GÖRÜNTÜ ÖN İŞLEME TEKNİKLERİ VE DERİN ÖĞRENME İLE BİTKİ ZARARLILARININ SINIFLANDIRILMASI

Year 2024, Volume: 12 Issue: 2, 455 - 465, 30.06.2024

Şevval Ezgi Eze Selcan Kaplan Berkaya

https://doi.org/10.21923/jesd.1490176

Abstract

Bitki zararlılarının erken dönemde, etkili bir şekilde tespit edilip kontrol altına alınmalarını sağlamak bitkilerin korunmasına, ürün veriminin artırılmasına ve tarım ekonomisindeki kayıpların azaltılmasına yardımcı olmaktadır. Bu çalışmada, bitki zararlılarının sınıflandırılması için derin öğrenme tabanlı yöntemler önerilmiştir. Aynı zamanda çeşitli görüntü ön işleme tekniklerinin performansa etkisi araştırılmıştır. Önerilen modeller, önceden eğitilmiş beş farklı derin sinir ağı (GoogLeNet, ResNet-18, ResNet-101, VGG-16 ve VGG-19) ile transfer öğrenimi ve bu ağlardan çıkarılan öznitelikler ile Destek Vektör Makinesi sınıflandırıcısını kullanmaktadır. Ayrıca yeşil renk kanalı çıkarımı, veri artırımı, histogram eşitleme, derin öğrenme tabanlı segmentasyon ile arka plan eliminasyonu gibi farklı görüntü ön işleme teknikleri ayrı ayrı ve birlikte kullanılarak kapsamlı bir performans analizi yapılmıştır. Deneyler, sırasıyla 10 ve 40 bitki zararlısı türü içeren Li ve D0 veri setleri üzerinde gerçekleştirilmiştir. Deneyler sonucunda iki veri setinde de veri artırımı ve ResNet-101 ağı ile transfer öğrenimi yöntemi kullanılarak sırasıyla %96.36 ve %99.63 doğruluk ile en yüksek performanslar elde edilmiştir. Deneysel sonuçlar, önerilen modellerin bitki zararlısı kontrolünde etkin bir şekilde kullanılabileceğini göstermektedir.

Keywords

Derin Öğrenme, Transfer Öğrenimi, Görüntü Ön İşleme, Bitki Zararlısı Sınıflandırma, Akıllı Tarım

CLASSIFICATION OF INSECT PESTS WITH DEEP LEARNING AND IMAGE PREPROCESSING TECHNIQUES

Abstract

Early and effective insect pest detection and control help to protect plants, increase crop yields, and reduce losses in the agricultural economy. In this paper, deep learning-based methods are proposed for classifying insect pests. Additionally, the impact of various image preprocessing techniques on performance has been investigated. The proposed models utilize transfer learning with five different pre-trained deep neural networks (GoogLeNet, ResNet-18, ResNet-101, VGG-16, and VGG-19), and Support Vector Machine classifier with features extracted from these networks. Furthermore, a comprehensive performance analysis is conducted using various image preprocessing techniques such as green color channel extraction, data augmentation, histogram equalization, and background elimination with deep learning based segmentation, individually and in combination. The experiments were conducted on the Li and D0 datasets containing 10 and 40 insect pest species, respectively. As a result of the experiments, the highest performances were obtained on both datasets, with accuracy rates of 96.36% and 99.63%, respectively, using data augmentation and transfer learning with the ResNet-101 network. The experimental results indicate that the proposed models can be effectively used in insect pest control.

Keywords

Deep Learning, Transfer Learning, Image Processing, Insect Pest Classification, Smart Agriculture.

References

• Chen, W., Gao, H., Ding, D., Dong, X., Luo, X., 2023. Chili Pepper Pests Recognition Based on Hsv Color Space and Convolutional Neural Networks. In 2023 IEEE 3rd International Conference on Electronic Technology, Communication and Information (ICETCI), pp. 241-245. Deng, L., Wang, Y., Han, Z., & Yu, R., 2018. Research on Insect Pest Image Detection and Recognition Based on Bio-Inspired Methods. Biosystems Engineering, 169, 139-148. He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep Residual Learning for Image Recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770-778. Li, Y., Wang, H., Dang, L. M., Sadeghi-Niaraki, A., Moon, H., 2020. Crop Pest Recognition in Natural Scenes Using Convolutional Neural Networks. Computers and Electronics in Agriculture, 169.
• Maharana, K., Mondal, S., Nemade, B., 2022. A review: Data Pre-Processing and Data Augmentation Techniques. Global Transitions Proceedings, 3(1), 91-99. Nanni, L., Maguolo, G., Pancino, F., 2020. Insect Pest Image Detection and Recognition Based on Bio-Inspired Methods. Ecological Informatics, 57, 101089. Qin, X., Zhang, Z., Huang, C., Dehghan, M., Zaiane, O. R., & Jagersand, M., 2020. U2-Net: Going Deeper with Nested U-structure for Salient Object Detection. Pattern Recognition, 106, 107404. Shorten, C., & Khoshgoftaar, T. M., 2019. A Survey on Image Data Augmentation for Deep Learning. Journal of Big Data, 6(1), 1-48. Simonyan, K., Zisserman, A., 2014. Very Deep Convolutional Networks for Large-Scale Image Recognition, arXiv preprint arXiv:1409.1556.
• Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Rabinovich, A., 2015. Going Deeper with Convolutions. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1-9.
• Thenmozhi, K., Reddy, U. S., 2019. Crop Pest Classification Based on Deep Convolutional Neural Network and Transfer Learning. Computers and Electronics in Agriculture, 164, 104906.
• Toscano-Miranda, R., Aguilar, J., Hoyos, W., Caro, M., Trebilcok, A., & Toro, M., 2024. Different Transfer Learning Approaches for Insect Pest Classification in Cotton. Applied Soft Computing, 153, 111283.
• Wang, C., Zhang, J., He, J., Luo, W., Yuan, X., Gu, L., 2023. A Two-Stream Network with Complementary Feature Fusion for Pest Image Classification. Engineering Applications of Artificial Intelligence, 124, 106563. Wu, X., Zhan, C., Lai, Y. K., Cheng, M. M., & Yang, J., 2019. Ip102: A Large-Scale Benchmark Dataset for Insect Pest Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 8787-8796). Xia, D., Chen, P., Wang, B., Zhang, J., Xie, C., 2018. Insect Detection and Classification Based on an Improved Convolutional Neural Network. Sensors, 18(12).
• Xiao, B., Ma, J. F., Cui, J. T., 2012. Combined Blur, Translation, Scale and Rotation Invariant Image Recognition by Radon and Pseudo-Fourier–Mellin Transforms. Pattern Recognition, 45(1), 314-321.
• Xie, C., Zhang, J., Li, R., Li, J., Hong, P., Xia, J., Chen, P., 2015. Automatic Classification for Field Crop Insects via Multiple-Task Sparse Representation and Multiple-Kernel Learning. Computers and Electronics in Agriculture, 119, 123–132.
• Xie, C., Wang, R., Zhang, J., Chen, P., Dong, W., Li, R., Chen, H., 2018. Multi-Level Learning Features for Automatic Classification of Field Crop Pests. Computers and Electronics in Agriculture, 152, 233-241.
• Yang, X., Luo, Y., Li, M., Yang, Z., Sun, C., Li, W., 2021. Recognizing Pests in Field-Based Images by Combining Spatial and Channel Attention Mechanism. IEEE Access, 9, 162448-162458.
• Yang, Z., Li, W., Li, M., Yang, X., 2021. Automatic Greenhouse Pest Recognition Based on Multiple Color Space Features. International Journal of Agricultural and Biological Engineering, 14(2), 188–195.