Fitting implicit polynomials to use as features in image understanding
N.J. Redding · 2002
The paper presents a method for extracting features from images using implicit polynomials in two dimensions. These features have been particularly designed to be appropriate for the image understanding problem involving oblique ionograms, but they are more generally applicable. Attention is given to the problem of selecting an appropriate model for the features, choosing a suitable fitting metric, and a robust procedure for achieving a good fit.