Methods of Characterizing Defects Based on Magnetic Flux Leakage Inspection

吴欣怡,赵伟,黄松岭
DOI: https://doi.org/10.3969/j.issn.1001-1390.2008.05.005
2008-01-01
Abstract:Ways to characterize a ferromagnetic pipeline defect detected by magnetic flux leakage inspection are presented.The paper shows principles,characters and results of these feasible ways.Because artificial neural network(ANN) is often used to characterize a defect,features of ANN used in defect sizing are introduced in detail.Furthermore,directions for improving the quality of quantitative recognition are proposed.
What problem does this paper attempt to address?