引用本文: | 张恂,郭桂蓉,庄钊文.基于多分辨分析的雷达目标识别方法[J].国防科技大学学报,1997,19(2):59-63.[点击复制] |
Zhang Xun,Guo Guirong,Zhuang Zhaowen.Radar Target Recognition Based on Multiresolution Analysis[J].Journal of National University of Defense Technology,1997,19(2):59-63[点击复制] |
|
|
|
本文已被:浏览 5997次 下载 5344次 |
基于多分辨分析的雷达目标识别方法 |
张恂, 郭桂蓉, 庄钊文 |
(国防科技大学 自动目标识别国家实验室 湖南 长沙 410073)
|
摘要: |
针对宽带高距离分辨率雷达的工作体制, 提出了一种基于多分辨分析和信息综合的目标识别算法。目标特征由小波变换在相邻分辨率上的能量之比的对数构成。利用多分辨分析, 将目标特征分解为反映目标结构概貌的低通特征和刻画目标结构细节的高通特征。利用辐射基函数神经网络分别对目标的低通特征和高通特征进行识别判决, 然后将基于目标的低通特征和高通特征的判决信息进行综合, 得到最终的识别结果。 |
关键词: 目标识别, 小波变换, 神经网络 |
DOI: |
投稿日期:1996-05-27 |
基金项目:国防预研基金资助项目 |
|
Radar Target Recognition Based on Multiresolution Analysis |
Zhang Xun, Guo Guirong, Zhuang Zhaowen |
(ATR National Laboratory, NUDT, Changsha, 410073)
|
Abstract: |
A target recognition approach based on multiresolution analysis is proposed, and tested on returns of high range resolution radar. Features of targets are extracted by computing the logarithms of the ratios of energies in adjacent scales defined by wavelet transform. With multiresolution analysis, the features are decomposed into the highpass features reflecting detailed structure of targets and the lowpass features showing the outline of the targets. A radial basis function neural network is applied for classification. |
Keywords: target recognition, wavelet transform, neural network |
|
|
|
|
|