DocumentCode :
2517281
Title :
Experts of probabilistic flow subspaces for robust monocular odometry in urban areas
Author :
Herdtweck, Christian ; Curio, Cristóbal
Author_Institution :
Max Planck Inst. for Biol. Cybern., Tubingen, Germany
fYear :
2012
fDate :
3-7 June 2012
Firstpage :
661
Lastpage :
667
Abstract :
Visual odometry has been promoted as a fundamental component for intelligent vehicles. Relying solely on monocular image cues would be desirable. Nevertheless, this is a challenge especially in dynamically varying urban areas due to scale ambiguities, independent motions, and measurement noise. We propose to use probabilistic learning with auxiliar depth cues. Specifically, we developed an expert model that specializes monocular egomotion estimation units on typical scene structures, i.e. statistical variations of scene depth layouts. The framework adaptively selects the best fitting expert. For on-line estimation of egomotion, we adopted a probabilistic subspace flow estimation method. Learning in our framework consists of two components: 1) Partitioning of datasets of video and ground truth odometry data based on unsupervised clustering of dense stereo depth profiles and 2) training a cascade of subspace flow expert models. A probabilistic quality measure from the estimates of the experts provides a selection rule overall leading to improvements of egomotion estimation for long test sequences.
Keywords :
automotive components; distance measurement; expert systems; motion estimation; pattern clustering; probability; stereo image processing; traffic engineering computing; unsupervised learning; video signal processing; auxiliar depth cues; dataset partitioning; dense stereo depth profile; ground truth odometry data; intelligent vehicle component; long test sequence; monocular egomotion estimation unit; monocular image cues; probabilistic flow subspace; probabilistic learning; probabilistic quality measure; probabilistic subspace flow estimation method; robust monocular odometry; scene depth layout; selection rule; statistical variation; subspace flow expert model; unsupervised clustering; urban area; visual odometry; Cameras; Estimation; Nonhomogeneous media; Probabilistic logic; Training; Vectors; Visualization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Vehicles Symposium (IV), 2012 IEEE
Conference_Location :
Alcala de Henares
ISSN :
1931-0587
Print_ISBN :
978-1-4673-2119-8
Type :
conf
DOI :
10.1109/IVS.2012.6232238
Filename :
6232238
Link To Document :
بازگشت