Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Pattern Recognition |
| Untertitel | Joint 34th DAGM and 36th OAGM Symposium, Proceedings |
| Seiten | 123-133 |
| Seitenumfang | 11 |
| ISBN (elektronisch) | 978-3-642-32717-9 |
| Publikationsstatus | Veröffentlicht - 2012 |
| Veranstaltung | Joint 34th Symposium of the German Association for Pattern Recognition, DAGM 2012 and 36th Symposium of the Austrian Association for Pattern Recognition, OAGM 2012 - Graz, Österreich Dauer: 28 Aug. 2012 → 31 Aug. 2012 |
Publikationsreihe
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Band | 7476 |
| ISSN (Print) | 0302-9743 |
| ISSN (elektronisch) | 1611-3349 |
Abstract
We present a feature-based framework that combines spatial feature clustering, guided sampling for pose generation, and model updating for 3D object recognition and pose estimation. Existing methods fails in case of repeated patterns or multiple instances of the same object, as they rely only on feature discriminability for matching and on the estimator capabilities for outlier rejection. We propose to spatially separate the features before matching to create smaller clusters containing the object. Then, hypothesis generation is guided by exploiting cues collected off- and on-line, such as feature repeatability, 3D geometric constraints, and feature occurrence frequency. Finally, while previous methods overload the model with synthetic features for wide baseline matching, we claim that continuously updating the model representation is a lighter yet reliable strategy. The evaluation of our algorithm on challenging video sequences shows the improvement provided by our contribution.
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
- Allgemeine Computerwissenschaft
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Pattern Recognition: Joint 34th DAGM and 36th OAGM Symposium, Proceedings. 2012. S. 123-133 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 7476 ).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - 3D Object Recognition and Pose Estimation for Multiple Objects Using Multi-Prioritized RANSAC and Model Updating
AU - Fenzi, Michele
AU - Dragon, Ralf
AU - Leal-Taixé, Laura
AU - Rosenhahn, Bodo
AU - Ostermann, Jörn
PY - 2012
Y1 - 2012
N2 - We present a feature-based framework that combines spatial feature clustering, guided sampling for pose generation, and model updating for 3D object recognition and pose estimation. Existing methods fails in case of repeated patterns or multiple instances of the same object, as they rely only on feature discriminability for matching and on the estimator capabilities for outlier rejection. We propose to spatially separate the features before matching to create smaller clusters containing the object. Then, hypothesis generation is guided by exploiting cues collected off- and on-line, such as feature repeatability, 3D geometric constraints, and feature occurrence frequency. Finally, while previous methods overload the model with synthetic features for wide baseline matching, we claim that continuously updating the model representation is a lighter yet reliable strategy. The evaluation of our algorithm on challenging video sequences shows the improvement provided by our contribution.
AB - We present a feature-based framework that combines spatial feature clustering, guided sampling for pose generation, and model updating for 3D object recognition and pose estimation. Existing methods fails in case of repeated patterns or multiple instances of the same object, as they rely only on feature discriminability for matching and on the estimator capabilities for outlier rejection. We propose to spatially separate the features before matching to create smaller clusters containing the object. Then, hypothesis generation is guided by exploiting cues collected off- and on-line, such as feature repeatability, 3D geometric constraints, and feature occurrence frequency. Finally, while previous methods overload the model with synthetic features for wide baseline matching, we claim that continuously updating the model representation is a lighter yet reliable strategy. The evaluation of our algorithm on challenging video sequences shows the improvement provided by our contribution.
UR - http://www.scopus.com/inward/record.url?scp=84865833134&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-32717-9_13
DO - 10.1007/978-3-642-32717-9_13
M3 - Conference contribution
AN - SCOPUS:84865833134
SN - 9783642327162
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 123
EP - 133
BT - Pattern Recognition
T2 - Joint 34th Symposium of the German Association for Pattern Recognition, DAGM 2012 and 36th Symposium of the Austrian Association for Pattern Recognition, OAGM 2012
Y2 - 28 August 2012 through 31 August 2012
ER -