3-D object recognition through next view planning

Abstract

DOI: 10.1049/ic.2009.0115 ISBN: 978-1-84919-193-7 Location: Maharashtra, India Conference date: 19-20 Sept. 2009 Format: PDF This paper summarises some work in active 3-D object recognition through next view planning, using an uncalibrated mobile camera. We delve on two strategies. The first is based on aspect graph-based modelling and recognition using noisy sensors. The second handles the case when the complete 3-D object does not fit into the camera's field of view. Both have a hierarchical knowledge representation scheme, and a robust probabilistic reasoning-based reactive object recognition strategy, which is scalable to memory and processing constraints. (4 pages) Inspec keywords: object recognition; uncertainty handling; inference mechanisms; knowledge representation Subjects: Image recognition; Computer vision and image processing techniques; Knowledge engineering techniques; Reasoning and inference in AI (theory); Image recognition