Generalized Hebbian learning for ellipse fitting
Sudanthi N. R. Wijewickrema, Andrew P. Papliński · Digital Library (University of West Bohemia) · 2005
In this paper, we investigate the use of a neural network employing Genralised Hebbian Learning for the approximation of an image of a hypothetically ellipsoidal object as an ellipse. Further, we discuss how the same algorithm is used with higher dimensional data to model hyperellipsoids, with the basic aim at a specific application, namely the modelling of an object as an ellipsoid given a set of 3-dimensional points.