Three-dimensional object recognition using spherical correlation

Although much research has been carried out on feature extraction from 3D object, validity and appropriateness of matching measures have not been given enough consideration. In 3D object recognition, object features are often represented by directional data, and treatment of such data needs different operation from that of the data on Euclidean space. In this paper, the authors propose a new method to define a matching measure between objects, and show that this represents the similarity reasonably well with good robustness. Moreover they show that the reliability of the partially input data can be estimated as a by-product of this method