Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture
The direct effect of global warming on viticulture is already apparent, with unexpected pests and diseases as one of the most concerning consequences. Deploying sticky traps on grape plantations to attract key insects has been the backbone of conventional pest management programs. However, they are...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Agronomy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4395/12/12/3052 |
| _version_ | 1797461820678602752 |
|---|---|
| author | João Gonçalves Eduardo Silva Pedro Faria Telmo Nogueira Ana Ferreira Cristina Carlos Luís Rosado |
| author_facet | João Gonçalves Eduardo Silva Pedro Faria Telmo Nogueira Ana Ferreira Cristina Carlos Luís Rosado |
| author_sort | João Gonçalves |
| collection | DOAJ |
| description | The direct effect of global warming on viticulture is already apparent, with unexpected pests and diseases as one of the most concerning consequences. Deploying sticky traps on grape plantations to attract key insects has been the backbone of conventional pest management programs. However, they are time-consuming processes for winegrowers, conducted through visual inspection via the manual identification and counting of key insects. Additionally, winegrowers usually lack taxonomy expertise for accurate species identification. This paper explores the usage of deep learning on the edge to identify and quantify pest counts automatically. Different mobile devices were used to acquire a dataset of yellow sticky and delta traps, consisting of 168 images with 8966 key insects manually annotated by experienced taxonomy specialists. Five different deep learning models suitable to run locally on mobile devices were selected, trained, and benchmarked to detect five different insect species. Model-centric, data-centric, and deployment-centric strategies were explored to improve and fine-tune the considered models, where they were tested on low-end and high-end mobile devices. The SSD ResNet50 model proved to be the most suitable architecture for deployment on edge devices, with accuracies per class ranging from 82% to 99%, the F1 score ranging from 58% to 84%, and inference speeds per trap image of 19.4 s and 62.7 s for high-end and low-end smartphones, respectively. These results demonstrate the potential of the approach proposed to be integrated into a mobile-based solution for vineyard pest monitoring by providing automated detection and the counting of key vector insects to winegrowers and taxonomy specialists. |
| first_indexed | 2024-03-09T17:25:48Z |
| format | Article |
| id | doaj.art-37ac4e6eef7a4051983fb375c8fba1f1 |
| institution | Directory Open Access Journal |
| issn | 2073-4395 |
| language | English |
| last_indexed | 2024-03-09T17:25:48Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agronomy |
| spelling | doaj.art-37ac4e6eef7a4051983fb375c8fba1f12023-11-24T12:45:37ZengMDPI AGAgronomy2073-43952022-12-011212305210.3390/agronomy12123052Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in ViticultureJoão Gonçalves0Eduardo Silva1Pedro Faria2Telmo Nogueira3Ana Ferreira4Cristina Carlos5Luís Rosado6Fraunhofer Portugal AICOS, 4200-135 Porto, PortugalFraunhofer Portugal AICOS, 4200-135 Porto, PortugalFraunhofer Portugal AICOS, 4200-135 Porto, PortugalGeoDouro—Consultoria e Topografia, Lda., 5100-196 Lamego, PortugalAssociação para o Desenvolvimento da Viticultura Duriense, 5000-033 Vila Real, PortugalAssociação para o Desenvolvimento da Viticultura Duriense, 5000-033 Vila Real, PortugalFraunhofer Portugal AICOS, 4200-135 Porto, PortugalThe direct effect of global warming on viticulture is already apparent, with unexpected pests and diseases as one of the most concerning consequences. Deploying sticky traps on grape plantations to attract key insects has been the backbone of conventional pest management programs. However, they are time-consuming processes for winegrowers, conducted through visual inspection via the manual identification and counting of key insects. Additionally, winegrowers usually lack taxonomy expertise for accurate species identification. This paper explores the usage of deep learning on the edge to identify and quantify pest counts automatically. Different mobile devices were used to acquire a dataset of yellow sticky and delta traps, consisting of 168 images with 8966 key insects manually annotated by experienced taxonomy specialists. Five different deep learning models suitable to run locally on mobile devices were selected, trained, and benchmarked to detect five different insect species. Model-centric, data-centric, and deployment-centric strategies were explored to improve and fine-tune the considered models, where they were tested on low-end and high-end mobile devices. The SSD ResNet50 model proved to be the most suitable architecture for deployment on edge devices, with accuracies per class ranging from 82% to 99%, the F1 score ranging from 58% to 84%, and inference speeds per trap image of 19.4 s and 62.7 s for high-end and low-end smartphones, respectively. These results demonstrate the potential of the approach proposed to be integrated into a mobile-based solution for vineyard pest monitoring by providing automated detection and the counting of key vector insects to winegrowers and taxonomy specialists.https://www.mdpi.com/2073-4395/12/12/3052viticulturepests monitoringinsect detectionobject detectiondeep learningmachine-learning |
| spellingShingle | João Gonçalves Eduardo Silva Pedro Faria Telmo Nogueira Ana Ferreira Cristina Carlos Luís Rosado Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture Agronomy viticulture pests monitoring insect detection object detection deep learning machine-learning |
| title | Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture |
| title_full | Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture |
| title_fullStr | Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture |
| title_full_unstemmed | Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture |
| title_short | Edge-Compatible Deep Learning Models for Detection of Pest Outbreaks in Viticulture |
| title_sort | edge compatible deep learning models for detection of pest outbreaks in viticulture |
| topic | viticulture pests monitoring insect detection object detection deep learning machine-learning |
| url | https://www.mdpi.com/2073-4395/12/12/3052 |
| work_keys_str_mv | AT joaogoncalves edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT eduardosilva edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT pedrofaria edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT telmonogueira edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT anaferreira edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT cristinacarlos edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture AT luisrosado edgecompatibledeeplearningmodelsfordetectionofpestoutbreaksinviticulture |