Abstract: The analysis and identification of defect signals was a key part of magnetic flux leakage inspection technology. In order to quickly identify pipe defect signals and improve the accuracy and efficiency of data analysis, a signal processing algorithm based on low-pass filtering and differential was developed. The noise reduction and optimization processing were carried out on the detection signals of each channel of the magnetic flux leakage detection probe, and pipe defect recognition rules were formulated. Engineering test results showed that the detection signals were clear and easy to identify quickly, the excavation detection results were consistent with the signal identification results, and the signal processing method helped to identify the pipe defect signal quickly and accurately.