Dinamik Oylama Tabanlı Topluluk Derin Öğrenme ile Benzer Mahsullerin Sınıflandırılması

Öz

Güncel derin öğrenme tekniklerinin görüntü işleme ve veri sınıflandırma için kullanılması, önemli sonuçlar elde etme açısından büyük bir potansiyel barındırmaktadır. Bu doğrultuda, derin öğrenme tabanlı yaklaşımlar tarımsal alanlarda, özellikle görüntü sınıflandırma ve veri analizi gibi görevlerde başarılı uygulamalar sergilemektedir. Bu çalışma, genellikle birbirine çok benzeyen tarımsal mahsul görüntülerini sınıflandırmayı amaçlayarak 17 farklı derin öğrenme modeli ve dinamik oylama yöntemini kullanmaktadır. Çalışmanın başlangıcında veri seti ve kullanılan derin öğrenme yöntemlerine ilişkin kısa bir genel bakış sunulmaktadır. Ardından eğitim ve test aşamaları etkili bir şekilde yürütülmektedir. Veri seti, kenevir, mısır, pirinç, şeker kamışı ve buğday olmak üzere beş farklı mahsul türüne ait toplam 804 görüntüden oluşmaktadır. Güvenilirliği sağlamak amacıyla 10 katlı çapraz doğrulama kullanılmış ve tüm modellerde aynı örnek setleriyle tutarlı deneyler gerçekleştirilmiştir. Elde edilen sonuçlar, hangi modellerin tarımsal mahsul görüntülerini daha doğru bir şekilde tespit edip sınıflandırabildiğini rapor etmektedir. Ayrıca önerilen topluluk yaklaşımı doğruluğu artırmakla kalmayıp daha fazla sağlamlık ve kararlılık sağlamaktadır. Deneysel bulgulara göre, Shufflenet test setinde %98,63 ile en yüksek bireysel doğruluğa ulaşmış, ancak topluluk yaklaşımı bu değeri %99,75'e yükseltmiştir.

Anahtar Kelimeler

• Derin Öğrenme
• Dinamik Oylama
• Mahsul Sınıflandırma

Dynamic Voting-Based Ensemble Deep Learning for Closely Resembling Crop Classification

Abstract

Utilizing modern deep learning techniques for image processing and data classification holds immense promise for yielding significant outcomes. Consequently, deep learning-based approaches have demonstrated successful applications in agricultural contexts, particularly in tasks such as image classification and data analysis. This study focuses on classifying agricultural crop images, which often bear close resemblance, employing 17 distinct deep learning models and dynamic voting. Initially, the paper provides a concise overview of the dataset and the deep learning methodologies employed. Subsequently, the training and testing phases are carried out effectively. The dataset comprises a total of 804 images depicting five different types of crops: jute, maize, rice, sugarcane, and wheat. To ensure robustness, 10-fold cross-validation is employed, and experiments are conducted consistently across all models using the same sample sets. The results obtained report which models can more accurately detect and classify agricultural crop images. Further, the proposed ensemble approach improves accuracy and ensures greater robustness and stability. According to the experimental findings, Shufflenet achieved the highest individual accuracy of 98.63% on the test set, but the ensemble approach increased this value to 99.75%.

Keywords

• Deep Learning
• Dynamic Voting
• Crop Classification

Kaynakça

1. [1] Gray H, Nuri KR. Differing Visions of Agriculture: Industrial-Chemical vs. Small Farm and Urban Organic Production. American Journal of Economics and Sociology. 2020; 79(3): 813-832. doi:10.1111/ajes.12344
2. [2] Bouguettaya A, Zarzour H, Kechida A, Taberkit AM. Deep learning techniques to classify agricultural crops through UAV imagery: A review. Neural computing and applications. 2022;34(12):9511-9536.
3. [3] Ngugi HN, Akinyelu AA, Ezugwu AE. Machine Learning and Deep Learning for Crop Disease Diagnosis: Performance Analysis and Review. Agronomy. 2024;14(12):3001.
4. [4] Jámbor A, Czine P, Balogh P. The impact of the coronavirus on agriculture: first evidence based on global newspapers. Sustainability. 2020;12(11):4535.
5. [5] Kamilaris A, Prenafeta-Boldú FX. Deep learning in agriculture: A survey. Comput Electron Agric. 2018;147:70-90.
6. [6] Abdullahi HS, Sheriff R, Mahieddine F. Convolution neural network in precision agriculture for plant image recognition and classification. In: 2017 Seventh International Conference on Innovative Computing Technology (INTECH). Vol 10 ; 2017:256-272.
7. [7] Yu S, Jia S, Xu C. Convolutional neural networks for hyperspectral image classification. Neurocomputing. 2017;219:88-98.
8. [8] Mohanty SP, Hughes DP, Salathé M. Using deep learning for image-based plant disease detection. Frontiers in plant science. 2016;7:1419.

9. [9] Ishimwe R, Abutaleb K, Ahmed F, others. Applications of thermal imaging in agriculture—A review. Advances in remote Sensing. 2014;3(03):128.
10. [10] Saxena L, Armstrong L. A survey of image processing techniques for agriculture. Published online 2014.
11. [11] Feng F, Gao M, Liu R, Yao S, Yang G. A deep learning framework for crop mapping with reconstructed Sentinel-2 time series images. Comput Electron Agric. 2023;213:108227.
12. [12] Pratama IPA, Atmadji ESJ, Purnamasar DA, Faizal E. Evaluating the performance of voting classifier in multiclass classification of dry bean varieties. Indonesian Journal of Data and Science. 2024;5(1):23-29.
13. [13] Schmidhuber J. Deep learning in neural networks: An overview. Neural networks. 2015;61:85-117.
14. [14] Pan SJ, Yang Q. A survey on transfer learning. IEEE Trans Knowl Data Eng. 2009;22(10):1345-1359.
15. [15] Mohammed A, Kora R. A comprehensive review on ensemble deep learning: Opportunities and challenges. Journal of King Saud University-Computer and Information Sciences. 2023;35(2):757-774.
16. [16] Amara J, Bouaziz B, Algergawy A. A deep learning-based approach for banana leaf diseases classification. Published online 2017.
17. [17] Mohammed A, Kora R. An effective ensemble deep learning framework for text classification. Journal of King Saud University-Computer and Information Sciences. 2022;34(10):8825-8837.
18. [18] Kim HC, Pang S, Je HM, Kim D, Bang SY. Constructing support vector machine ensemble. Pattern Recognit. 2003;36(12):2757-2767.
19. [19] Montgomery JM, Hollenbach FM, Ward MD. Improving predictions using ensemble Bayesian model averaging. Political Analysis. 2012;20(3):271-291.
20. [20] Latif-Shabgahi GR. A novel algorithm for weighted average voting used in fault tolerant computing systems. Microprocess Microsyst. 2004;28(7):357-361.
21. [21] Atay Y, Yildirim MO, Dogan CU. High performance classification of cancer types with gene microarray datasets: hybrid approach. Gazi University Journal of Science Part C: Design and Technology. 2021;9(4):811-827.
22. [22] Ayan E, Erbay H, Varç\in F. Crop pest classification with a genetic algorithm-based weighted ensemble of deep convolutional neural networks. Comput Electron Agric. 2020;179:105809.
23. [23] Chen M, Wang J, Chen Y, Guo M, Zheng N. Weight-based ensemble method for crop pest identification. Ecol Inform. Published online 2024:102693.
24. [24] Shahid MF, Khanzada TJS, Aslam MA, Hussain S, Baowidan SA, Ashari RB. An ensemble deep learning models approach using image analysis for cotton crop classification in AI-enabled smart agriculture. Plant Methods. 2024;20(1):104.
25. [25] Hyder U, Talpur MRH. Detection of cotton leaf disease with machine learning model. Turkish Journal of Engineering. 2024;8(2):380-393. doi:10.31127/tuje.1406755
26. [26] Nanni L, Maguolo G, Pancino F. Insect pest image detection and recognition based on bio-inspired methods. Ecol Inform. 2020;57:101089.
27. [27] Varma WK, Kumar V. Analysis of Crop Leaf Image Classification using Deep Learning Models over Novel Dataset. In: 2023 10th International Conference on Signal Processing and Integrated Networks (SPIN). 2023:14-19.
28. [28] Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks. Adv Neural Inf Process Syst. 2012;25.
29. [29] Alom MZ, Taha TM, Yakopcic C, et al. The history began from alexnet: A comprehensive survey on deep learning approaches. arXiv preprint arXiv:180301164. Published online 2018.
30. [30] Al-Haija QA, Smadi M, Al-Bataineh OM. Identifying phasic dopamine releases using darknet-19 convolutional neural network. In: 2021 IEEE International IOT, Electronics and Mechatronics Conference (IEMTRONICS). 2021:1-5.
31. [31] Rajkumar R, Gopalakrishnan S, Praveena K, Venkatesan M, Ramamoorthy K, Hephzipah JJ. Darknet-53 convolutional neural network-based image processing for breast cancer detection. Mesopotamian Journal of Artificial Intelligence in Healthcare. 2024;2024:59-68.
32. [32] Huang G, Liu Z, Van Der Maaten L, Weinberger KQ. Densely connected convolutional networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017:4700-4708.
33. [33] Tan M, Le Q. Efficientnet: Rethinking model scaling for convolutional neural networks. In: International Conference on Machine Learning. 2019:6105-6114.
34. [34] Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2015:1-9.
35. [35] Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z. Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016:2818-2826.
36. [36] Sandler M, Howard A, Zhu M, Zhmoginov A, Chen LC. Mobilenetv2: Inverted residuals and linear bottlenecks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018:4510-4520.
37. [37] Zoph B, Vasudevan V, Shlens J, Le Q V. Learning transferable architectures for scalable image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018:8697-8710.
38. [38] Targ S, Almeida D, Lyman K. Resnet in resnet: Generalizing residual architectures. arXiv preprint arXiv:160308029. Published online 2016.
39. [39] He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016:770-778.
40. [40] Zhang X, Zhou X, Lin M, Sun J. Shufflenet: An extremely efficient convolutional neural network for mobile devices. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018:6848-6856.
41. [41] Iandola FN. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size. arXiv preprint arXiv:160207360. Published online 2016.
42. [42] Simonyan K. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:14091556. Published online 2014.
43. [43] Chollet F. Xception: Deep learning with depthwise separable convolutions. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017:1251-1258.
44. [44] Aman Jaiswal. Agriculture crop images. 2020. https://www.kaggle.com/datasets/aman2000jaiswal/agriculture-crop-images