Convolutional Neural Networks for Detection of European Apple Canker
Abstract
European Apple Canker affects orchards from several countries and emerged in Brazil in Rio Grande do Sul spreading to other states that represent a large portion of the production of apple. The disease is a threat due to its rapid spread and for the difficulty to detect it. In order to solve this problem, a CNN was proposed for disease classification. The result in validation set for F-Beta was 0.795, obtained by VGG19, with 0.878 in precision and 0.591 in recall. The results show the possibility of using CNN for european canker detection, although they still can be improved.
References
Branco Neto, W. C., Araújo, L., Pinto, F., Machado, R., Ribeiro, Y., Junior, W. C., e Mattos, K. (2021). Cancontrol: plataforma para diagnóstico do cancro europeu da macieira. In Anais do XIII Congresso Brasileiro de Agroinformática, pages 44–52, Porto Alegre, RS, Brasil. SBC.
Goodfellow, I., Bengio, Y., e Courville, A. (2016). Deep Learning. MIT Press. Hughes, D. P. e Salathe, M. (2016). An open access repository of images on plant health to enable the development of mobile disease diagnostics.
Lecun, Y., Bottou, L., Bengio, Y., e Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324.
Liang, Q., Xiang, S., Hu, Y., Coppola, G., Zhang, D., e Sun, W. (2019). Pd2senet: Computer-assisted plant disease diagnosis and severity estimation network. Computers and Electronics in Agriculture, 157:518–529.
Sagar, A. e Jacob, D. (2021). On using transfer learning for plant disease detection. bioRxiv.
TorchVision (2016). TorchVision: PyTorch’s Computer Vision library.
Weber, R. (2014). Biology and control of the apple canker fungus neonectria ditissima (syn. n. galligena) from a northwestern european perspective. Erwerbs-Obstbau, 56:95–107.
