日本地球惑星科学連合2022年大会

講演情報

[E] ポスター発表

セッション記号 M (領域外・複数領域) » M-SD 宇宙開発・地球観測

[M-SD41] Geospatial applications for natural resources, environment and agriculture

2022年5月29日(日) 11:00 〜 13:00 オンラインポスターZoom会場 (29) (Ch.29)

コンビーナ:Mohamed Shariff Abdul Rashid Bin(Universiti Putra Malaysia )、コンビーナ:高橋 幸弘(北海道大学・大学院理学院・宇宙理学専攻)、Chairperson:Abdul Rashid Bin Mohamed Shariff(Universiti Putra Malaysia)、Anuar Ahmad(Universiti Teknokogi Malaysia)、成瀬 延康(滋賀医科大学 医学部医学科)

11:00 〜 13:00

[MSD41-P01] スペクトルライブラリを活用したリモートセンシングの戦略

★招待講演

*成瀬 延康1高橋 幸弘2 (1.滋賀医科大学 医学部医学科、2.北海道大学 大学院理学院 宇宙理学専攻)

キーワード:スペクトルライブラリ、マルチスペクトル

In the next generation agriculture, it is important to have remote sensing that can monitor the condition of crops and soil on a farm with high frequency and high accuracy. If we can understand the growth stage of crops, predict crop yields, detect damage from pests and diseases at an early stage, and determine the excess or deficiency of fertilizer and water in the soil without much effort, we can contribute to the automation, large-scale, and stable operation of agriculture, as well as to the problem of aging farmers in Japan. Recently, remote sensing of agriculture using satellites and drones has become a hot topic, but the problem is that farms are not being monitored with sufficient accuracy and frequency. The reason is that most of the current remote sensing is based on ordinary color photography (three broadband bands) or several bands added to it. The images acquired by these methods lose a lot of information. It is not possible to obtain accurate and precise information on crops and soils that can be used for practical purposes.
To solve this problem, multi-wavelength spectral imaging with a narrow band (wavelength) can dramatically improve the accuracy and precision of the images and make a significant contribution to agriculture, as many case studies have proven. Nevertheless, spectral measurement is not widely used because 1) the equipment is too large, heavy, and expensive, and the operation cost is too high; 2) the reflection spectrum depends on the combination of the angle of solar radiation and the angle of the camera, and the relationship between them can be fixed to measure only a narrow area directly under the flying object, resulting in a significant decrease in the frequency of observation. Therefore, the observation frequency is significantly reduced. As a result, the conventional multi-wavelength spectrum measurement is close to manual work, not worth the cost, and not applicable.
In order to overcome the limitations of agricultural remote sensing, it is important to construct a detailed database "spectral library" of the relationship between illumination, measurement direction, and spectra on the ground in a highly efficient manner and to link it with data on crop conditions and yield. This presentation will introduce the efforts of a group at Hokkaido University for precise remote sensing using a spectral library, which they have been working on for three years since 2020. This presentation will demonstrate how we have established a library of 300,000 spectra in total and how we can collect and view the information using an edge server.
Part of the content of this research was supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT)'s Regional Innovation Ecosystem Formation Program under the theme of "Creation of an 'Innovative Remote Sensing Business' Using Hokkaido University's Spectral Measurement Technology - Breakthrough for Next-generation Agriculture through the Construction of a Spectral Library.