OPTIMASI MODEL EFFICIENTNETB0 UNTUK KLASIFIKASI PENYAKIT JAMBU BIJI MELALUI CLAHE DAN AUGMENTASI TEKSTUR

Iqbal; Martanto; Yudhistira Arie Wijaya; Mulyawan; Irfan Ali

Authors

Iqbal STMIK IKMI Cirebon, Indonesia
Martanto
Yudhistira Arie Wijaya
Mulyawan
Irfan Ali

Keywords:

EfficientNetB0, CLAHE, augmentasi tekstur, penyakit jambu biji, deep learning, klasifikasi citra

Abstract

Penelitian ini mengusulkan kerangka klasifikasi penyakit jambu biji berbasis EfficientNetB0 yang dioptimalkan melalui teknik Contrast Limited Adaptive Histogram Equalization (CLAHE) dan augmentasi tekstur. Dataset Guava Disease sebanyak 3.785 citra mencakup tiga kelas—Anthracnose, Fruit Fly, dan Healthy—dengan pembagian 70%:15%:15%. Tahapan prapemrosesan meliputi resize 224×224, normalisasi, CLAHE, serta augmentasi tekstur probabilistik menggunakan kombinasi LBP dan Gabor. Model EfficientNetB0 dilatih menggunakan transfer learning, batch size 32, 20 epoch, dan seed 42. Hasil menunjukkan bahwa integrasi CLAHE dan augmentasi tekstur meningkatkan performa model secara signifikan, dengan akurasi validasi 95,63% dan akurasi pengujian 94,71%. Metrik precision, recall, dan F1-score pada ketiga kelas mendekati nilai sempurna. Studi ablation membuktikan bahwa konfigurasi CLAHE + augmentasi tekstur menghasilkan kinerja terbaik dibanding model baseline, CLAHE-only, maupun texture-only. Temuan ini menegaskan bahwa normalisasi pencahayaan dan pengayaan tekstur sangat efektif menangani variasi iluminasi dan karakter bercak penyakit pada permukaan jambu biji. Model yang dioptimalkan ini berpotensi diterapkan dalam sistem pemantauan penyakit tanaman berbasis pertanian cerdas.

References

Ali, H. (2025). A fine tuned EfficientNet B0 convolutional neural network for leaf disease classification. Scientific Reports. https://doi.org/10.1038/s41598-025-04479-2

Archana, R., & Jeevaraj, P. S. (2024). Deep learning models for digital image processing: A review. Artificial Intelligence Review, 57. https://doi.org/10.1007/s10462-023-10631-z

Braik, M., Al Betar, M. A., Mahdi, M. A., Al Shalabi, M., Ahamad, S., & Saad, S. A. (2024). Enhancement of satellite images based on CLAHE and augmented elk herd optimizer. Artificial Intelligence Review, 58(2), 38. https://doi.org/10.1007/s10462-024-11022-8

González Briones, A. (2025). Enhancing Plant Disease Detection: Incorporating Deep Learning and Field Image Challenges. Plant Signal & Behavior. https://doi.org/10.1007/s44196-025-00835-2

Härtinger, P. (2024). Adaptive histogram equalization in constant time. Journal of Real Time Image Processing, 19(1), 76–88. https://doi.org/10.1007/s11554-024-01465-1

Law, C. X., Chen, D., & Li, Q. (2025). A review on anthracnose disease caused by Colletotrichum spp.: host range, epidemiology and economic impact on tropical fruits. Crop Protection, 178, 106279. https://doi.org/10.1016/j.cropro.2025.106279

Liu, Q., Xu, W., & Zhang, Y. (2023). Benchmarking data augmentation strategies for convolutional neural networks in agricultural image recognition. IEEE Access, 11, 45321–45337.

Shihab, M. R. (2025). Image dataset for classification of diseases in guava fruits: anthracnose, scab, styler-end-rot and leaf diseases. Data in Brief, 53, 109456. https://doi.org/10.1016/j.dib.2025.109456

Tan, M., & Le, Q. (2021). EfficientNetV2: Smaller Models and Faster Training. https://arxiv.org/abs/2104.00298

Upadhyay, A., Bhargava, A., & Gupta, P. (2025). Deep learning and computer vision in plant disease detection. Artificial Intelligence Review. https://doi.org/10.1007/s10462-024-11100-x

Xu, M. (2023). A comprehensive survey of image augmentation: taxonomy, relationships and trends. Image and Vision Computing, 135, 108479. https://doi.org/10.1016/j.image.2023.108479