Classification of Green Coffee Beans by Convolutional Neural Network and its Implemnentation on Rasperry and Camera Module
Classification of Green Coffee Beans by Convolutional Neural Network and its Implemnentation on Rasperry and Camera Module
Resumo
The coffee is the most important agricultural product of Timor-Leste for the acquisition of foreign currency. Nevertheless, there are almost no rationalizations at local production sites. The efficient enhancement of the value of coffee beans is desired. The grade of a lot of coffee beans greatly depends on the number of defect beans. However, the defect beans are currently removed by hand pick in Timor-Leste. The final objective of our study is to develop the automatic coffee beans sorting system for the producers of coffee beans in Timor-Leste using state-of-the-art machine learning techniques and cheap single-board computer. As the first step, we developed an image processing system which classifies the images of green coffee beans into each type of defect by deep convolutional neural networks. Next, the trained artificial neural network was implemented into Raspberry Pi compute module with camera module. We evaluated the performances of inspection speed and accuracy of the sorting system for practical realization.
Keywords: Agricultural engineering, Manufacturing automation, Convolutional neural networks, Image processing, Raspberry Pi
Publicação Completa
A publicação completa está disponível em PDF.
Citar esta Publicação
Hidekazu Fukai et al. (2018). Classification of Green Coffee Beans by Convolutional Neural Network and its Implemnentation on Rasperry and Camera Module. Hera.
Hidekazu Fukai et al. “Classification of Green Coffee Beans by Convolutional Neural Network and its Implemnentation on Rasperry and Camera Module.” Hera, 2018.
Hidekazu Fukai et al. 2018. “Classification of Green Coffee Beans by Convolutional Neural Network and its Implemnentation on Rasperry and Camera Module.” Hera.
Detalhes da Publicação
- Categoria
- Artigo de Revista
- Âmbito
- Nacional
- Ano
- 2018
- Local de Publicação
- Hera
- ISBN
- 2617 - 4952 (Online)
Autores
-
Hidekazu Fukai,
-
Junya Furukawa,
-
Carlito Pinto,
-
Carmelita Afonso