Early diagnosis of plant diseases is of great importance for improving agricultural productivity and preventing crop losses. In this study, an AI-powered mobile system was developed for the early detection of leaf diseases in plants, which cause significant losses in agricultural production. The deep learning-based model was trained using the PlantVillage [1] dataset through the ResNet50 architecture with transfer learning and fine-tuning techniques. During the training process, methods such as data augmentation, normalization, and layer freezing were applied to optimize model performance. In the initial phase, only the classification layers were trained; then, the final layers of ResNet50 were unfrozen for fine-tuning. Based on the results obtained, the model achieved an accuracy rate of 97%, demonstrating strong generalization capability. The model was integrated into a mobile application using the FastAPI framework, and a user-friendly interface was developed with Flutter. The application analyzes leaf images uploaded via camera or gallery, diagnoses plant diseases, provides recommendations based on the diagnosis, and stores previous results to offer feedback to users. This system aims to contribute to agricultural sustainability and facilitate farmers' decision-making processes by enabling fast and accurate on-field detection of plant diseases.

Keywords - Plant Disease Detection, Image Processing, Deep Learning, Convolutional Neural Networks (CNN), Agricultural Productivity

Bitkilerde hastalıkların erken teşhisi, tarımsal verimliliğin artırılması ve ürün kayıplarının önlenmesi açısından büyük öneme sahiptir. Bu çalışmada, tarımsal üretimde önemli kayıplara yol açan bitki yaprak hastalıklarının erken teşhisine yönelik bir yapay zekâ destekli mobil sistem geliştirilmiştir. Derin öğrenme temelli model, PlantVillage [1] veri seti kullanılarak ResNet50 mimarisi üzerinden transfer öğrenme ve fine-tuning teknikleriyle eğitilmiştir. Eğitim süreci boyunca veri artırma, normalize etme ve katman dondurma gibi yöntemlerle modelin öğrenme başarımı optimize edilmiştir. Eğitimin ilk aşamasında yalnızca sınıflandırma katmanları eğitilmiş, ardından ResNet50'nin son katmanları açılarak ince ayar uygulanmıştır. Elde edilen sonuçlar doğrultusunda, modelin doğruluğu %97 seviyesine ulaşmıştır ve modelin genelleme kabiliyetinin yüksek olduğu görülmüştür. Model, FastAPI altyapısıyla mobil uygulamaya entegre edilmiş ve Flutter kullanılarak kullanıcı dostu bir arayüz geliştirilmiştir. Uygulama, kullanıcıların kamera veya galeri aracılığıyla yüklediği yaprak görsellerini analiz ederek hastalık teşhisi yapmakta, teşhis edilen hastalıklara göre öneriler sunmakta ve önceki sonuçları kayıt altına alarak kullanıcıya geri bildirim sağlamaktadır. Geliştirilen bu sistem, bitki hastalıklarının sahada hızlı ve doğru şekilde tespit edilmesine olanak tanıyarak, hem tarımsal sürdürülebilirliğe katkı sağlamayı hem de çiftçilerin karar destek süreçlerini kolaylaştırmayı hedeflemektedir.

Keywords^Turkish - Bitki Hastalığı Tespiti, Görüntü İşleme, Derin Öğrenme, Convolutional Neural Networks (CNN), Tarımsal Verimlilik

[1] A. Ali. (2019). PlantVillage Dataset. Available:
https://www.kaggle.com/datasets/abdallahalidev/plantvillage-dataset

[2] G. Lilienthal, N. Ahmad, and P. Hodgkinson, “The conflict between farmers rights in the Food and Agriculture Organization (FAO) treaty and corporate plant breeder laws,” Intellectual Property Forum, 2018.

[3] E.-C. Oerke and H.-W. Dehne, “Safeguarding production—losses in major crops and the role of crop protection,” Crop protection, vol. 23, pp. 275-285, 2004.

[4] J. Hearty, Advanced machine learning with Python: Packt Publishing Ltd, 2016.

[5] C. Topçu and P. Güneş, "Bitki hastalıklarını tespitte derin öğrenme: ResNet modelinin etkinliği," Anadolu Bil Meslek Yüksekokulu Dergisi, vol. 19, pp. 31-65, 2024.

[6] S. Lindow and R. Webb, "Quantification of foliar plant disease symptoms by microcomputer-digitized video image analysis," Phytopathology, vol. 73, pp. 520-524, 1983.

[7] D. P. Martin and E. P. Rybicki, "Microcomputer-based quantification of maize streak virus symptoms in Zea mays," Phytopathology, vol. 88, pp. 422-427, 1998.

[8] J. G. A. Barbedo, "A review on the main challenges in automatic plant disease identification based on visible range images," Biosystems engineering, vol. 144, pp. 52-60, 2016.

[9] J. Boulent, S. Foucher, J. Théau, and P.-L. St-Charles, "Convolutional neural networks for the automatic identification of plant diseases," Frontiers in plant science, vol. 10, p. 941, 2019.

[10] S. H. Lee, C. S. Chan, S. J. Mayo, and P. Remagnino, "How deep learning extracts and learns leaf features for plant classification," Pattern recognition, vol. 71, pp. 1-13, 2017.

[11] M. Loey, A. ElSawy, and M. Afify, "Deep learning in plant diseases detection for agricultural crops: a survey," International Journal of Service Science, Management, Engineering, and Technology (IJSSMET), vol. 11, pp. 41-58, 2020.

[12] N. Fu, C. Wang, and X. Ji, "Study on Visual Detection Device of Plant Leaf Disease," in 2019 IEEE International Conference on Mechatronics and Automation (ICMA), 2019, pp. 86-90.

[13] M. A. Iqbal and K. H. Talukder, "Detection of potato disease using image segmentation and machine learning," in 2020 international conference on wireless communications signal processing and networking (WiSPNET), 2020, pp. 43-47.

[14] F. Qin, D. Liu, B. Sun, L. Ruan, Z. Ma, and H. Wang, "Identification of alfalfa leaf diseases using image recognition technology," PloS one, vol. 11, p. e0168274, 2016.

[15] A. Ahmad, D. Saraswat, and A. El Gamal, "A survey on using deep learning techniques for plant disease diagnosis and recommendations for development of appropriate tools," Smart Agricultural Technology, vol. 3, p. 100083, 2023.

[16] S. Ltd. (2019). Agrio. Available: https://agrio.app/Tree-Doctor-App-Revolutionizing-Agronomy-with-AI/

[17] Bayer. (2022). MagicScout. Available: https://magicscout.app/en

[18] Mohanty, S. P., Hughes, D. P., & Salathé, M. (2016). Using deep learning for image-based plant disease detection. Frontiers in Plant Science, 7, 1419. https://doi.org/10.3389/fpls.2016.01419

[19] Brahimi, M., Boukhalfa, K., & Moussaoui, A. (2017). Deep Learning for Tomato Diseases: Classification and Symptoms Visualization. Applied Artificial Intelligence, 31(4), 299–315. https://doi.org/10.1080/08839514.2017.1315516

[20] Ferentinos, K. P. (2018). Deep learning models for plant disease detection and diagnosis. Computers and Electronics in Agriculture, 145, 311–318. https://doi.org/10.1016/j.compag.2018.01.009