Abstract

This paper presents a geometric approach to recognizing smooth ob jects from their outlines. We define a signature function that associates feature vectors with ob jects and baselines connecting pairs of possible viewpoints. Feature vectors, which can be pro jective, affine, or Euclidean, are computed using the planes that pass through a fixed baseline and are also tangent to the ob ject’s surface. In the proposed framework, matching a test outline to a set of training outlines is equiv- alent to finding intersections in feature space between the images of the training and the test signature functions. The paper presents experimen- tal results for the case of internally calibrated perspective cameras, where the feature vectors are angles between epipolar tangent planes.