A merging procedure for connecting fitted implicit polynomials for features

Author

Redding, Nicholas J. ; Newsam, Garry N.

Author_Institution

Div. of Microwave Radar, Defence Sci. & Technol. Organ., Salisbury, SA, Australia

fYear

1996

fDate

18-20 Nov 1996

Firstpage

299

Lastpage

303

Abstract

Describes a merging procedure to merge features extracted from images. The feature extraction process fits implicit polynomials in two dimensions to curved features in the images. This process sometimes results in more than one feature vector for the one underlying curved feature when the curve is broken into two or more segments. The procedure described is designed to correct this when it occurs. It decides when to merge using a second-order approximation to the joint fit computed using Hessians from each segment´s individual fit. This procedure is designed to be one step in a feature extraction stage for image understanding of oblique ionograms

Keywords

Hessian matrices; curve fitting; feature extraction; image recognition; merging; polynomials; vectors; Hessians; curve segments; curved image features; feature extraction; feature vectors; fitted implicit polynomial connection; image understanding; joint fit; merging procedure; oblique ionograms; second-order approximation; Australia; Curve fitting; Feature extraction; Image segmentation; Joining processes; Merging; Polynomials; Radar imaging; Space technology; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Information Systems, 1996., Australian and New Zealand Conference on

Conference_Location

Adelaide, SA

Print_ISBN

0-7803-3667-4

Type

conf

DOI

10.1109/ANZIIS.1996.573964

Filename

573964