{"created":"2021-03-01T06:20:46.193965+00:00","id":4113,"links":{},"metadata":{"_buckets":{"deposit":"320828a3-7899-443f-bdfd-e50738245c6d"},"_deposit":{"id":"4113","owners":[],"pid":{"revision_id":0,"type":"depid","value":"4113"},"status":"published"},"_oai":{"id":"oai:repository.dl.itc.u-tokyo.ac.jp:00004113","sets":["34:105:330","9:233:280"]},"item_7_alternative_title_1":{"attribute_name":"その他のタイトル","attribute_value_mlt":[{"subitem_alternative_title":"照度差ステレオの一般化 : 3次元モデル化のための光学的手法と幾何学的手法の融合"}]},"item_7_biblio_info_7":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicIssueDates":{"bibliographicIssueDate":"2011-03-24","bibliographicIssueDateType":"Issued"},"bibliographic_titles":[{}]}]},"item_7_date_granted_25":{"attribute_name":"学位授与年月日","attribute_value_mlt":[{"subitem_dategranted":"2011-03-24"}]},"item_7_degree_grantor_23":{"attribute_name":"学位授与機関","attribute_value_mlt":[{"subitem_degreegrantor":[{"subitem_degreegrantor_name":"University of Tokyo (東京大学)"}]}]},"item_7_degree_name_20":{"attribute_name":"学位名","attribute_value_mlt":[{"subitem_degreename":"博士(情報理工学)"}]},"item_7_description_5":{"attribute_name":"抄録","attribute_value_mlt":[{"subitem_description":"Three-dimensional (3-D) models created by computer vision and computer graphics are widely used in many important applications and for a variety of purposes. However, creating 3-D models manually involves significant costs. Therefore, automatic generation of such models is attracting considerable interest, and this process in turn requires efficient 3-D scanning. Numerous systems exist for 3-D scanning using a laser range sensor and image-based methods such as multi-view stereo, structured light, and photometric stereo. These are classified into two approaches: a photometric approach and a geometric approach. Generally, the photometric approach, as typified by photometric stereo, estimates surface normals to be suitable for representing fine details of the surfaces. On the other hand, the geometric approach recovers a depth of the target object to acquire a rough shape. This dissertation proposes efficient 3-D modeling methods that combine photometric and geometric approaches. Combining photometric stereo with a laser range sensor or multi-view stereo allows us to introduce practical constraints for 3-D modeling. Also, we propose practical methods of photometric stereo for handling object with various reflection properties. The first method fuses a laser range sensor and a camera with an attached camera flash for 3-D modeling. The laser range sensor captures the basic shape of the target object. Meanwhile, photometric stereo estimates surface normals as a bump map for detailed surfaces. Then accurate reflection parameters are estimated by using the surface normals. The second method combines photometric stereo and multi-view stereo to simultaneously estimate shape and surface normals. The method uses a simple configuration with a camera and a camera flash. Furthermore, by using color information, the method is extended to robustly handle specularities and occlusions. Generally, photometric stereo assumes that the target object has only diffuse reflection. However, many materials have specular reflection components, which cause an estimation error in photometric stereo. To overcome this problem, a real-time specular removal method is proposed. With a known light source color, the method enables the removal of specular reflection components faster than conventional methods by using a defined color space. In addition, a new photometric stereo method is proposed to handle a wide range of surface reflectances. The method avoids imposing restricting assumptions on surface reflectances and expands the applicability of photometric stereo using three reflection properties. Moreover, our method eliminates the necessity of radiometric calibration and any dependency on ambient illumination. In this dissertation, the theory of these methods is presented, and qualitative and quantitative experiments are performed to demonstrate the effectiveness of each method.","subitem_description_type":"Abstract"},{"subitem_description":"近年コンピュータビジョンやコンピュータグラフィックスの技術を駆使した3次元モデルは, 「3D元年」という言葉にも代表されるように, 多くのアプリケーションや様々な目的のために幅広く利用されてきている. しかし3次元モデルを人手で作るコストや手間は膨大であるため, 3次元モデルを自動で生成しようとする試みに関心が高まってきているとともに, 効果的な3次元計測手法が必要とされてきている. 3次元モデルを計測する方法には, レーザーレンジセンサを用いたり, 画像ベースで推定をおこなう多視点ステレオ, 光切断法, 照度差ステレオなどの様々な手法が存在する. これらの手法は大きく分けて光学的手法と幾何学的手法に分けることができる. 一般的に光学的手法は, 照度差ステレオに代表されるように, 物体表面の法線を推定する手法であり, 細かな凹凸や滑らかな面を推定するのに適している. 一方で幾何学的手法は対象物体の奥行きを計測する手法であり, 物体の外形を推定するのに適している. 本論文ではこれらの相反する光学的手法と幾何学的手法を組合せることで, より効果的な3次元モデル化手法の提案をおこなう. 具体的には, 光学的手法である照度差ステレオと, レーザーレンジセンサや, 多視点ステレオとを組み合わせることで, 新たな拘束を導いて3次元モデル化をおこなう. さらに, 様々な反射特性の物体に対して照度差ステレオを扱えるようにするための手法を提案する. 1つ目の手法はレンジセンサとフラッシュ付きのカメラを融合した3次元モデル化手法である. 大まかな形状をレーザレンジセンサで取得し, 細かな凹凸などは照度差ステレオによって法線を求め, バンプマップとして表現する. さらに得られた法線を用いて正確な反射パラメータを推定する. 2つ目の手法は照度差ステレオと多視点ステレオを組合せることで, 法線と形状を同時に推定する手法である. 本手法はレーザーレンジセンサを使わず, カメラとフラッシュだけの簡易なセットアップで3次元モデルを推定することができる. さらに色情報を利用することによって本手法を拡張し, 鏡面反射や遮蔽にロバストな手法を提案する. 一般的に照度差ステレオを用いる場合は, 対象物体が拡散反射であることを仮定する. しかし物体の中には鏡面反射成分が含まれる物体も数多く存在し, これが推定に悪影響を与えてしまう, そこでリアルタイムに鏡面反射成分を除去する手法を提案する. 本手法は光源色を既知として独自の色空間を用いることで, 従来手法よりも高速に鏡面反射成分を除去することができる. さらに, 鏡面反射成分だけでなく, 幅広い反射特性の物体の法線を推定するための手法を提案する. 特定の反射モデルを仮定せず, 基本的な反射特性から導いた拘束を用いて法線を推定する. 本手法はカメラの特性や環境光の影響に対してもロバストである. 本論文では, これらの手法の原理を示し, 定性的及び定量的な実験を通してそれぞれの手法の実用性を示す.","subitem_description_type":"Abstract"}]},"item_7_dissertation_number_26":{"attribute_name":"学位授与番号","attribute_value_mlt":[{"subitem_dissertationnumber":"甲第27291号"}]},"item_7_full_name_3":{"attribute_name":"著者別名","attribute_value_mlt":[{"nameIdentifiers":[{"nameIdentifier":"9470","nameIdentifierScheme":"WEKO"}],"names":[{"name":"肥後, 智昭"}]}]},"item_7_identifier_registration":{"attribute_name":"ID登録","attribute_value_mlt":[{"subitem_identifier_reg_text":"10.15083/00004104","subitem_identifier_reg_type":"JaLC"}]},"item_7_select_21":{"attribute_name":"学位","attribute_value_mlt":[{"subitem_select_item":"doctoral"}]},"item_7_subject_13":{"attribute_name":"日本十進分類法","attribute_value_mlt":[{"subitem_subject":"007","subitem_subject_scheme":"NDC"}]},"item_7_text_22":{"attribute_name":"学位分野","attribute_value_mlt":[{"subitem_text_value":"Information Science and Technology (情報理工学)"}]},"item_7_text_24":{"attribute_name":"研究科・専攻","attribute_value_mlt":[{"subitem_text_value":"Department of Information and Communication Engineering, Graduate School of Information Science and Technology (情報理工学系研究科電子情報学専攻)"}]},"item_7_text_27":{"attribute_name":"学位記番号","attribute_value_mlt":[{"subitem_text_value":"博情第329号"}]},"item_7_text_4":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"東京大学大学院情報理工学系研究科電子情報学専攻"}]},"item_creator":{"attribute_name":"著者","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Higo, Tomoaki"}],"nameIdentifiers":[{"nameIdentifier":"9469","nameIdentifierScheme":"WEKO"}]}]},"item_files":{"attribute_name":"ファイル情報","attribute_type":"file","attribute_value_mlt":[{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2017-06-01"}],"displaytype":"detail","filename":"48087413.pdf","filesize":[{"value":"7.6 MB"}],"format":"application/pdf","licensetype":"license_note","mimetype":"application/pdf","url":{"label":"48087413.pdf","url":"https://repository.dl.itc.u-tokyo.ac.jp/record/4113/files/48087413.pdf"},"version_id":"2df658fb-a59e-4b8c-81b5-a6fdcd7105c3"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"eng"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourcetype":"thesis","resourceuri":"http://purl.org/coar/resource_type/c_46ec"}]},"item_title":"GENERALIZATION OF PHOTOMETRIC STEREO : FUSING PHOTOMETRIC AND GEOMETRIC APPROACHES FOR 3-D MODELING","item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"GENERALIZATION OF PHOTOMETRIC STEREO : FUSING PHOTOMETRIC AND GEOMETRIC APPROACHES FOR 3-D MODELING"}]},"item_type_id":"7","owner":"1","path":["280","330"],"pubdate":{"attribute_name":"公開日","attribute_value":"2012-11-09"},"publish_date":"2012-11-09","publish_status":"0","recid":"4113","relation_version_is_last":true,"title":["GENERALIZATION OF PHOTOMETRIC STEREO : FUSING PHOTOMETRIC AND GEOMETRIC APPROACHES FOR 3-D MODELING"],"weko_creator_id":"1","weko_shared_id":null},"updated":"2022-12-19T03:45:53.586185+00:00"}