Abstract: Magnetic flux leakage detection technology is one of the most commonly used methods for damage detection in oil and gas pipelines. It is a nondestructive testing technology that uses magnetic sensors to detect magnetic flux leakage signals and then detect pipeline defects. This paper first briefly introduced the principle and operation process of magnetic flux leakage testing, and then summarized the important achievements of magnetic flux leakage testing technology in signal preprocessing, anomaly recognition, and defect quantification in recent years. Among them, anomaly recognition and defect quantification were the key points in the magnetic flux leakage detection process. The focus was on introducing deep learning based anomaly recognition and defect quantification methods, and then analyzing the current shortcomings of these methods. Finally, the future development of pipeline magnetic flux leakage testing was prospected.