<![CDATA[Abstrak—Penelitian ini dimaksudkan untuk pengembangan model deteksi penyakit kulit pada sapi yang berbasis citra digital menggunakan arsitektur MobileNetV2. Dataset yang digunakan terdiri dari 1.319 citra dengan tiga kelas, yaitu LSD, Ringworms, dan Sapi Sehat. Data tersebut kemudian dibagi menjadi data latih, data validasi, dan data uji dengan rasio 70:20:10. Tahapan penelitian meliputi pra-pemrosesan citra, augmentasi data, pelatihan model, fine-tuning, dan evaluasi menggunakan accuracy, precision, recall, F1-score, serta confusion matrix. Hasil pengujian menunjukkan bahwa model memperoleh akurasi sebesar 75,56 persen dengan nilai test loss sebesar 0,6646. Kelas Sapi Sehat memperoleh performa terbaik dengan F1-score sebesar 0,86. Hasil penelitian menunjukkan bahwa MobileNetV2 dapat digunakan sebagai dasar sistem deteksi awal penyakit kulit sapi, meskipun peningkatan dataset dan kualitas citra masih diperlukan. Kata Kunci— citra digital, MobileNetV2, penyakit kulit sapi Abstract— This study is intended for the development of a cattle skin disease detection model based on digital images using the MobileNetV2 architecture. The dataset used consists of 1,319 images with three classes, namely LSD, Ringworms, and Healthy Cattle. The data were then divided into training data, validation data, and test data with a ratio of 70:20:10. The research stages include image preprocessing, data augmentation, model training, fine-tuning, and evaluation using accuracy, precision, recall, F1-score, and confusion matrix. The test results showed that the model achieved an accuracy of 75.56 percent with a test loss of 0.6646. The Healthy Cattle class achieved the best performance with an F1-score of 0.86. The results indicate that MobileNetV2 can be used as the basis for an early detection system for cattle skin diseases, although improvements in the dataset and image quality are still needed. Keywords— cattle skin disease, digital image, MobileNetV2]]>
Copyrights © 2026
