基于敲击声音信号的墙体内管道探测方法

Wall pipe detection method based on percussion sound signal

为了有效地探测墙体内的埋设管道,分析了基于敲击声音信号的墙体内管道探测原理,对采集到的声音信号进行了时频域分析,再采用Welch方法估计功率谱,提取功率谱峰值作为特征参数输入到支持向量机(SVM)分类模型进行管道有无的分类,提出了基于敲击声音信号的墙体内管道探测方法。实验证明,所提探测方法探测准确率达到了86.85%,能有效地探测出墙壁中有无埋设管道。

In order to effectively detect the wall inside the buried pipeline,this paper analyzes the percussion sound signal within the walls of pipeline detection principle,to the voice of the collected signals are analyzed in the time-frequency domain,Welch method is used to estimate the power spectrum,to extract the characteristic parameters as the input to the power spectrum peak pipes with and without the classification model of support vector machine(SVM)classification,a method for detecting pipes in walls based on percussive sound signals is proposed.Experimental results show that the detection accuracy of the proposed method reaches 86.85%,and it can effectively detect whether there are buried pipes in the wall.