| Item type |
SIG Technical Reports(1) |
| 公開日 |
2015-01-15 |
| タイトル |
|
|
タイトル |
隠消現実感における隠背景投影誤差の軽減（3）－隠背景平面拘束の緩和とVDTMの利用－ |
| 言語 |
|
|
言語 |
jpn |
| 資源タイプ |
|
|
資源タイプ識別子 |
http://purl.org/coar/resource_type/c_18gh |
|
資源タイプ |
technical report |
| 著者所属 |
|
|
|
立命館大学大学院情報理工学研究科 |
| 著者所属 |
|
|
|
立命館大学大学院情報理工学研究科 |
| 著者所属 |
|
|
|
立命館大学情報理工学部 |
| 著者所属 |
|
|
|
立命館大学情報理工学部 |
| 著者所属 |
|
|
|
立命館大学情報理工学部 |
| 著者所属 |
|
|
|
立命館大学総合科学技術研究機構 |
| 著者名 |
森, 尚平
酒井, 章伸
松見, 優一
柴田, 史久
木村, 朝子
田村, 秀行
|
| 論文抄録 |
|
|
内容記述タイプ |
Other |
|
内容記述 |
This paper describes a Diminished Reality (DR) technique using View-Dependent Texture Mapping (VDTM) for visually removing real objects in a scene. Our previous work warps one of pre-observed hidden view images and overlays it to the source view on the assumption that the scene consists of planar surfaces. In this paper, the previous work is extended to be applied to 3D scenes. Our main contributions are the followings; 3D hidden scene recovery using VDTM and vision-based 6 degrees of freedom camera tracking using VDTM. Proposed method reconstructs a 3D scene and describes it as 3D surfaces, images, and camera poses corresponding to the images. Hidden views in the source view are recovered using a VDTM manner. To obtain input camera pose for VDTM, the system estimates camera pose by tracking feature points between a recovered view and the source view. |
| 論文抄録(英) |
|
|
内容記述タイプ |
Other |
|
内容記述 |
This paper describes a Diminished Reality (DR) technique using View-Dependent Texture Mapping (VDTM) for visually removing real objects in a scene. Our previous work warps one of pre-observed hidden view images and overlays it to the source view on the assumption that the scene consists of planar surfaces. In this paper, the previous work is extended to be applied to 3D scenes. Our main contributions are the followings; 3D hidden scene recovery using VDTM and vision-based 6 degrees of freedom camera tracking using VDTM. Proposed method reconstructs a 3D scene and describes it as 3D surfaces, images, and camera poses corresponding to the images. Hidden views in the source view are recovered using a VDTM manner. To obtain input camera pose for VDTM, the system estimates camera pose by tracking feature points between a recovered view and the source view. |
| 書誌レコードID |
|
|
収録物識別子タイプ |
NCID |
|
収録物識別子 |
AA11131797 |
| 書誌情報 |
研究報告コンピュータビジョンとイメージメディア（CVIM）
巻 2015-CVIM-195,
号 6,
p. 1-6,
発行日 2015-01-15
|
| Notice |
|
|
|
SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc. |
| 出版者 |
|
|
言語 |
ja |
|
出版者 |
情報処理学会 |
IPSJ-CVIM15195006.pdf (6.3 MB)
