Classification of Grapevine Leaves with LeNet and ResNet Deep Learning Models

Abstract

The main objective of this study is to investigate the effectiveness of deep learning techniques in accurately classifying grapevine leaves according to their species. The integration of LeNet and ResNet architectures has been used as a way to realise this classification. For the required dataset, a comprehensive collection of 500 grapevine leaf images representing five different species was used. A significant emphasis was placed on feature selection as a critical component to optimise classification performance. By careful selection of relevant features and elimination of redundant ones, it was aimed to improve the accuracy of the models used. By utilizing the LeNet-5 approach with the selected deep features, a good accuracy of 93.99% was achieved. This exceeded the performance of other state-of-the-art methods used for grapevine leaf classification. Based on these remarkable findings, several promising avenues for future research have been identified. These include the exploration of alternative deep learning architectures, a thorough investigation of various feature selection methods, and the extension of these techniques to cover the identification of leaves from other plant species.

Keywords

• LeNet-5
• ResNet
• Deep Learning
• Convolutional Neural Networks
• Image Processing

LeNet ve ResNet Derin Öğrenme Modelleri ile Asma Yapraklarının Sınıflandırması

Abstract

Bu çalışmanın temel amacı, asma yapraklarının türlerine göre doğru bir şekilde sınıflandırılmasında derin öğrenme tekniklerinin etkinliğini araştırmaktır. LeNet ve ResNet mimarilerinin entegrasyonu, bu sınıflandırmayı gerçekleştirmenin bir yolu olarak kullanılmıştır. Gerekli veri seti için, beş farklı türü temsil eden 500 asma yaprağı görüntüsünden oluşan kapsamlı bir koleksiyon kullanılmıştır. Sınıflandırma performansını optimize etmek için kritik bir bileşen olarak özellik seçimine önemli bir vurgu yapıldı. İlgili özelliklerin dikkatli bir şekilde seçilmesi ve gereksiz olanların ortadan kaldırılmasıyla, kullanılan modellerin doğruluğunun artırılması amaçlanmıştır. Seçilen derin özelliklerle birlikte LeNet-5 yaklaşımından yararlanılarak %93.99 gibi iyi bir doğruluk oranına ulaşılmıştır. Bu, asma yaprağı sınıflandırması için kullanılan diğer son teknoloji yöntemlerin performansını aşmıştır. Bu kayda değer bulgulara dayanarak, gelecekteki araştırmalar için umut verici birkaç yol belirlenmiştir. Bunlar arasında alternatif derin öğrenme mimarilerinin araştırılması, çeşitli özellik seçim yöntemlerinin kapsamlı bir şekilde incelenmesi ve bu tekniklerin diğer bitki türlerinden yaprakların tanımlanmasını kapsayacak şekilde genişletilmesi yer almaktadır.

Keywords

• LeNet-5
• Derin öğrenme
• Evrişimsel sinir ağları
• Görüntü İşleme

References

1. Koklu M, Unlersen MF, Ozkan IA, Aslan MF, Sabanci K. "A CNN-SVM study based on selected deep features for grapevine leaves classification". Measurement. 188,110425,2022.
2. Bharadi V, Mukadam AI, Panchbhai MN, Rode NN. "Image classification using deep learning". Int J Eng Res Technol. 2017.
3. Krishna MM, Neelima M, Harshali M, Rao MVG. "Image classification using deep learning". Int J Eng Technol, 7(2.7),614–7, 2018.
4. Hu B, Ergu D, Yang H, Liu K, Cai Y. "Document images classification based on deep learning". Procedia Comput Sci.,162,514–22,2019.
5. Yadav S, Sawale MD. "A review on image classification using deep learning". World J Adv Res Rev., 17(1),480–2, 2023.
6. You J. "Leaf Image Classification Using Deep Learning Network". Acad J Comput Inf Sci., 4(3),109–15, 2021.
7. Yang K, Zhong W, Li F. "Leaf segmentation and classification with a complicated background using deep learning". Agronomy, 10(11),1721, 2020.
8. Nguyen Thanh TK, Truong QB, Truong QD, Huynh Xuan H. "Depth learning with convolutional neural network for leaves classifier based on shape of leaf vein". In: Intelligent Information and Database Systems: 10th Asian Conference, ACIIDS 2018, Dong Hoi City, Vietnam, March 19-21, 2018, Proceedings, Part I 10. Springer, p. 565–75, 2018.

9. Liu Z, Zhu L, Zhang X-P, Zhou X, Shang L, Huang Z-K, et al. "Hybrid deep learning for plant leaves classification." In: Intelligent Computing Theories and Methodologies: 11th International Conference, ICIC 2015, Fuzhou, China, August 20-23, 2015, Proceedings, Part II 11. Springer, p. 115–23, 2015.
10. Minowa Y, Kubota Y. "Identification of broad-leaf trees using deep learning based on field photographs of multiple leaves". J For Res., 27(4),246–54, 2022.
11. Shah MP, Singha S, Awate SP. "Leaf classification using marginalized shape context and shape+ texture dual-path deep convolutional neural network". In: 2017 IEEE International conference on image processing (ICIP). IEEE, p. 860–4, 2017.
12. Sugata TLI, Yang CK. "Leaf App: Leaf recognition with deep convolutional neural networks". In: IOP Conference Series: Materials Science and Engineering. IOP Publishing. p. 12004, 2017.
13. Araújo VM, Britto AS, Brun AL, Koerich AL, Oliveira LES. "Fine-grained hierarchical classification of plant leaf images using fusion of deep models". In: 2018 IEEE 30th International Conference on Tools with Artificial Intelligence (ICTAI). IEEE, p. 1–5, 2018.
14. Beikmohammadi A, Faez K. "Leaf classification for plant recognition with deep transfer learning". In: 2018 4th Iranian Conference on Signal Processing and Intelligent Systems (ICSPIS). IEEE, p. 21–6, 2018.
15. Işık Ş, Özkan K. "Overview of handcrafted features and deep learning models for leaf recognition". J Eng Res., 9(1), 2021.
16. Hasan MA, Riana D, Swasono S, Priyatna A, Pudjiarti E, and Prahartiwi LI. “Identification of grape leaf diseases using convolutional neural network,” in Journal of Physics: Conference Series, vol. 1641, no. 1, p. 12007, 2020.
17. Nagaraju Y, Swetha S, and Stalin S. “Apple and grape leaf diseases classification using transfer learning via fine-tuned classifier,” in 2020 IEEE International Conference on Machine Learning and Applied Network Technologies (ICMLANT), pp. 1–6, 2020.
18. Ji M, Zhang L, and Wu Q. “Automatic grape leaf diseases identification via UnitedModel based on multiple convolutional neural networks,” Inf. Process. Agric., vol. 7, no. 3, pp. 418–426, 2020.
19. Liu B, Ding Z, Tian L, He D, Li S, and Wang H. “Grape leaf disease identification using improved deep convolutional neural networks,” Front. Plant Sci., vol. 11, p. 1082, 2020.
20. Padol PB and Yadav AA. “SVM classifier based grape leaf disease detection,” in 2016 Conference on advances in signal processing (CASP), pp. 175–179, 2016.
21. Ghoury S, Sungur C, and Durdu A. “Real-time diseases detection of grape and grape leaves using faster r-cnn and ssd mobilenet architectures,” in International conference on advanced technologies, computer engineering and science (ICATCES 2019), pp. 39–44, 2019.
22. Kayaalp K. “Classification of Medicinal Plant Leaves for Types and Diseases with Hybrid Deep Learning Methods,” Inf. Technol. Control, vol. 53, no. 1, pp. 19–36, 2024.
23. Kayaalp K. “A deep ensemble learning method for cherry classification,” Eur. Food Res. Technol., vol. 250, no. 5, pp. 1513–1528, 2024.