Aplicação de aprendizado de máquina profundo para detecção por imagens De doenças em frutos do cacaueiro
International Journal of Development Research
Aplicação de aprendizado de máquina profundo para detecção por imagens De doenças em frutos do cacaueiro
Received 10th February, 2021; Received in revised form 20th March, 2021; Accepted 14th April, 2021; Published online 30th May, 2021
Copyright © 2021, Maria Eliana da Silva Holanda et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The state of Pará is the largest cocoa producer in Brazil, with 51% of national production, involving 26 thousand producers, generating 64 thousand direct and 225 thousand indirect jobs. However, diseases that affect this culture are responsible for high losses in yield. This study presents an approach based on deep learning to identify diseases that affect the cocoa culture. A public database with 4,389 fruit images was used, covering the diseases black pod rot and pod borer. The experiments using the techniques of data augmentation and convolutional neural networks (CNN) indicate an average accuracy of 95% in the images' classification. In this way, the present work aims to contribute effectively proposing a tool that can help in the improvement of the cocoa production chain in the state of Pará.
