| 電子情報通信学会 研究会発表申込システム 講演論文 詳細 |
技報閲覧サービス [ログイン] 技報アーカイブ |
| トップに戻る | 前のページに戻る | [Japanese] / [English] |
| 講演抄録／キーワード | ||
| 講演名 | 2023-12-08 08:50 Towards Satellite Data Fusion For Active Fire Detection ○Nur Fajar Trihantoro・Simon Jones・Karin Reinke（RMIT Univ.）  SANE2023-65 |
|
| 抄録 | （和） | Numerous Earth observation satellite instruments show potential for monitoring wildfire, yet current fire detection algorithms mostly rely on data from a single sensor. This highlights an opportunity to explore a multi-sensor approach, considering the abundance of available Earth observation data. This study introduces a novel data-fusion-based algorithm utilizing Earth observation satellite data. It aims to create a reliable fire detection system, merging middle infrared data from geostationary sensors like Himawari-8/9 AHI, GeoKompsat-2 AMI, and low earth orbit sensors like Sentinel-3 SLSTR. The Kalman filter is the data fusion method, which enhances data accuracy by filtering out noise. Following Roberts and Wooster's [1] approach, a Diurnal Temporal Cycle model provides background temperature information. Preliminary results show promise, with detection rates exceeding 95% for various scenarios, highlighting the algorithm's potential. The presented approach showcases a promising advancement in active fire detection, combining the benefits of data fusion and conventional methodology for improved accuracy and adaptability in disaster management scenarios. The method's sensor agnosticism, iterative detection process, and consideration of resolution discrepancies mark significant advancements, addressing limitations observed in existing fire detection methodologies. Future steps involve refining the algorithm's performance across varied study cases, addressing spatial and temporal resolution disparities among input sensors, and scaling the method for near real-time implementation to enhance its utility in disaster response further. |
| （英） | Numerous Earth observation satellite instruments show potential for monitoring wildfire, yet current fire detection algorithms mostly rely on data from a single sensor. This highlights an opportunity to explore a multi-sensor approach, considering the abundance of available Earth observation data. This study introduces a novel data-fusion-based algorithm utilizing Earth observation satellite data. It aims to create a reliable fire detection system, merging middle infrared data from geostationary sensors like Himawari-8/9 AHI, GeoKompsat-2 AMI, and low earth orbit sensors like Sentinel-3 SLSTR. The Kalman filter is the data fusion method, which enhances data accuracy by filtering out noise. Following Roberts and Wooster's [1] approach, a Diurnal Temporal Cycle model provides background temperature information. Preliminary results show promise, with detection rates exceeding 95% for various scenarios, highlighting the algorithm's potential. The presented approach showcases a promising advancement in active fire detection, combining the benefits of data fusion and conventional methodology for improved accuracy and adaptability in disaster management scenarios. The method's sensor agnosticism, iterative detection process, and consideration of resolution discrepancies mark significant advancements, addressing limitations observed in existing fire detection methodologies. Future steps involve refining the algorithm's performance across varied study cases, addressing spatial and temporal resolution disparities among input sensors, and scaling the method for near real-time implementation to enhance its utility in disaster response further. | |
| キーワード | （和） | Kalman Filter / Data Fusion / Wildfire Detection / Earth Observation / Brightness Temperature / Middle Infrared / Diurnal Temperature Cycle / |
| （英） | Kalman Filter / Data Fusion / Wildfire Detection / Earth Observation / Brightness Temperature / Middle Infrared / Diurnal Temperature Cycle / | |
| 文献情報 | 信学技報, vol. 123, no. 298, SANE2023-65, pp. 35-35, 2023年12月. | |
| 資料番号 | SANE2023-65 | |
| 発行日 | 2023-11-30 (SANE) | |
| ISSN | Online edition: ISSN 2432-6380 | |
| 著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します．(許諾番号：10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) | |
| PDFダウンロード | SANE2023-65 |
|
| 研究会情報 | |
| 研究会 | SANE |
| 開催期間 | 2023-12-07 - 2023-12-09 |
| 開催地（和） | インドネシア ジャワ県ソロ市 |
| 開催地（英） | Surakarta, Indonesia |
| テーマ（和） | ICSANE |
| テーマ（英） | ICSANE |
| 講演論文情報の詳細 | |
| 申込み研究会 | SANE |
| 会議コード | 2023-12-SANE |
| 本文の言語 | 英語 |
| タイトル（和） | |
| サブタイトル（和） | |
| タイトル（英） | Towards Satellite Data Fusion For Active Fire Detection |
| サブタイトル（英） | |
| キーワード(1)（和/英） | Kalman Filter / Kalman Filter |
| キーワード(2)（和/英） | Data Fusion / Data Fusion |
| キーワード(3)（和/英） | Wildfire Detection / Wildfire Detection |
| キーワード(4)（和/英） | Earth Observation / Earth Observation |
| キーワード(5)（和/英） | Brightness Temperature / Brightness Temperature |
| キーワード(6)（和/英） | Middle Infrared / Middle Infrared |
| キーワード(7)（和/英） | Diurnal Temperature Cycle / Diurnal Temperature Cycle |
| キーワード(8)（和/英） | / |
| 第1著者 氏名（和/英/ヨミ） | Nur Fajar Trihantoro / Nur Fajar Trihantoro / |
| 第1著者 所属（和/英） | RMIT University (略称： RMIT大) RMIT University (略称： RMIT Univ.) |
| 第2著者 氏名（和/英/ヨミ） | Simon Jones / Simon Jones / |
| 第2著者 所属（和/英） | RMIT University (略称： RMIT大) RMIT University (略称： RMIT Univ.) |
| 第3著者 氏名（和/英/ヨミ） | Karin Reinke / Karin Reinke / |
| 第3著者 所属（和/英） | RMIT University (略称： RMIT大) RMIT University (略称： RMIT Univ.) |
| 講演者 | 第1著者 |
| 発表日時 | 2023-12-08 08:50:00 |
| 発表時間 | 20分 |
| 申込先研究会 | SANE |
| 資料番号 | SANE2023-65 |
| 巻番号（vol） | vol.123 |
| 号番号（no） | no.298 |
| ページ範囲 | p.35 |
| ページ数 | 1 |
| 発行日 | 2023-11-30 (SANE) |
[研究会発表申込システムのトップページに戻る]