Automation of Coffee Biotic Stresses Classification Using Convolutional Neural Networks and Image Preprocessing Techniques

Daniel Imai Yamakawa; Ana Claudia Patrocinio; Pedro Moises de Sousa; Pedro Henrique Aparecido Damaso de Melo

doi:10.22456/2175-2745.150942

Authors

Daniel Imai Yamakawa Universidade Federal de Viçosa (UFV) https://orcid.org/0009-0009-2930-0350
Ana Claudia Patrocinio Universidade Federal de Uberlândia (UFU)
Pedro Moises de Sousa Universidade Federal de Viçosa (UFV) https://orcid.org/0000-0003-4563-0033
Pedro Henrique Aparecido Damaso de Melo Universidade Federal de Viçosa (UFV) https://orcid.org/0000-0002-8054-5523

DOI:

https://doi.org/10.22456/2175-2745.150942

Keywords:

coffee, neural networks, artificial intelligence, wavelet

Abstract

Coffee is one of the most consumed beverages worldwide, with Brazil leading exports in 2024. However, production is threatened by biotic stresses such as cercospora, phoma, rust, and leaf miner, which reduce grain quality and cause major losses. Disease identification is usually performed manually by specialists, a process that is slow, limited, and error-prone. This study proposes automating classification through Convolutional Neural Networks (CNNs) combined with various image processing techniques. The JMunBEN and JMunBEN2 datasets were used for this task, together with image preprocessing techniques such as cropping, augmentation, Gaussian, Clahe, Graythresh and Wavelet. Three CNN architectures—ResNet50, MobileNetV2, and AlexNet—were tested for multiclass classification, trained for 100 epochs with Adam optimizer and CrossEntropy loss. The article proposed wavelet filter yielded the best results, with ResNet achieving 99.07% accuracy, confirming CNNs with image pre-processing techniques to be efficient tools for rapid, reliable detection of coffee biotic stresses.

Downloads

Download data is not yet available.

References

[1] CECAFE. Conselho dos Exportadores de Café do Brasil. 2024. Acesso em: 30 de agosto de 2025. Disponível em: ⟨http://www.consorciopesquisacafe.com.br/images/stories/noticias/2021/2024/Dezembro/CECAFE_Relatorio_Mensal_DEZEMBRO_2024.pdf⟩.

[2] RODRIGUES, F. et al. Desafios do manejo no controle de doenças do café - fitossanidade. In: . [S.l.: s.n.], 2019. p. 52 – 54. ISBN 18066403.

[3] SHARMA, R. Artificial intelligence in agriculture: A review. In: 2021 5th International Conference on Intelligent Computing and Control Systems (ICICCS). [S.l.: s.n.], 2021. p. 937–942.

[4] MISHRA, S.; SACHAN, R.; RAJPAL, D. Deep convolutional neural network based detection system for real-time corn plant disease recognition. Procedia Computer Science, v. 167, p. 2003–2010, 2020. ISSN 1877-0509. International Conference on Computational Intelligence and Data Science. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S187705092030702X⟩.

[5] BEVERS, N.; SIKORA, E. J.; HARDY, N. B. Soybean disease identification using original field images and transfer learning with convolutional neural networks. Computers and Electronics in Agriculture, v. 203, p. 107449, 2022. ISSN 0168-1699. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S0168169922007578⟩.

[6] TEDESCO-OLIVEIRA, D. et al. Convolutional neural networks in predicting cotton yield from images of commercial fields. Computers and Electronics in Agriculture, v. 171, p. 105307, 2020. ISSN 0168-1699. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S0168169919319878⟩.

[7] KRANZ, J. Measuring plant disease. In: . Experimental Techniques in Plant Disease Epidemiology. Berlin, Heidelberg: Springer Berlin Heidelberg, 1988. p. 35–50. ISBN 978-3-642-95534-1. Disponível em: ⟨https://doi.org/10.1007/978-3-642-95534-1_4⟩.

[8] ESGARIO, J. G.; KROHLING, R. A.; VENTURA, J. A. Deep learning for classification and severity estimation of coffee leaf biotic stress. Computers and Electronics in Agriculture, v. 169, p. 105162, 2020. ISSN 0168-1699. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S0168169919313225⟩.

[9] ESGARIO, J. G. et al. An app to assist farmers in the identification of diseases and pests of coffee leaves using deep learning. Information Processing in Agriculture, v. 9, n. 1, p. 38–47, 2022. ISSN 2214-3173. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S2214317321000044⟩.

[10] WANI, M. A.; BHAT, F. A.; AFZAL, S.; KHAN, A. I. Advances in Deep Learning. Singapore: Springer Singapore, 2019.

[11] KRIZHEVSKY, A.; SUTSKEVER, I.; HINTON, G. E. Imagenet classification with deep convolutional neural networks. In: PEREIRA, F. et al. (Ed.). Advances in Neural Information Processing Systems. Curran Associates, Inc., 2012. v. 25. Disponível em: ⟨https://proceedings.neurips.cc/paper_files/paper/2012/file/c399862d3b9d6b76c8436e924a68c45b-Paper.pdf⟩.

[12] SIMONYAN, K.; ZISSERMAN, A. Very Deep Convolutional Networks for Large-Scale Image Recognition. 2015.

[13] SZEGEDY, C. et al. Going deeper with convolutions. In: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). [S.l.: s.n.], 2015. p. 1–9.

[14] HE, K. et al. Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). [S.l.: s.n.], 2016. p. 770–778.

[15] ZHANG, D. Wavelet transform. In: . Fundamentals of Image Data Mining: Analysis, Features, Classification and Retrieval. Cham: Springer International Publishing, 2019. p. 35–44. ISBN 978-3-030-17989-2. Disponível em: ⟨https://doi.org/10.1007/978-3-030-17989-2_3⟩.

[16] RONNEBERGER, O.; FISCHER, P.; BROX, T. U-net: Convolutional networks for biomedical image segmentation. In: NAVAB, N. et al. (Ed.). Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. Cham: Springer International Publishing, 2015. p. 234–241. ISBN 978-3-319-24574-4.

[17] ZHAO, H. et al. Pyramid Scene Parsing Network. 2017.

[18] BOA SORTE, L. X. et al. Coffee leaf disease recognition based on deep learning and texture attributes. Procedia Computer Science, v. 159, p. 135–144, 2019. ISSN 1877-0509. Knowledge-Based and Intelligent Information Engineering Systems: Proceedings of the 23rd International Conference KES2019. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S1877050919313468⟩.

[19] SANDLER, M. et al. MobileNetV2: Inverted Residuals and Linear Bottlenecks. 2019.

[20] JEPKOECH, J. et al. Arabica coffee leaf images dataset for coffee leaf disease detection and classification. Data in Brief, v. 36, p. 107142, 2021. ISSN 2352-3409. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/S2352340921004261⟩.

[21] D’HAEYER, J. P. Gaussian filtering of images: A regularization approach. Signal Processing, v. 18, n. 2, p. 169–181, 1989. ISSN 0165-1684. Disponível em: ⟨https://www.sciencedirect.com/science/article/pii/0165168489900480⟩.

[22] PIZER, S. et al. Contrast-limited adaptive histogram equalization: speed and effectiveness. In: [1990] Proceedings of the First Conference on Visualization in Biomedical Computing. [S.l.: s.n.], 1990. p. 337–345.

[23] OTSU, N. A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man, and Cybernetics, v. 9, n. 1, p. 62–66, 1979.

[24] PONTI, M. A.; COSTA, G. B. P. da. Como funciona o Deep Learning. 2018. Disponível em: ⟨https://arxiv.org/abs/1806.07908⟩.