Wir prasentieren einen Ansatz, um lange Bildfolgen einer omnidirektionalen Kamera mittels Bundelausgleichung auszuwerten. Wir nutzen Bilder des Multikamerasystems Ladybug3 von PointGrey, welches aus sechs Einzelkameras besteht. Die gegenseitige Uberdeckung aufeinanderfolgender Bilder ist gros; Verknupfungen zwischen weit entfernten Bildern kommen nur uber Schleifenschlusse zustande. Zwei Probleme sind zu losen: (1) Die Bundelausgleichung muss Bilder einer omnidirektionalen Kamera verarbeiten und (2) Ausreissersuche und Naherungswertbestimmung mussen mit der speziellen Aufnahmegeometrie umgehen konnen. Wir losen Problem (1) indem wir die Einzelkameras der Ladybug zu einer virtuellen Kamera zusammenfassen und fur die Bundelausgleichung ein spharisches Modell verwenden. Die Ausreisserdetektion (2) erfolgt uber lokale Bundelausgleichungen benachbarter Bilder und anschliesende robuste Gesamtbundelausgleichung. Ein Inertialnavigationssystem liefert die benotigten Naherungswerte fur die Kamerapositionen. We present a method to perform bundle adjustment using long sequences of digital images from an omnidirectional camera. We use the Ladybug3 camera from PointGrey, which consists of six individual cameras pointing in different directions. There is large overlap between successive images but only a few loop closures provide connections between distant camera positions. We face two challenges: (1) to perform a bundle adjustment with images of an omnidirectional camera and (2) implement outlier detection and estimation of initial parameters for the geometry described above. Our program combines the Ladybug’s individual cameras to a single virtual camera and uses a spherical imaging model within the bundle adjustment, solving problem (1). Outlier detection (2) is done using bundle adjustments with small subsets of images followed by a robust adjustment of all images. Approximate values in our context are taken from an on-board inertial navigation system.
[1]
Michel Dhome,et al.
Generic and real-time structure from motion using local bundle adjustment
,
2009,
Image Vis. Comput..
[2]
David Nistér,et al.
An efficient solution to the five-point relative pose problem
,
2003,
2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..
[3]
Jan-Michael Frahm,et al.
Building Rome on a Cloudless Day
,
2010,
ECCV.
[4]
Manolis I. A. Lourakis,et al.
SBA: A software package for generic sparse bundle adjustment
,
2009,
TOMS.
[5]
Luis Puig,et al.
Visual SLAM with an Omnidirectional Camera
,
2010,
2010 20th International Conference on Pattern Recognition.
[6]
H. Mayer.
ISSUES FOR IMAGE MATCHING IN STRUCTURE FROM MOTION
,
2008
.
[7]
Maxime Lhuillier,et al.
Automatic Structure and Motion using a Catadioptric Camera
,
2005
.
[8]
T. Läbe.
AUTOMATIC RELATIVE ORIENTATION OF IMAGES
,
2006
.
[9]
Taehee Lee,et al.
Robust 3D street-view reconstruction using sky motion estimation
,
2009,
2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops.
[10]
Steven M. Seitz,et al.
Photo tourism: exploring photo collections in 3D
,
2006,
ACM Trans. Graph..
[11]
Wolfgang Förstner,et al.
BENCHMARKING AUTOMATIC BUNDLE ADJUSTMENT RESULTS
,
2008
.
[12]
Robert C. Bolles,et al.
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography
,
1981,
CACM.
[13]
Matthijs C. Dorst.
Distinctive Image Features from Scale-Invariant Keypoints
,
2011
.