FEATURE INFLUENCE ON TARGET CLASSIFICATION PERFORMANCE USING THE HIGH RANGE RESOLUTION PROFILES

Document Type : Original Article

Authors

1 Microwave and Radar Laboratory, EMP, BP 17, Bordj El Bahri, Algiers,Tel: 00 213 20 31 43 89 Fax: 00 213 21 86 32 04.

2 Microwave and Radar Laboratory, EMP, BP 17, Bordj El Bahri, Algiers, Tel: 00 213 20 31 43 89 Fax: 00 213 21 86 32 04.

Abstract

We propose a supervised classification of aircrafts using the high range resolution profiles extracted from inverse synthetic aperture radar images. The principal components are extracted from pre-processed synthetic radar data before applying the 1-nearest neighbour classification. The classification performances are presented in terms of the correct classification probability. A classification performances comparison is presented for two types of data: The HRR profiles and the principal components of HRR profiles.

Keywords