Abstract
Representations for many three dimensional object recognition systems lack highly salient visual details. Here, eight second-order volumetric primitives are defined to extend the range of descriptive features usable for object representation and thus recognition. While defined independent from the sensory modality, the shapes can be extracted from range data, using a classification based on local surface shape. With descriptions of both model and scene features from this vocabulary, model matching is more efficient, because the rich shape vocabulary reduces the combinatorial generation of hypotheses. With a set of model-to-data correspondences, accurate three dimensional model location is possible. Results for each stage of the process are shown.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
