{"created":"2023-06-19T10:29:42.467858+00:00","id":9685,"links":{},"metadata":{"_buckets":{"deposit":"7437fa06-ea59-452b-973d-256b46d052d0"},"_deposit":{"created_by":18,"id":"9685","owners":[18],"pid":{"revision_id":0,"type":"depid","value":"9685"},"status":"published"},"_oai":{"id":"oai:muroran-it.repo.nii.ac.jp:00009685","sets":["41:227"]},"author_link":["55182"],"item_81_date_granted_17":{"attribute_name":"学位授与年月日","attribute_value_mlt":[{"subitem_dategranted":"2018-03-23"}]},"item_81_degree_grantor_10":{"attribute_name":"学位授与機関","attribute_value_mlt":[{"subitem_degreegrantor":[{"subitem_degreegrantor_language":"ja","subitem_degreegrantor_name":"室蘭工業大学"},{"subitem_degreegrantor_language":"en","subitem_degreegrantor_name":"Muroran Institute of Technology"}],"subitem_degreegrantor_identifier":[{"subitem_degreegrantor_identifier_name":"10103","subitem_degreegrantor_identifier_scheme":"kakenhi"}]}]},"item_81_degree_name_11":{"attribute_name":"学位名","attribute_value_mlt":[{"subitem_degreename":"博士（工学）","subitem_degreename_language":"ja"}]},"item_81_description_25":{"attribute_name":"フォーマット","attribute_value_mlt":[{"subitem_description":"application/pdf","subitem_description_type":"Other"}]},"item_81_description_7":{"attribute_name":"抄録","attribute_value_mlt":[{"subitem_description":"高レベル放射線廃棄物の処分場設計においては、周辺岩盤の力学的安定性や水理特性等を、モデルを通じて評価する。対象岩盤が亀裂性の特徴を有する場合、モデル化のために岩盤中の不連続面の物理的特性を取得する必要がある。不連続面は地質専門家による目視での地質観察に基づいて取得されるが、その際の省力化と平準化が求められている。そのために注目を集めているのが、3Dレーザースキャナを用いて得られる点群データである。しかし、不連続面の取得に際し手動での操作が多く、自動的な点群処理を行っている例は少ない。そのため、現状のままでは、高レベル放射線廃棄物のような大規模な地下施設の岩盤評価に活用できない。そこで本研究では、点群データから不連続面を自動で抽出する処理について検討し、点群から自動で面を推定する可変格子分割アルゴリズム（Variable-Box Segmentation ：VBS）を開発した。また、推定した面について不連続面を判別する処理についても検討し、表面粗さに関する従来のハンドクラフトな特徴量とディープラーンニングにより自動抽出した特徴量の2つに関してSupport Vector Machine（SVM）による判別を試みた。具体的な研究内容は、VBSアルゴリズムの開発については、アルゴリズムを「大格子分割」「小格子分割」「結合」の３つのプロセスから構成し、場所により適切なサイズの格子をボトムアップ的に定め、適切な面を推定した。実験では、実際の点群データからVBSにより坑道壁面の表面形状を推定し、従来処理のアルゴリズムと比べ、VBSの判別精度や自動化への寄与について検討した。不連続面の判別処理については、ハンドクラフトな特徴量として、4つの表面粗さ（フラクタル次元・算術平均粗さ・最大高さ粗さ・Roughness-length）を用いて、SVMによる判別を試み、適合率・再現率・F値を確認した。自動抽出した特徴量については、点群データをレンジ画像データに変換し、Convolutional Neural Network（CNN）の一種であるAlexNetから特徴量抽出を試み、ハンドクラフトな特徴量と比べ再現率の増加を確認した。以上の結果から、不連続面・掘削面共に約6割の精度で判別できることを確認した。すなわち、地質観察の省力化や平準化に対して、本研究で提案するVBSを用いて点郡データから適切な面を自動推定した後、CNNで表面の特徴量を抽出しSVMで不連続面を学習・判別する手法は充分活用できると判断する。従って、点郡データを用いた不連続面の自動推定は将来の処分場設計や設置場所の選定等の検討に役立つことができると考える。","subitem_description_language":"ja","subitem_description_type":"Abstract"},{"subitem_description":"The mechanical stabilities and water-flow properties around the rock mass are valid from a model in designing of the high-level radioactive wastes final disposal. For modeling, it is necessary to obtain the properties of the discontinuities in the rock mass. The geologists survey discontinuities from the rock mass manually, then the qualities of the obtained properties are depending on the skill of the geologist. It is therefore necessary to optimize and standardize the surveying method. The three-dimensional point cloud is focused on the optimization and standardization of the surveying, recently. However, all the developed method to obtain discontinuities processes point clouds manually or semi-automatically using some tools. Therefore, we developed the Variable-Box Segmentation algorithm (VBS) to estimate planes from point cloud automatically. The VBS has three processes: first segmentation, second segmentation and combining. The box suited to the surface shape are obtained through these processes, and after that suitable planes are estimated. The estimated planes are compared with planes estimated using DiAna, the semi-automatic method. The result of comparison shows that the VBS planes are more similar to planes estimated manually than the DiAna planes. Next, we tried to discriminate the estimated planes between discontinuity and excavated plane using four hand-craft surface features, fractal dimension, arithmetic average roughness, maximum height roughness and roughness length, through the Support Vector Machine (SVM). However, the precisions, recalls and F values are not appropriate to practical application. We therefore obtained features automatically using AlexNet, a Convolutional Neural Network. The recalls are higher than the recalls of the hand-craft features. In the future, the three-dimensional point clouds are obtained from the rock mass and the discontinuities are obtained automatically using the VBS algorithm developed in this research. The surveying of the rock mass will be optimized and standardize by VBS and it will support the construction of the high-level radioactive wastes repository.","subitem_description_language":"en","subitem_description_type":"Abstract"}]},"item_81_dissertation_number_13":{"attribute_name":"学位授与番号","attribute_value_mlt":[{"subitem_dissertationnumber":"甲第416号"}]},"item_81_identifier_registration":{"attribute_name":"ID登録","attribute_value_mlt":[{"subitem_identifier_reg_text":"10.15118/00009637","subitem_identifier_reg_type":"JaLC"}]},"item_81_subject_9":{"attribute_name":"日本十進分類法","attribute_value_mlt":[{"subitem_subject":"501","subitem_subject_scheme":"NDC"}]},"item_81_text_12":{"attribute_name":"学位の種別","attribute_value_mlt":[{"subitem_text_language":"ja","subitem_text_value":"課程博士"}]},"item_81_text_14":{"attribute_name":"報告番号","attribute_value_mlt":[{"subitem_text_language":"ja","subitem_text_value":"甲第416号"}]},"item_81_text_15":{"attribute_name":"学位記番号","attribute_value_mlt":[{"subitem_text_language":"ja","subitem_text_value":"博甲第416号"}]},"item_81_text_16":{"attribute_name":"研究科・専攻","attribute_value_mlt":[{"subitem_text_language":"ja","subitem_text_value":"工学専攻・先端環境創生工学コース"}]},"item_81_version_type_24":{"attribute_name":"著者版フラグ","attribute_value_mlt":[{"subitem_version_resource":"http://purl.org/coar/version/c_970fb48d4fbd8a85","subitem_version_type":"VoR"}]},"item_access_right":{"attribute_name":"アクセス権","attribute_value_mlt":[{"subitem_access_right":"open access","subitem_access_right_uri":"http://purl.org/coar/access_right/c_abf2"}]},"item_creator":{"attribute_name":"著者","attribute_type":"creator","attribute_value_mlt":[{"creatorAffiliations":[{"affiliationNameIdentifiers":[],"affiliationNames":[{"affiliationName":""}]}],"creatorNames":[{"creatorName":"松川, 瞬","creatorNameLang":"ja"},{"creatorName":"MATSUKAWA, Syun","creatorNameLang":"en"},{"creatorName":"マツカワ, シュン","creatorNameLang":"ja-Kana"}],"familyNames":[{},{},{}],"givenNames":[{},{},{}],"nameIdentifiers":[{}]}]},"item_files":{"attribute_name":"ファイル情報","attribute_type":"file","attribute_value_mlt":[{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2018-06-06"}],"displaytype":"detail","filename":"A416.pdf","filesize":[{"value":"12.6 MB"}],"format":"application/pdf","licensetype":"license_note","mimetype":"application/pdf","url":{"label":"A416","objectType":"fulltext","url":"https://muroran-it.repo.nii.ac.jp/record/9685/files/A416.pdf"},"version_id":"40ec8e30-b1db-4aeb-a6c9-a7598d13d9d7"},{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2018-06-06"}],"displaytype":"detail","filename":"A416_summary.pdf","filesize":[{"value":"186.3 kB"}],"format":"application/pdf","licensetype":"license_note","mimetype":"application/pdf","url":{"label":"A416_summary","objectType":"abstract","url":"https://muroran-it.repo.nii.ac.jp/record/9685/files/A416_summary.pdf"},"version_id":"47f10410-9d7e-4fa6-ab5b-07a1531c4467"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourcetype":"doctoral thesis","resourceuri":"http://purl.org/coar/resource_type/c_db06"}]},"item_title":"３次元点群データを用いた坑道壁面における不連続面の自動推定に関する研究","item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"３次元点群データを用いた坑道壁面における不連続面の自動推定に関する研究","subitem_title_language":"ja"}]},"item_type_id":"81","owner":"18","path":["227"],"pubdate":{"attribute_name":"PubDate","attribute_value":"2018-06-06"},"publish_date":"2018-06-06","publish_status":"0","recid":"9685","relation_version_is_last":true,"title":["３次元点群データを用いた坑道壁面における不連続面の自動推定に関する研究"],"weko_creator_id":"18","weko_shared_id":-1},"updated":"2024-01-22T01:45:04.559447+00:00"}