Aircraft recognition based on nonparametrical statistics

Abstract

The problem of aircraft recognition in a single image is analyzed. A novel method combined Fourier series representation and nonparametric statistics is introduced to do pattern recognition for aircraft. Silhouettes obtained by image segmentation are described by Fourier series, after which empirical distribution function of curvature is computed. Fast matching is performed with coarse hypothesis test for partial classification. The re#ned hypothesis test is used to get detailed matching result. Experimental results show the effectiveness of the algorithm for shape recognition.