• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Reassessment of the role of wild grass and weed species in the agro-ecosystem services for adaptive crop and natural resource management

Research Project

Project/Area Number 18KT0087
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section特設分野
Research Field Agricultural Resources for the Next Generation
Research InstitutionThe University of Tokyo

Principal Investigator

KATO YOICHIRO  東京大学, 大学院農学生命科学研究科(農学部), 教授 (50463881)

Co-Investigator(Kenkyū-buntansha) 深野 祐也  東京大学, 大学院農学生命科学研究科(農学部), 助教 (70713535)
郭 威  東京大学, 大学院農学生命科学研究科(農学部), 助教 (70745455)
小山 明日香  国立研究開発法人森林研究・整備機構, 森林総合研究所, 主任研究員 等 (90812462)
Project Period (FY) 2018-07-18 – 2022-03-31
Project Status Completed (Fiscal Year 2021)
Budget Amount *help
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2020: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2019: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2018: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Keywords雑草 / 野外植物 / 作物栽培 / 農資源管理学
Outline of Final Research Achievements

This study aimed to develop an automatic plant species identification system for images of seedlings or vegetative organs. The development processes consisted of preparing a database of training data and developing a deep learning algorithm for automatic plant species identification. Using the prepared image database, a deep learning model was constructed, and a species identification rate of over 95% was obtained. Since it was anticipated that further improvement of the identification rate would require a dramatic increase in the amount of training data, we tried to develop the plant image synthesis using Generative Adversarial Network, and a practical algorithm with an average identification accuracy of 91% was successfully prepared.

Academic Significance and Societal Importance of the Research Achievements

作物生産と生態系保全が高度に両立した農業システムを確立するには、耕地・草地における生物多様性の高精度モニタリング技術の確立が不可欠である。植物種の自動認識技術は、環境に優しい局所雑草管理(必要な箇所にスポット状に適切な除草剤を散布)の確立にも貢献する。本研究は、農耕地の主要な雑草種を極めて高い精度で自動識別する技術を開発した。それだけでなく、現実世界の実生画像とほぼ変わりない精度の合成画像データを深層学習によって人為的に作出することに成功した。大量のトレーニングデータ取得の困難さゆえに深層学習技術を適用しづらかった野外植物科学の現状の突破につながる画期的な発見である。

Report

(5 results)
  • 2021 Annual Research Report   Final Research Report ( PDF )
  • 2020 Research-status Report
  • 2019 Research-status Report
  • 2018 Research-status Report
  • Research Products

    (3 results)

All 2021 2019

All Presentation (3 results) (of which Int'l Joint Research: 1 results)

  • [Presentation] 絶滅の負債を抱えた草原性植物の過去10年間の変化2021

    • Author(s)
      小柳知代、小山明日香
    • Organizer
      日本生態学会
    • Related Report
      2020 Research-status Report
  • [Presentation] Non-native plants act as a seasonal pollen source for native honeybees in suburban ecosystems in Japan2019

    • Author(s)
      Koyama A, Egawa C, Taki H, Yasuda M, Kanzaki N, Ide T, Okabe K
    • Organizer
      International Association for Vegetation Science (IAVS)
    • Related Report
      2019 Research-status Report
    • Int'l Joint Research
  • [Presentation] 住宅団地内の空地は草原生植物のレフュージアになるか?2019

    • Author(s)
      小倉梨央奈, 小山明日香, 大黒俊哉
    • Organizer
      第66回日本生態学会大会
    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-07-20   Modified: 2023-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi