Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Pattern Recognition - 40th German Conference, GCPR 2018, Proceedings |
| Herausgeber/-innen | Andrés Bruhn, Mario Fritz, Thomas Brox |
| Herausgeber (Verlag) | Springer Verlag |
| Seiten | 407-421 |
| Seitenumfang | 15 |
| ISBN (Print) | 9783030129385 |
| Publikationsstatus | Veröffentlicht - 9 Okt. 2018 |
| Veranstaltung | 40th German Conference on Pattern Recognition, GCPR 2018 - Stuttgart, Deutschland Dauer: 9 Okt. 2018 → 12 Okt. 2018 |
Publikationsreihe
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Band | 11269 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (elektronisch) | 1611-3349 |
Abstract
In this paper, we propose a new approach for dense depth estimation based on multimodal stereo images. Our approach employs a combined cost function utilizing robust metrics and a transformation to an illumination independent representation. Additionally, we present a confidence based weighting scheme which allows a pixel-wise weight adjustment within the cost function. We demonstrate the capabilities of our approach using RGB- and thermal images. The resulting depth maps are evaluated by comparing them to depth measurements of a Velodyne HDL-64E LiDAR sensor. We show that our method outperforms current state of the art dense matching methods regarding depth estimation based on multimodal input images.
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
- Allgemeine Computerwissenschaft
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Pattern Recognition - 40th German Conference, GCPR 2018, Proceedings. Hrsg. / Andrés Bruhn; Mario Fritz; Thomas Brox. Springer Verlag, 2018. S. 407-421 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 11269 LNCS).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Multimodal Dense Stereo Matching
AU - Mehltretter, Max
AU - Kleinschmidt, Sebastian P.
AU - Wagner, Bernardo
AU - Heipke, Christian
N1 - Funding information:. This work was supported by the German Research Foundation (DFG) as a part of the Research Training Group i.c.sens [GRK2159] and the MOBILISE initiative of the Leibniz Universität Hannover and TU Braunschweig.
PY - 2018/10/9
Y1 - 2018/10/9
N2 - In this paper, we propose a new approach for dense depth estimation based on multimodal stereo images. Our approach employs a combined cost function utilizing robust metrics and a transformation to an illumination independent representation. Additionally, we present a confidence based weighting scheme which allows a pixel-wise weight adjustment within the cost function. We demonstrate the capabilities of our approach using RGB- and thermal images. The resulting depth maps are evaluated by comparing them to depth measurements of a Velodyne HDL-64E LiDAR sensor. We show that our method outperforms current state of the art dense matching methods regarding depth estimation based on multimodal input images.
AB - In this paper, we propose a new approach for dense depth estimation based on multimodal stereo images. Our approach employs a combined cost function utilizing robust metrics and a transformation to an illumination independent representation. Additionally, we present a confidence based weighting scheme which allows a pixel-wise weight adjustment within the cost function. We demonstrate the capabilities of our approach using RGB- and thermal images. The resulting depth maps are evaluated by comparing them to depth measurements of a Velodyne HDL-64E LiDAR sensor. We show that our method outperforms current state of the art dense matching methods regarding depth estimation based on multimodal input images.
UR - http://www.scopus.com/inward/record.url?scp=85063499082&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-12939-2_28
DO - 10.1007/978-3-030-12939-2_28
M3 - Conference contribution
AN - SCOPUS:85063499082
SN - 9783030129385
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 407
EP - 421
BT - Pattern Recognition - 40th German Conference, GCPR 2018, Proceedings
A2 - Bruhn, Andrés
A2 - Fritz, Mario
A2 - Brox, Thomas
PB - Springer Verlag
T2 - 40th German Conference on Pattern Recognition, GCPR 2018
Y2 - 9 October 2018 through 12 October 2018
ER -