2022 Fiscal Year Final Research Report
Development of an AI topographic classification model by the use of Landform Evolution Models with geohistorical transition AI training data
| Project/Area Number |
20K20726
|
|---|
| Research Category |
Grant-in-Aid for Challenging Research (Exploratory)
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Medium-sized Section 4:Geography, cultural anthropology, folklore, and related fields
|
|---|
| Research Institution | Okayama University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2020-07-30 – 2023-03-31
|
| Keywords | AIによる地形分類 / 地形変化シミュレータ / AIの地形区分評価関数 / AIによる地形面区分図 |
|---|
| Outline of Final Research Achievements |
The traditional work of terrain classification and creation of maps based on the experience and judgment of experts is carried out using high-resolution digital elevation models and artificial intelligence (AI) machine learning in this study. One of the original aspects of this work is the use of computer simulations of landform evolution to provide the large amount of training data needed for AI. To solve the initial topography problem of the simulation, I selected the Ito pyroclastic flow distribution area in the Satsuma Peninsula, Kagoshima Prefecture, and the marine terrace topography of Kikai Island, Kagoshima Prefecture. To confirm the generality of the AI in this study, a topographic classification map of the Muroto Peninsula in Kochi Prefecture, is created. As a result, the AI could provide a topographic classification map of the marine terrace topography that is consistent with the topographic classification created by conventional experts.
|
| Free Research Field |
変動地形学
|
| Academic Significance and Societal Importance of the Research Achievements |
近年のAI技術の実装環境の普及を受けて，画像などデジタルデータに記録されている形状の判別や分類の自動化の研究の進展は著しいが，地表面形状の分類は，画像から読み取れる形状の相違だけでは成しえない．しかし，人が地形の形成プロセスを単元ごとに教科書や野外実習で学ぶのと同様に，AIが河川の侵食・堆積のプロセス，斜面の従順化や地すべりなどのプロセス，また，海水準変動や地殻変動などの広域変化プロセスと逐次自動で地形分類を更新する履歴情報までを学習できれば，そこで得られるAIの地形分類の評価関数は，従来の専門家判断による地形分類と比較が可能なレベルに高度化できる可能性があると考え，本研究を実施した．
|
