This paper describes a magnetic flux leak (MFL) model of pipeline defect inspection, and presents a recognition algorithm based on dynamic wavelet basis function (WBF) neural network. The dynamic network utilizes multiscale and multiresolution orthogonal wavelet, through signals backwards propagation, has more significant advantages than BP or other neural networks used in MFL inspection. It also can control the accuracy of the predicted defect profiles, high-speed convergence possessing and well approaching feature. The performance applying the algorithm based on the network to predict defect profile from experimental MFL signals is presented.