Pamuk Yaprağı Hastalıklarının Tespiti için DenseNet201 ve ResNet50 Tabanlı Yeni Bir Derin Özellik Çıkarım Yaklaşımı

Öz

Bu çalışmada, pamuk bitkisi yapraklarında görülen hastalıkların otomatik tespiti için yeni bir derin özellik çıkarımı yaklaşımı önerilmiştir. Önerilen yaklaşımda, DenseNet201 ve ResNet50 derin öğrenme mimarileri kullanılarak özellik çıkarımı gerçekleştirilmiş ve elde edilen özellik vektörleri birleştirilmiştir. Daha sonra, Yinelemeli Chi2 algoritması ile en bilgilendirici özellikler seçilmiş ve Support Vector Machine (SVM) sınıflandırıcısı kullanılarak hastalık tespiti yapılmıştır. Geliştirilen model, 2137 pamuk yaprağı görüntüsünden oluşan ve 7 farklı sınıf içeren (1 sağlıklı, 6 hastalıklı) açık erişimli bir veri seti üzerinde test edilmiştir. Test aşamasında 10-fold cross-validation ve 80:20 hold-out cross-validation stratejileri uygulanmıştır. Herhangi bir veri artırma tekniği kullanılmadan gerçekleştirilen testler sonucunda, sırasıyla %97,29 ve %96,96 sınıflandırma doğruluğu elde edilmiştir. Önerilen yaklaşım, dengesiz veri seti üzerinde yüksek başarı göstermesi ve hesapsal olarak lightweight bir mimari sunması açısından literatüre önemli katkılar sağlamaktadır.

Anahtar Kelimeler

• Derin özellik çıkarımı
• Pamuk yaprağı hastalıkları
• Özellik seçimi
• Makine öğrenmesi
• Hafif siklet sınıflandırma modeli

A Novel Deep Feature Extraction Approach Based on DenseNet201 and ResNet50 for Cotton Leaf Disease Detection

Abstract

In this study, a new deep feature extraction approach is proposed for automatic detection of diseases observed on cotton plant leaves. In the proposed approach, feature extraction is performed using DenseNet201 and ResNet50 deep learning architectures. and the obtained feature vectors are combined. Then, the most informative features are selected with the Iterative Chi2 algorithm, and disease detection is performed using the Support Vector Machine (SVM) classifier. The developed model is tested on an open access dataset consisting of 2,137 cotton leaf images and 7 different classes (1 healthy, 6 diseased). 10-fold cross-validation and 80:20 hold-out cross-validation strategies are applied in the testing phase. As a result of the tests performed without using any data augmentation technique, 97.29% and 96.96% classification accuracies are obtained, respectively. The proposed approach makes significant contributions to the literature in terms of showing high success on the imbalanced dataset and providing a computationally lightweight architecture.

Keywords

• Deep feature extraction
• Cotton leaf diseases
• Feature selection
• Machine learning
• Lightweight classification model

References

1. [1] R. F. Caldeira, W. E. Santiago, and B. Teruel, “Identification of cotton leaf lesions using deep learning techniques,” Sensors, vol. 21, no. 9, 2021, doi: 10.3390/s21093169.
2. [2] P. R. Rothe and R. V. Kshirsagar, “Cotton leaf disease identification using pattern recognition techniques,” 2015 Int. Conf. Pervasive Comput. Adv. Commun. Technol. Appl. Soc. ICPC 2015, vol. 00, no. c, pp. 1–6, 2015, doi: 10.1109/PERVASIVE.2015.7086983.
3. [3] B.-A. S. Bashish Al Dheeb,Braik Malik, “Detection and Classification of Leaf Diseases using K-means-based Segmentation and Neural-networks-based Classification,” vol. 10, 2011.
4. [4] T. Rumpf, A. K. Mahlein, U. Steiner, E. C. Oerke, H. W. Dehne, and L. Plümer, “Early detection and classification of plant diseases with Support Vector Machines based on hyperspectral reflectance,” Comput. Electron. Agric., vol. 74, no. 1, pp. 91–99, 2010, doi: 10.1016/j.compag.2010.06.009.
5. [5] C. Hillnhütter and A. K. Mahlein, “Neue Ansätze zur frühzeitigen Erkennung und Lokalisierung von Zuckerrübenkrankheiten,” Gesunde Pflanz., vol. 60, no. 4, pp. 143–149, 2008, doi: 10.1007/s10343-008-0196-0.
6. [6] A. Jenifa, R. Ramalakshmi, and V. Ramachandran, “Cotton Leaf Disease Classification using Deep Convolution Neural Network for Sustainable Cotton Production,” 2019 Int. Conf. Clean Energy Energy Effic. Electron. Circuit Sustain. Dev. INCCES 2019, pp. 19–21, 2019, doi: 10.1109/INCCES47820.2019.9167715.
7. [7] B. S. Prajapati, V. K. Dabhi, and H. B. Prajapati, “A survey on detection and classification of cotton leaf diseases,” Int. Conf. Electr. Electron. Optim. Tech. ICEEOT 2016, pp. 2499–2506, 2016, doi: 10.1109/ICEEOT.2016.7755143.
8. [8] N. Parashar and P. Johri, “Deep Learning for Cotton Leaf Disease Detection,” Proc. - 2nd IEEE Int. Conf. Device Intell. Comput. Commun. Technol. DICCT 2024, no. Dl, pp. 158–162, 2024, doi: 10.1109/DICCT61038.2024.10533021.

9. [9] M. M. Islam et al., “A deep learning model for cotton disease prediction using fine-tuning with smart web application in agriculture,” Intell. Syst. with Appl., vol. 20, no. January, p. 200278, 2023, doi: 10.1016/j.iswa.2023.200278.
10. [10] M. Azath, M. Zekiwos, and A. Bruck, “Deep Learning-Based Image Processing for Cotton Leaf Disease and Pest Diagnosis,” J. Electr. Comput. Eng., vol. 2021, 2021, doi: 10.1155/2021/9981437.
11. [11] C. K. Rai and R. Pahuja, “Classification of Diseased Cotton Leaves and Plants Using Improved Deep Convolutional Neural Network,” Multimed. Tools Appl., vol. 82, no. 16, pp. 25307–25325, 2023, doi: 10.1007/s11042-023-14933-w.
12. [12] A. Kaur, V. Kukreja, M. Kumar, A. Choudhary, and R. Sharma, “A Fine-tuned Deep Learning-based VGG16 Model for Cotton Leaf Disease Classification,” 2024 5th Int. Conf. Emerg. Technol. INCET 2024, pp. 1–6, 2024, doi: 10.1109/INCET61516.2024.10593164.
13. [13] H. Kukadiya, N. Arora, D. Meva, and S. Srivastava, “An ensemble deep learning model for automatic classification of cotton leaves diseases,” Indones. J. Electr. Eng. Comput. Sci., vol. 33, no. 3, pp. 1942–1949, 2024, doi: 10.11591/ijeecs.v33.i3.pp1942-1949.
14. [14] B. Kavinandhan, R. Pranav, and M. Ganesan, “A Hybrid Approach for Cotton Leaf Disease Detection using DCGAN and Diverse CNN Models,” 2024 5th Int. Conf. Innov. Trends Inf. Technol. ICITIIT 2024, pp. 1–8, 2024, doi: 10.1109/ICITIIT61487.2024.10580758.
15. [15] C. K. Rai and R. Pahuja, An ensemble transfer learning-based deep convolution neural network for the detection and classification of diseased cotton leaves and plants, no. 0123456789. Springer US, 2024. doi: 10.1007/s11042-024-18963-w.
16. [16] A. Saleh et al., “Machine Learning-based classification of cotton diseases using mobilenet and Support Vector Machine,” 2024 Int. Telecommun. Conf. ITC-Egypt 2024, pp. 165–171, 2024, doi: 10.1109/ITC-Egypt61547.2024.10620532.
17. [17] W. S. Noble, “What is a support vector machine?,” Nat. Biotechnol., vol. 24, no. 12, pp. 1565–1567, 2006.
18. [18] P. Bishshash, M. A. S. Nirob, M. H. Shikder, M. A. H. Sarower, D. T. Bhuiyan, and S. R. H. Noori, “A Comprehensive Cotton Leaf Disease Dataset for Enhanced Detection and Classification,” Data Br., vol. 57, p. 110913, 2024, doi: 10.1016/j.dib.2024.110913.
19. [19] G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 4700–4708.
20. [20] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778.
21. [21] M. Erten and T. Tuncer, “Automated differential diagnosis method for iron deficiency anemia and beta thalassemia trait based on iterative Chi2 feature selector,” Int. J. Lab. Hematol., vol. 44, no. 2, pp. 430–436, 2022.