Aiming at problem that at present

most magnetic induction detection data processing algorithms used for automatic thickness measurement in pipelines have disadvantages of high false positive rate

low precision and long time consumption

high-precision calibration algorithm for small dirt thickness in the oil and gas pipeline is studied. For static calibration

a comprehensive algorithm based on data preprocessing

improved median filtering and least squares fitting is proposed. Aiming at fluctuation of measurement results caused by factors such as temperature

humidity and component aging in practical applications

a dynamic calibration algorithm is proposed in measurement. Actual test results validate that the proposed method satisfy the requirements of high-precision measurement with measurement precision of ±（ 1 + % 1） μm.