Next Article in Journal
Image Collection Summarization Method Based on Semantic Hierarchies
Next Article in Special Issue
Carrot Yield Mapping: A Precision Agriculture Approach Based on Machine Learning
Previous Article in Journal
Cities of the Future? The Potential Impact of Artificial Intelligence
Open AccessArticle

A Study on CNN-Based Detection of Psyllids in Sticky Traps Using Multiple Image Data Sources

1
Embrapa Agricultural Informatics, Campinas 13083-886, Brazil
2
CromAI, São Paulo 05319-090, Brazil
*
Author to whom correspondence should be addressed.
AI 2020, 1(2), 198-208; https://doi.org/10.3390/ai1020013
Received: 3 April 2020 / Revised: 1 May 2020 / Accepted: 13 May 2020 / Published: 18 May 2020
(This article belongs to the Special Issue Artificial Intelligence in Agriculture)
Deep learning architectures like Convolutional Neural Networks (CNNs) are quickly becoming the standard for detecting and counting objects in digital images. However, most of the experiments found in the literature train and test the neural networks using data from a single image source, making it difficult to infer how the trained models would perform under a more diverse context. The objective of this study was to assess the robustness of models trained using data from a varying number of sources. Nine different devices were used to acquire images of yellow sticky traps containing psyllids and a wide variety of other objects, with each model being trained and tested using different data combinations. The results from the experiments were used to draw several conclusions about how the training process should be conducted and how the robustness of the trained models is influenced by data quantity and variety. View Full-Text
Keywords: deep learning; model robustness; data variety; citrus; HLB deep learning; model robustness; data variety; citrus; HLB
Show Figures

Figure 1

MDPI and ACS Style

Barbedo, J.G.A.; Castro, G.B. A Study on CNN-Based Detection of Psyllids in Sticky Traps Using Multiple Image Data Sources. AI 2020, 1, 198-208.

Show more citation formats Show less citations formats

Article Access Map by Country/Region

1
Back to TopTop