奇异目标信号的多尺度定位识别方法  

Method of Singularity Target Signal Multiscale Orientation and Recognition

作　　者：许先凡[1] 明军[1] 

机构地区：[1]安徽大学电子学院,安徽合肥230039

出　　处：《计算机技术与发展》2006年第3期63-65,共3页Computer Technology and Development

基　　金：安徽省自然科学基金资助项目(01042105)

摘　　要：在复杂环境下检测微弱目标信号是信号检测技术中关键问题。基于奇异目标信号和噪声干扰在多尺度小波变换下表现出截然不同的特性,利用奇异目标信号的小波变换模极大值随尺度增大而增大,而随机白噪声小波变换模极大值随尺度增大而减小,进行复杂环境中的目标定位识别。理论和实验表明该方法特别适合弱信号的定位识别。The weak signal orientation and recognition in overlapped random noise is the bottleneck problem. Based on completely different properties of the signal and noise in multiseale wavelet transform, the singularity target signal wavelet transform modulus maxima increment by scale enlargement, and the noise＇ s wavelet transform modulus maxima increment by scale reducing, target signal is orientated and recognized in complex circumstance.Theoretical analysis and experimental results show that the method is particularly suitable for weak signal orientation and recognition.

关 键 词：小波变换 奇异性 目标定位 

分 类 号：TN911.23[电子电信—通信与信息系统]