The radial basis function (RBF) network can be used as an effective method of function approximation with any precision and a better sorting tool in pattern recognition. On the other hand, wavelet packet transform (WPT) can be used as a new efficient tool to precisely analyze signal properties. In this paper, a conception and algorithm of wavelet packet energy feature (WPEF) is quoted firstly. Then WPEF is applied to features extraction of five types of typical ground target echoes, whose computation results show that the separable characteristic of WPEF of ground target is much better than Fourier transform feature. Finally, combined with RBF network and WPEF, an algorithm for ground target recognition is proposed. Simulation results show that five kinds of typical ground targets are completely identified, which indicates that the proposed approach of ground target identification has better feasibility and stronger anti-noise-interference ability.
展开▼
