Recognition by Linear Combinations of Models
Visual object recognition requires the matching of an image with a set of models stored in memory. In this paper we propose an approach to recognition in which a 3-D object is represented by the linear combination of 2-D images of the object. If M = {M1,...Mk} is the set of pictures representing a g...
| Main Authors: | , |
|---|---|
| Language: | en_US |
| Published: |
2004
|
| Online Access: | http://hdl.handle.net/1721.1/6516 |
| _version_ | 1826198559601459200 |
|---|---|
| author | Ullman, Shimon Basri, Ronen |
| author_facet | Ullman, Shimon Basri, Ronen |
| author_sort | Ullman, Shimon |
| collection | MIT |
| description | Visual object recognition requires the matching of an image with a set of models stored in memory. In this paper we propose an approach to recognition in which a 3-D object is represented by the linear combination of 2-D images of the object. If M = {M1,...Mk} is the set of pictures representing a given object, and P is the 2-D image of an object to be recognized, then P is considered an instance of M if P = Eki=aiMi for some constants ai. We show that this approach handles correctly rigid 3-D transformations of objects with sharp as well as smooth boundaries, and can also handle non-rigid transformations. The paper is divided into two parts. In the first part we show that the variety of views depicting the same object under different transformations can often be expressed as the linear combinations of a small number of views. In the second part we suggest how this linear combinatino property may be used in the recognition process. |
| first_indexed | 2024-09-23T11:06:46Z |
| id | mit-1721.1/6516 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T11:06:46Z |
| publishDate | 2004 |
| record_format | dspace |
| spelling | mit-1721.1/65162019-04-12T08:31:21Z Recognition by Linear Combinations of Models Ullman, Shimon Basri, Ronen Visual object recognition requires the matching of an image with a set of models stored in memory. In this paper we propose an approach to recognition in which a 3-D object is represented by the linear combination of 2-D images of the object. If M = {M1,...Mk} is the set of pictures representing a given object, and P is the 2-D image of an object to be recognized, then P is considered an instance of M if P = Eki=aiMi for some constants ai. We show that this approach handles correctly rigid 3-D transformations of objects with sharp as well as smooth boundaries, and can also handle non-rigid transformations. The paper is divided into two parts. In the first part we show that the variety of views depicting the same object under different transformations can often be expressed as the linear combinations of a small number of views. In the second part we suggest how this linear combinatino property may be used in the recognition process. 2004-10-04T15:14:26Z 2004-10-04T15:14:26Z 1989-08-01 AIM-1152 http://hdl.handle.net/1721.1/6516 en_US AIM-1152 3715484 bytes 2640916 bytes application/postscript application/pdf application/postscript application/pdf |
| spellingShingle | Ullman, Shimon Basri, Ronen Recognition by Linear Combinations of Models |
| title | Recognition by Linear Combinations of Models |
| title_full | Recognition by Linear Combinations of Models |
| title_fullStr | Recognition by Linear Combinations of Models |
| title_full_unstemmed | Recognition by Linear Combinations of Models |
| title_short | Recognition by Linear Combinations of Models |
| title_sort | recognition by linear combinations of models |
| url | http://hdl.handle.net/1721.1/6516 |
| work_keys_str_mv | AT ullmanshimon recognitionbylinearcombinationsofmodels AT basrironen recognitionbylinearcombinationsofmodels |