探地雷达回波相邻道信号具有很大的相似性,本文利用相邻道信号间的相关性,改 进了二次样条小波模极大值搜索算法,提出了一种基于信号相关性的奇异点检测算法,并加以实现。实验结果表明改进后的算法能够进一步去除噪声产生的奇异点,更加准确的判断目标回波的奇异点位置,为地下埋藏物位置的判断提供有利的判据。 关键词:小波变换,样条小波,Lipschitz 指数,奇异性,探地雷达 Abstract: Based on the wavelet module maxima theory of Mallat and analyzing the feature of the GPR’s echo wave signals in detail. Because the GPR echo wave neighboring scan signals are similar, the paper utilizes the signal correlativity and improves the wavelet module maxima algorithm. The result of experiments validates the improved algorithm can wipe off the noise’s disturbance more efficiently and get the exact positions of the singularities. It is important for showing the position of object under the ground. Key words: GPR, spline wavelet, Lipschitz exponent, singularity