An unsupervised neural network classifier for automatic aerial image recognition

Abstract

This article describes the application of the adaptive resonance theory (ART 2-A) network to the problem of automatic aerial image recognition (AAIR). The classification of aerial images independently of their position and orientation is required for automatic tracking and target recognition. Invariance is achieved by using different invariant feature spaces in combination with an unsupervised neural network. The performance of the neural network based classifier in conjunction with several types of invariant AAIR global features, such as the Fourier transform (FT) space, Zernike moments, central moments and polar transforms, are examined. The advantages of this approach are discussed. The ART 2-A distinguished itself with its speed and low number of training vectors. Although a large image data base would be necessary before this approach could be fully validated, the initial results are very promising.