Deep learning for image-based cassava disease detection

Cassava is the third largest source of carbohydrates for human food in the world but is vulnerable to virus diseases, which threaten to destabilize food security in sub-Saharan Africa. Novel methods of cassava disease detection are needed to support improved control which will prevent this crisis. I...

Descripción completa

Detalles Bibliográficos
Autores principales: Ramcharan, A., Baranowski, K., McCloskey, P., Ahmed, B., Legg, James P., Hughes, D.P.
Formato: Journal Article
Lenguaje:Inglés
Publicado: Frontiers Media 2017
Materias:
Acceso en línea:https://hdl.handle.net/10568/89938
_version_ 1855529031835844608
author Ramcharan, A.
Baranowski, K.
McCloskey, P.
Ahmed, B.
Legg, James P.
Hughes, D.P.
author_browse Ahmed, B.
Baranowski, K.
Hughes, D.P.
Legg, James P.
McCloskey, P.
Ramcharan, A.
author_facet Ramcharan, A.
Baranowski, K.
McCloskey, P.
Ahmed, B.
Legg, James P.
Hughes, D.P.
author_sort Ramcharan, A.
collection Repository of Agricultural Research Outputs (CGSpace)
description Cassava is the third largest source of carbohydrates for human food in the world but is vulnerable to virus diseases, which threaten to destabilize food security in sub-Saharan Africa. Novel methods of cassava disease detection are needed to support improved control which will prevent this crisis. Image recognition offers both a cost effective and scalable technology for disease detection. New deep learning models offer an avenue for this technology to be easily deployed on mobile devices. Using a dataset of cassava disease images taken in the field in Tanzania, we applied transfer learning to train a deep convolutional neural network to identify three diseases and two types of pest damage (or lack thereof). The best trained model accuracies were 98% for brown leaf spot (BLS), 96% for red mite damage (RMD), 95% for green mite damage (GMD), 98% for cassava brown streak disease (CBSD), and 96% for cassava mosaic disease (CMD). The best model achieved an overall accuracy of 93% for data not used in the training process. Our results show that the transfer learning approach for image recognition of field images offers a fast, affordable, and easily deployable strategy for digital plant disease detection.
format Journal Article
id CGSpace89938
institution CGIAR Consortium
language Inglés
publishDate 2017
publishDateRange 2017
publishDateSort 2017
publisher Frontiers Media
publisherStr Frontiers Media
record_format dspace
spelling CGSpace899382025-11-11T10:32:49Z Deep learning for image-based cassava disease detection Ramcharan, A. Baranowski, K. McCloskey, P. Ahmed, B. Legg, James P. Hughes, D.P. cassava food security disease control epidemiology deep learning convolutional neural networks transfer learning mobile epidemiology Cassava is the third largest source of carbohydrates for human food in the world but is vulnerable to virus diseases, which threaten to destabilize food security in sub-Saharan Africa. Novel methods of cassava disease detection are needed to support improved control which will prevent this crisis. Image recognition offers both a cost effective and scalable technology for disease detection. New deep learning models offer an avenue for this technology to be easily deployed on mobile devices. Using a dataset of cassava disease images taken in the field in Tanzania, we applied transfer learning to train a deep convolutional neural network to identify three diseases and two types of pest damage (or lack thereof). The best trained model accuracies were 98% for brown leaf spot (BLS), 96% for red mite damage (RMD), 95% for green mite damage (GMD), 98% for cassava brown streak disease (CBSD), and 96% for cassava mosaic disease (CMD). The best model achieved an overall accuracy of 93% for data not used in the training process. Our results show that the transfer learning approach for image recognition of field images offers a fast, affordable, and easily deployable strategy for digital plant disease detection. 2017 2018-01-08T10:37:55Z 2018-01-08T10:37:55Z Journal Article https://hdl.handle.net/10568/89938 en Open Access application/pdf Frontiers Media Ramcharan, A., Baranowski, K., McCloskey, P., Ahamed, B., Legg, J. & Hughes, D.P. (2017). Deep learning for image-based cassava disease detection. Frontiers in Plant Science, 8, 1-7.
spellingShingle cassava
food security
disease control
epidemiology
deep learning
convolutional neural networks
transfer learning
mobile epidemiology
Ramcharan, A.
Baranowski, K.
McCloskey, P.
Ahmed, B.
Legg, James P.
Hughes, D.P.
Deep learning for image-based cassava disease detection
title Deep learning for image-based cassava disease detection
title_full Deep learning for image-based cassava disease detection
title_fullStr Deep learning for image-based cassava disease detection
title_full_unstemmed Deep learning for image-based cassava disease detection
title_short Deep learning for image-based cassava disease detection
title_sort deep learning for image based cassava disease detection
topic cassava
food security
disease control
epidemiology
deep learning
convolutional neural networks
transfer learning
mobile epidemiology
url https://hdl.handle.net/10568/89938
work_keys_str_mv AT ramcharana deeplearningforimagebasedcassavadiseasedetection
AT baranowskik deeplearningforimagebasedcassavadiseasedetection
AT mccloskeyp deeplearningforimagebasedcassavadiseasedetection
AT ahmedb deeplearningforimagebasedcassavadiseasedetection
AT leggjamesp deeplearningforimagebasedcassavadiseasedetection
AT hughesdp deeplearningforimagebasedcassavadiseasedetection