WEKO3
アイテム
水中ロボットを用いたコバルトリッチマンガンクラストの定量的分布推定法
https://doi.org/10.18997/00007976
https://doi.org/10.18997/00007976545fbc7a-c842-44fc-8a25-4f16279ef99b
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
sei_k_369.pdf (60.7 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2020-11-30 | |||||||
| 資源タイプ | ||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_db06 | |||||||
| 資源タイプ | doctoral thesis | |||||||
| タイトル | ||||||||
| タイトル | Quantitative Estimation Method of Deep Sea Cobalt-rich Manganese Crust Distribution using Underwater Robots | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | 水中ロボットを用いたコバルトリッチマンガンクラストの定量的分布推定法 | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 著者 |
Umesh, Neettiyath
× Umesh, Neettiyath
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | A method to efficiently map the distribution of Cobalt-rich Manganese Crusts （Mncrust） using data collected by autonomous underwater vehicles and remotely operated vehicles is developed. Volumetric measurements of Mn-crusts are made using a highfrequency sub-surface sonar and a 3D visual mapping instrument mounted on these vehicles. This thesis proposes a fully automated algorithmic approach to estimate Mncrust distribution by combining the continuous sub-surface thickness measurements with the exposed surface area identified in the 3D maps. This method is applied to data collected from field surveys in the deep sea and the results are validated using physical samples. Manganese crusts （Mn-crust） are a type of mineral deposit commonly found on seamounts and guyots at depths varying from 800???? to 5500????. They are precipitated from ambient seawater creating deposits up to 250???????? in thickness. Mn-crusts are rich in Cobalt, Nickel, other rare minerals and rare earth metals, making it a potential target for deep sea mining. Since they contain the historical record of millions of years of ocean conditions and fossils of ancient crustaceans, it is of high scientific interest. The present methods of studying Mn-crusts by collecting physical samples provides a low spatial resolution of the order of several km and cannot thus capture the local variability of its distribution. Since the deposits are thin compared to the typical resolutions of sub-bottom sonars, in-situ surveys using dedicated sensors are required to accurately measure the Mn-crust thickness. Towed camera surveys and ROV video feeds have been used by researchers to visually confirm the presence of Mn-crusts, but cannot be used to make accurate thickness or volume measurements. Automated methods for classification is suitable for analyzing large volumes of seafloor data to create estimates of Mn-crust coverage. 3D colour seafloor maps are more suitable to distinguish the unique texture and shape of Mn-crust deposits. The proposed method creates volumetric distribution estimates of Mn-crusts using data collected using a high-frequency sub-surface sonar and a 3D visual mapping instrument mounted on an underwater robot suitable for surveying Mn-crusts. The proposed sensor fusion method consists of 3 algorithms for measuring the percentage cover （lateral coverage）, thickness and unit mass coverage （mass of crust per unit area） of Mn-crusts respectively and can be scaled to large regions of seafloor. 3D colour reconstructions made by the visual mapping instrument is analyzed by using a Support Vector Machine classifier to identify Mn-crusts and other seafloor types present and estimate a percentage cover value. In the areas covered with exposed Mncrust, the sub-surface sonar data is analyzed to measure the continuous thickness of Mn-crust. These thickness measurements are then extrapolated into the entire region containing exposed Mn-crust. From the extrapolated thickness map, the total volume of crust is calculated and the mass coverage estimates are calculated by integrating the thickness values over an area of influence. The density of Mn-crusts measured from samples is used in the mass calculations. This method is applied to field data collected from three expeditions at Takuyo Daigo seamount in the northwestern Pacific ocean at depths ranging from 1350???? to 1600????. The total transect lengths add to about 11 ???????? with 12, 510????2 mapped. The results showed that 52% of the surveyed area is covered by Mn-crusts with a mean thickness of 69.6????????. The mean Mn-crust occurrence is 69.6 ????????/????2 with a maximum of 204 ????????/????2 in the mapped region. In order to validate the proposed approach, the results of thickness were compared to Mn-crust samples retrieved from the surveyed area by researchers. The results generated by the method agree with estimates made from samples retrieved from the area, and shows more detailed distribution patterns. By looking at the variability of crust coverage in different transects, it is seen that the coverage of crust can vary by a large margin. Therefore, a continuous in-situ survey is required to accurately assess the distribution of Mn-crust over large regions. The proposed method is therefore effective for efficiently estimating Mn-crust distributions and inventories at hectare scale areas. This is the first method suitable for estimating volumetric distribution estimates of Mn-crust for hectare-scale or larger areas at centimeter resolution. A method for sensor fusion using a 3D map and secondary sensor data and a method for high accuracy classification of seafloor 3D colour point clouds are developed. The information generated can provide valuable insights into the study of Mn-crust distribution, the ocean processes of Mn-crust formation and inputs for making policy decisions regarding deep sea mining and planning future courses of action. | |||||||
| 目次 | ||||||||
| 内容記述タイプ | TableOfContents | |||||||
| 内容記述 | 1: Introduction||2: Literature Review and Research Objectives||3: Methods and Algorithms||4: Field Survey Results||5: Conclusions and Future Work | |||||||
| 備考 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 九州工業大学博士学位論文 学位記番号：生工博甲第369号 学位授与年月日：令和2年3月25日 | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | マンガンクラスト （Mn-crust） | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Autonomous Underwater Vehicles （AUV） | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Machine Learning | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Subbottom acousttics | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Deep-sea survey | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Sensor fusion | |||||||
| アドバイザー | ||||||||
| 石井, 和男 | ||||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第369号 | |||||||
| 学位名 | ||||||||
| 学位名 | 博士（工学） | |||||||
| 学位授与年月日 | ||||||||
| 学位授与年月日 | 2020-03-25 | |||||||
| 学位授与機関 | ||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||
| 学位授与機関識別子 | 17104 | |||||||
| 学位授与機関名 | 九州工業大学 | |||||||
| 学位授与年度 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 令和元年度 | |||||||
| 出版タイプ | ||||||||
| 出版タイプ | VoR | |||||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||
| アクセス権 | ||||||||
| アクセス権 | open access | |||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| ID登録 | ||||||||
| ID登録 | 10.18997/00007976 | |||||||
| ID登録タイプ | JaLC | |||||||
