摘要
探地雷达以宽频带记录回波信号, 在记录各种有效信号的同时, 不可避免地记录下各种干扰噪声, 能否有效去除干扰噪声将直接影响到探地雷达对地下目标体的识别能力。探地雷达回波信号属于非平稳信号, FFT只适用于平稳信号的滤波, 而小波变换对非平稳信号却能取得良好的滤波效果。为此, 文中提出了采用小波阈值萎缩进行数据预处理的方法, 以去除雷达回波噪声, 并给出了基于DSP的电路和软件实现方法。
Ground Penetrating Radar(GPR) records echo signals with broad bandwidth. When GPR records valid signals, the noises are recorded inevitably at the same time. Whether or not the noises can be removed effectively will directly influence the recognition ability of GPR to underground objects. GPR echo signals are nonstationary. FFT can only be applicable to stationary signals' filtering, but wavelet transform can get better effects of filtering than FFT does toward nonstationary signals.A method of data preprocess using wavelet threshold shrinkage algorithm is presented to remove GPR echo signals' noises and the realization of the circuit based on DSP and software is given.
出处
《计算机测量与控制》
CSCD
2005年第3期259-261,共3页
Computer Measurement &Control
