The Real-Time Quantitation and Display Method for Incomplete Defect MFL Signals

Defect real-time quantitation and display is critical for tank floor magnetic flux leakage (MFL) testing. Since the detection process is continuous, which means the data processing area may contain the incomplete MFL signals. This paper proposed a novel real-time quantitation and display method for the condition of incomplete defect MFL signals. This method is mainly for defect edge recognition and depth estimation. Firstly, to identify the defect edge for the incomplete MFL signals, the edge identification method based on the Sobel discrete different operator has been putting forward. Then, to estimate the defect depth, putting forward rapid depth estimation method based on the quantification neural network, key characterise abstraction and signal equivalent processing. Finally, sums up the real-time display method after estimating the edge and depth of the defects. By experimental verification, this method can realize the defects’ quantitative and display for the incomplete defect MFL signals effectively.