基于改进小波降噪的雷达一维距离像识别  

Radar One-Dimensional Range Profile Recognition Based on Improved Wavelet Denoising

作　　者：郭小康 简涛 董云龙 GUO Xiaokang;JIAN Tao;DONG Yunlong(Research Institute of Information Fusion,Naval Aviation University,Yantai 264001,China)

机构地区：[1]海军航空大学信息融合研究所

出　　处：《雷达科学与技术》2019年第4期360-364,370,共6页Radar Science and Technology

基　　金：国家自然科学基金(No.61471379,61790551,61102166);国防科技项目基金(No.2102028);装备发展部“十三五”预研项目(No.41413060101);泰山学者工程专项经费资助项目

摘　　要：针对低信噪比下雷达目标一维距离像质量不高、影响目标识别率的问题,将小波阈值降噪的方法应用到雷达目标一维距离像识别研究中,设计了一种新的小波阈值函数,提出了基于小波阈值降噪的雷达一维距离像识别的方法。利用仿真数据进行实验验证,以LVQ(Learning Vector Quantization)神经网络作为分类识别器,进行目标的分类识别研究。结果表明,将小波阈值降噪用于雷达目标一维距离像识别,在低信噪比时能够有效地降低噪声,提高距离像的质量,从而提高目标一维距离像识别率,同时实验也验证了所提出的新阈值函数相较于原阈值函数能更加有效地降低噪声,提高识别率。For the problem that low one-dimensional range profile quality of radar target has affect on the target recognition rate at low SNR,a new threshold function is designed and a method of radar one-dimensional range profile recognition based on wavelet threshold denoising is proposed.The simulation data is used to make experimental verification.The LVQ neural network is used as the classification recognizer to carry out the target classification and recognition.The results show that the wavelet threshold denoising can effectively reduce the noise and improve the range profile quality at low SNR so as to improve the target one-dimensional range profile recognition rate.The experiment also verifies that the proposed new threshold function can reduce the noise and improve the recognition rate more effectively than the original threshold function.

关 键 词：小波 阈值函数 一维距离像 目标识别 LVQ神经网络 

分 类 号：TN957[电子电信—信号与信息处理]