Image monitoring system to find hidden objects with time series data and different angle data

2020 Fiscal Year Final Research Report

• Project/Area Number: 17K08036
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Agricultural environmental engineering/Agricultural information engineering
• Research Institution: National Agriculture and Food Research Organization
• Principal Investigator: FUKATSU Tokihiro 国立研究開発法人農業・食品産業技術総合研究機構, 農業技術革新工学研究センター, 上級研究員 (40355483)
• Co-Investigator(Kenkyū-buntansha): 内藤 裕貴 国立研究開発法人農業・食品産業技術総合研究機構, 農業技術革新工学研究センター, 研究員 (20794118)
• Project Period (FY): 2017-04-01 – 2021-03-31
• Keywords: 花果実検出 / 画像解析 / 時系列データ / 移動計測

Outline of Final Research Achievements

It is important for agricultural science to detect objects such as flowers or fruits from image data. However, these objects are sometimes hidden by leaves or stems, so that it may be difficult to detect them from the image data. In this study, we propose a new image monitoring method to find hidden objects using time series image data and different angle image data in order to deal with the issues. We have developed a movable platform with high-resolution image system so that it enables to collect spatio-temporal images automatically. We also constructed a framework to analyze image data with defined metadata that shows relationship to each image data. By using related image data extracted based on metadata, our proposed system enables to find hidden objects efficiently.

Free Research Field

農業センサネットワーク

Academic Significance and Societal Importance of the Research Achievements

本研究は、深層学習をはじめとする画像検出手法に貢献するものであり、特に作物生育といった場面に有効に作用するものと思われる。また本研究で開発された要素技術は、農業ビッグデータの構築や高度な生育モニタリングなどに利用できる基盤技術であり、社会的・学術的にも意義があるものである。