The paper proposes a reliable method for estimating quadric surfaces from 3D range data in the framework of object recognition and localization or object modelling. Instead of estimating a quadric surface individually the approach fits all the surfaces captured in the scene together taking into account the geometric relationships between them and their specific characteristics. The technique is compared with other methods through experiments performed on real objects.
Naoufel Werghi, Anthony Ashbrook, Robert B. Fisher