Fuzzy ARTMAP classification of invariant features derived using angle of rotation from a neural network

Conventional regular moment functions have been proposed as pattern sensitive features in image classi®cation and recognition applications. But conventional regular moments are only invariant to translation, rotation and equal scaling. It is shown that the conventional regular moment invariants remain no longer invariant when the image is scaled unequally in the xand y-axis directions. We address this problem by presenting a technique to make the regular moment functions invariant to unequal scaling. However, the technique produces a set of features that are only invariant to translation, unequal/equal scaling and re ̄ection. They are not invariant to rotation. To make them invariant to rotation, moments are calculated with respect to the principal axis of the image. To perform this, the exact angle of rotation must be known. But the method of using the second-order moments to determine this angle will also be inclusive of an undesired tilt angle. Therefore, in order to correctly determine the amount of rotation, the tilt angle which di€ers for di€erent scaling factors in the xand y-axis directions for the particular image must be obtained. In order to solve this problem, a neural network using the back-propagation learning algorithm is trained to estimate the tilt angle of the image and from this the amount of rotation for the image can be determined. Next, the new moments are derived and a Fuzzy ARTMAP network is used to classify these Information Sciences 000 (2000) 000±000 ELSEVIER SCIENCE Inc. [DTD 4.1.0] JOURNAL INS ARTICLE No. 6429 PAGES 01-18 DISPATCH 27 Octobe r 2000 INS 6429 P R O D . T Y P E : F RO M D IS K A www.elsevier.com/locate/ins * Corresponding author. Tel.: +60-379-595-253; fax: +60-377-847-981. E-mail address: [email protected] (P. Raveendran). 0020-0255/00/$ see front matter Ó 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 0 0 2 5 5 ( 0 0 ) 0 0 0 8 7 6 UN CO RR EC TE D PR OO F images into their respective classes. Sets of experiments involving images rotated and scaled unequally in the xand y-axis directions are carried out to demonstrate the validity of the proposed technique. Ó 2000 Elsevier Science Inc. All rights reserved.