Dynamic temporal analysis for early detection and intervention strategies of tomato bacterial leaf diseases

---

Şanver U., Pınarer Ö., Tekin A. B.

15th European Conference on Precision Agriculture, Barcelona, İspanya, 29 Haziran - 03 Temmuz 2025, ss.29-37, (Tam Metin Bildiri)

• Yayın Türü: Bildiri / Tam Metin Bildiri
• Doi Numarası: 10.1163/9789004725232_002
• Basıldığı Şehir: Barcelona
• Basıldığı Ülke: İspanya
• Sayfa Sayıları: ss.29-37
• Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
• Galatasaray Üniversitesi Adresli: Evet

Özet

Early detection of tomato bacterial leaf diseases is vital for sustainable agriculture. This study proposed a hybrid model combining machine learning and deep learning techniques for accurate detection and prediction. Focused on Pseudomonas syringae pv. tomato and Xanthomonas euvesicatoria, the system integrated ML structured feature extraction with DL hierarchical learning through a stacking strategy. Transfer learning enhanced robustness by fine-tuning pre-trained models, while automated feature selection optimized performance. Experiments indicated that the hybrid model outperformed individual ML and DL approaches, offering a scalable and effective tool for early disease management.