Region-aggregated attention CNN for disease detection in fruit images.

<h4>Background</h4>Diseases and pests have a profound effect on a yearly harvest and productivity in agriculture. A precise and accurate detection of the diseases and pests could facilitate timely treatment and management of the diseases and pests and lessen the resultant loss in economy...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Chang Hee Han, Eal Kim, Tan Nhu Nhat Doan, Dongil Han, Seong Joon Yoo, Jin Tae Kwak
Formato: article
Lenguaje:EN
Publicado: Public Library of Science (PLoS) 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/9224d388654740718fb25ea241f04112
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:9224d388654740718fb25ea241f04112
record_format dspace
spelling oai:doaj.org-article:9224d388654740718fb25ea241f041122021-12-02T20:13:33ZRegion-aggregated attention CNN for disease detection in fruit images.1932-620310.1371/journal.pone.0258880https://doaj.org/article/9224d388654740718fb25ea241f041122021-01-01T00:00:00Zhttps://doi.org/10.1371/journal.pone.0258880https://doaj.org/toc/1932-6203<h4>Background</h4>Diseases and pests have a profound effect on a yearly harvest and productivity in agriculture. A precise and accurate detection of the diseases and pests could facilitate timely treatment and management of the diseases and pests and lessen the resultant loss in economy and health. Herein, we propose an improved design of the disease detection system for plant images.<h4>Methods</h4>Built upon the two-stage framework of object detection neural networks such as Mask R-CNN, the proposed network involves three types of extensions, including the addition of additional level of feature pyramids to improve the exploration and proposal of candidate regions, the aggregation of feature maps from all levels of feature pyramids per candidate region to fully exploit the information from feature pyramids, and the introduction of a squeeze-and-excitation block to the construction of feature pyramids and the aggregated feature maps to improve the representation of feature maps.<h4>Results</h4>The proposed network was evaluated using 74 images of infected apple fruits. In 3-fold cross-validation, the proposed network achieved averaged precision (AP) of 72.26, AP at 0.5 threshold of 88.51 and AP at 0.75 threshold of 82.30. In the comparative experiments, the proposed network outperformed the other competing networks. The utility of the three extensions was also demonstrated in comparison to Mask R-CNN.<h4>Conclusions</h4>The experimental results suggest that the proposed network could identify and localize the symptom of the disease with high accuracy, leading to an early diagnosis and treatment of the disease, and thus holding the potential for improving crop yield and quality.Chang Hee HanEal KimTan Nhu Nhat DoanDongil HanSeong Joon YooJin Tae KwakPublic Library of Science (PLoS)articleMedicineRScienceQENPLoS ONE, Vol 16, Iss 10, p e0258880 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Chang Hee Han
Eal Kim
Tan Nhu Nhat Doan
Dongil Han
Seong Joon Yoo
Jin Tae Kwak
Region-aggregated attention CNN for disease detection in fruit images.
description <h4>Background</h4>Diseases and pests have a profound effect on a yearly harvest and productivity in agriculture. A precise and accurate detection of the diseases and pests could facilitate timely treatment and management of the diseases and pests and lessen the resultant loss in economy and health. Herein, we propose an improved design of the disease detection system for plant images.<h4>Methods</h4>Built upon the two-stage framework of object detection neural networks such as Mask R-CNN, the proposed network involves three types of extensions, including the addition of additional level of feature pyramids to improve the exploration and proposal of candidate regions, the aggregation of feature maps from all levels of feature pyramids per candidate region to fully exploit the information from feature pyramids, and the introduction of a squeeze-and-excitation block to the construction of feature pyramids and the aggregated feature maps to improve the representation of feature maps.<h4>Results</h4>The proposed network was evaluated using 74 images of infected apple fruits. In 3-fold cross-validation, the proposed network achieved averaged precision (AP) of 72.26, AP at 0.5 threshold of 88.51 and AP at 0.75 threshold of 82.30. In the comparative experiments, the proposed network outperformed the other competing networks. The utility of the three extensions was also demonstrated in comparison to Mask R-CNN.<h4>Conclusions</h4>The experimental results suggest that the proposed network could identify and localize the symptom of the disease with high accuracy, leading to an early diagnosis and treatment of the disease, and thus holding the potential for improving crop yield and quality.
format article
author Chang Hee Han
Eal Kim
Tan Nhu Nhat Doan
Dongil Han
Seong Joon Yoo
Jin Tae Kwak
author_facet Chang Hee Han
Eal Kim
Tan Nhu Nhat Doan
Dongil Han
Seong Joon Yoo
Jin Tae Kwak
author_sort Chang Hee Han
title Region-aggregated attention CNN for disease detection in fruit images.
title_short Region-aggregated attention CNN for disease detection in fruit images.
title_full Region-aggregated attention CNN for disease detection in fruit images.
title_fullStr Region-aggregated attention CNN for disease detection in fruit images.
title_full_unstemmed Region-aggregated attention CNN for disease detection in fruit images.
title_sort region-aggregated attention cnn for disease detection in fruit images.
publisher Public Library of Science (PLoS)
publishDate 2021
url https://doaj.org/article/9224d388654740718fb25ea241f04112
work_keys_str_mv AT changheehan regionaggregatedattentioncnnfordiseasedetectioninfruitimages
AT ealkim regionaggregatedattentioncnnfordiseasedetectioninfruitimages
AT tannhunhatdoan regionaggregatedattentioncnnfordiseasedetectioninfruitimages
AT dongilhan regionaggregatedattentioncnnfordiseasedetectioninfruitimages
AT seongjoonyoo regionaggregatedattentioncnnfordiseasedetectioninfruitimages
AT jintaekwak regionaggregatedattentioncnnfordiseasedetectioninfruitimages
_version_ 1718374779019329536