Defect 3D Profile Reconstruction Using Magnetic Flux Leakage Signals Based on Error Estimation

Magnetic leakage detection as a nondestructive testing technology, is widely used in in-service pipeline. Effective 3-D profile reconstruction of defect is very important for magnetic flux leakage (MFL) detection. In this paper, a novel 3-D profile reconstruction method using random forest (RF) based on error estimation is proposed. RF is developed to estimate reconstructed profile errors by inputting signal errors. Then errors between estimated signals and real signals are applied to update defect profiles. Finally, 3-D reconstructed profiles are obtained. Experimental results show that the proposed method has good precision of defect reconstruction.