Skip to Main Content
A method based on the concept of fuzzy sets for automatic target recognition is proposed in this paper. It involves automatic threshold selection, feature extraction, and classification. An optimal threshold is selected by the fuzzy 2D entropic thresholding, namely, so as to separate a given image into foreground (object image) and background. In order to evaluate the thresholding method that we proposed in this paper, the uniformity measure and edge measure for object-background segmentation are used. When an image is segmented, a set of invariant features called homomorphic invariant moments, which is derived from the spectrum histogram of the target image, is calculated. The classification is then accomplished using the membership function of the feature space of an image and stored patterns. By simulation results, we find that the fuzzy 2D entropic thresholding achieves significant performance according to the uniformity and edge measures, and the fuzzy classifier with the invariant features has good performance even in low signal-to-noise ratio conditions.