid and accurate diagnosis. In Indonesia, monkeypox cases reached 88 victims as of August, 2024. The complex symptoms of monkeypox, which often resemble those of other diseases, require advanced technology to distinguish them in a short amount of time. The purpose of this research is to enhance the accuracy and efficiency of the model in identifying monkeypox skin lesions automatically and quickly, thereby supporting the medical diagnosis process more effectively. This study proposes an innovative approach by combining Gray Level Co-occurrence Matrix (GLCM) feature extraction with deep learning architectures ResNet50 and VGG16 for detecting monkeypox in skin lesion im-ages The results show a significant improvement in classification accuracy for both Res-Net50 and VGG16. The GLCM-VGG16 model achieved an accuracy of 95.75%, an im-provement of 18.57% from its original 77.18% without GLCM features. The GLCM-ResNet50 model reached an accuracy of 98.07%, marking a 44.82% increase from the initial 53.25%. The training time of models with GLCM features was also faster compared to models without GLCM. The integration of GLCM successfully captured unique texture characteristics in monkeypox lesions, thereby enhancing the model's ability to distinguish them from other skin diseases. These findings indicate that the combination of GLCM with CNN architectures can be an effective approach for accurately and efficiently detect-ing skin diseases.]]>
Copyrights © 2025
