Pregled bibliografske jedinice broj: 1122415
Large-scale image mosaicking using multimodal hyperedge constraints from multiple registration methods within the Generalized Graph SLAM framework
Large-scale image mosaicking using multimodal hyperedge constraints from multiple registration methods within the Generalized Graph SLAM framework // 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems
Chicago (IL), Sjedinjene Američke Države: Institute of Electrical and Electronics Engineers (IEEE), 2014. str. 4564-4570 doi:10.1109/iros.2014.6943209 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 1122415 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Large-scale image mosaicking using multimodal
hyperedge constraints from multiple registration
methods within the Generalized Graph SLAM
framework
Autori
Pfingsthorn, Max ; Birk, Andreas ; Ferreira, Fausto ; Veruggio, Gianmarco ; Caccia, Massimo ; Bruzzone, Gabriele
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
2014 IEEE/RSJ International Conference on Intelligent Robots and Systems
/ - : Institute of Electrical and Electronics Engineers (IEEE), 2014, 4564-4570
ISBN
978-1-4799-6934-0
Skup
2014 IEEE/RSJ International Conference on Intelligent Robots and Systems
Mjesto i datum
Chicago (IL), Sjedinjene Američke Države, 14.09.2014. - 18.09.2014
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
graph theory , image registration , image segmentation , marine control , mobile robots , robot vision , robust control , SLAM (robots) , underwater vehicles
Sažetak
Underwater image mosaicking is an important tool for visual surveys, object detection, and as a means to control the underwater robot if done online. Such application areas can benefit significantly from a recent focus on robust methods for graph-based Simultaneous Localization and Mapping (SLAM). This paper focuses on two contributions: An approach to combine registration results from multiple methods in multimodal constraints and, up to the authors' knowledge, the first method to generate hyperedge constraints from state-of-the-art place recognition techniques. Both contributions are implemented within the Generalized Graph SLAM framework. Experimental results show that the methods generate informative constraints and that the authors' Prefilter method outperforms related methods on a large underwater image dataset processed with these methods.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika, Računarstvo
