Construction and exploitation of a 3D model from 2D image features
This paper proposes a trainable computer vision approach for visual object registration relative to a collection of training images obtained a priori. The algorithm first identifies whether or not the image belongs to the scene location, and should it belong, it will identify objects of interest wit...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
SPIE
2010
|
| Online Access: | http://hdl.handle.net/1721.1/58559 |
| _version_ | 1811083785261809664 |
|---|---|
| author | Bliss, Nadya T. Sun, Zachary Ni, Karl S. Snavely, Noah |
| author2 | Lincoln Laboratory |
| author_facet | Lincoln Laboratory Bliss, Nadya T. Sun, Zachary Ni, Karl S. Snavely, Noah |
| author_sort | Bliss, Nadya T. |
| collection | MIT |
| description | This paper proposes a trainable computer vision approach for visual object registration relative to a collection of training images obtained a priori. The algorithm first identifies whether or not the image belongs to the scene location, and should it belong, it will identify objects of interest within the image and geo-register them. To accomplish this task, the processing chain relies on 3-D structure derived from motion to represent feature locations in a proposed model. Using current state-of- the-art algorithms, detected objects are extracted and their two-dimensional sizes in pixel quantities are converted into relative 3-D real-world coordinates using scene information, homography, and camera geometry. Locations can then be given with distance alignment information. The tasks can be accomplished in an efficient manner. Finally, algorithmic evaluation is presented with receiver operating characteristics, computational analysis, and registration errors in physical distances. |
| first_indexed | 2024-09-23T12:39:04Z |
| format | Article |
| id | mit-1721.1/58559 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T12:39:04Z |
| publishDate | 2010 |
| publisher | SPIE |
| record_format | dspace |
| spelling | mit-1721.1/585592022-10-01T10:18:11Z Construction and exploitation of a 3D model from 2D image features Bliss, Nadya T. Sun, Zachary Ni, Karl S. Snavely, Noah Lincoln Laboratory Bliss, Nadya T. Bliss, Nadya T. Sun, Zachary Ni, Karl S. This paper proposes a trainable computer vision approach for visual object registration relative to a collection of training images obtained a priori. The algorithm first identifies whether or not the image belongs to the scene location, and should it belong, it will identify objects of interest within the image and geo-register them. To accomplish this task, the processing chain relies on 3-D structure derived from motion to represent feature locations in a proposed model. Using current state-of- the-art algorithms, detected objects are extracted and their two-dimensional sizes in pixel quantities are converted into relative 3-D real-world coordinates using scene information, homography, and camera geometry. Locations can then be given with distance alignment information. The tasks can be accomplished in an efficient manner. Finally, algorithmic evaluation is presented with receiver operating characteristics, computational analysis, and registration errors in physical distances. United States. Dept. of the Air Force (contract FA8721-05-C-0002) 2010-09-15T20:44:12Z 2010-09-15T20:44:12Z 2010-01 Article http://purl.org/eprint/type/JournalArticle 0277-786X Proc. SPIE, Vol. 7533, 75330J (2010) http://hdl.handle.net/1721.1/58559 Ni, Karl et al. “Construction and exploitation of a 3D model from 2D image features.” Computational Imaging VIII. Ed. Charles A. Bouman, Ilya Pollak, & Patrick J. Wolfe. San Jose, California, USA: SPIE, 2010. 75330J-10. ©201 SPIE. en_US http://dx.doi.org/10.1117/12.849919 Proceedings of SPIE--the International Society for Optical Engineering Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf SPIE SPIE |
| spellingShingle | Bliss, Nadya T. Sun, Zachary Ni, Karl S. Snavely, Noah Construction and exploitation of a 3D model from 2D image features |
| title | Construction and exploitation of a 3D model from 2D image features |
| title_full | Construction and exploitation of a 3D model from 2D image features |
| title_fullStr | Construction and exploitation of a 3D model from 2D image features |
| title_full_unstemmed | Construction and exploitation of a 3D model from 2D image features |
| title_short | Construction and exploitation of a 3D model from 2D image features |
| title_sort | construction and exploitation of a 3d model from 2d image features |
| url | http://hdl.handle.net/1721.1/58559 |
| work_keys_str_mv | AT blissnadyat constructionandexploitationofa3dmodelfrom2dimagefeatures AT sunzachary constructionandexploitationofa3dmodelfrom2dimagefeatures AT nikarls constructionandexploitationofa3dmodelfrom2dimagefeatures AT snavelynoah constructionandexploitationofa3dmodelfrom2dimagefeatures |