Radar target recognition using a radial basis function neural network

被引：24

作者：

Zhao, Q

Bao, Z

机构：

[1] Research Institute, Xidian University, Xi'an

来源：

NEURAL NETWORKS | 1996年 / 9卷 / 04期

关键词：

target recognition; neural network; radial basis function; range profile; kernel classifier; Bayesian probability;

D O I：

10.1016/0893-6080(96)00088-3

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The problem of radar target recognition based on high resolution range profiles is investigated in this paper, using a radial basis function network (RBFN). After analysing the signatures of the range profile, an effective preprocessing method iv proposed to obtain stable and shift invariant patterns which are classified by RBFN. Then the classification mechanism of RBFN is described, suggesting that the RBFN has better performance than the conventional kernel classifier. It is shown from theoretical analysis and experimental results which were obtained with data acquired in a microwave anechoic chamber that the method proposed in this paper offers promise for target recognition. Copyright (C) 1996 Elsevier Science Ltd

引用

页码：709 / 720

页数：12

共 21 条

[1]

[Anonymous], 1981, HIGH RESOLUTION RADA

[2]

Bourlard H., 1989, Advances in Neural Information Processing Systems, VI, P502

[3] ORTHOGONAL LEAST-SQUARES LEARNING ALGORITHM FOR RADIAL BASIS FUNCTION NETWORKS [J].