In the detection of overhead pipelines in petrochemical enterprises, ultrasonic guided wave detection technology has become an important technical means and development direction for pressure pipeline in-service detection and evaluation. In the ultrasonic guided wave inspection of overhead pipelines, it is very important to identify defect signals and suppress noise signals. In this paper, wavelet transform (WT) and empirical mode decomposition (EMD) are used for signal processing, time-frequency transition, signal decomposition, superposition, etc., to reduce noise and extract important signals successfully. Through experiments, it is found that the decomposed signal in WT can better preserve defect information and reduce the interference of noise signals. But the signal processed by the EMD is better than WT. Finally, in the ultrasonic guided wave signal processing, the defect information in the raw signal is retained, so as to reduce the interference of the noise signal and assist inspector in the identification of the defect in the raw signal.