Intelligent defect recognition from magnetic flux leakage inspection

THIS paper presents a new approach for detection of oil-gas pipeline corrosion defects in which pipeline magnetic flux leakage (MFL) detection system is adopted. To test the new approach, some artificial defects are simulated mathematically on the surface of oil-gas pipes to simulate different defects happening in practice. The interference eliminating method is introduced for compensation of defected MFL field and helping map extraction from the background. Artificial neural networks, Multilayer perceptron (MLP) are applied for identifying erosion defects. and are described and employed in the paper. Salient results that are mentioned for this devised algorithm, is presented well in this paper that could guide to achieve fine, accurate and fast decision.