Research on the defects identify of the weld-line's image of the in-service pipeline based on BP neural network

Abstract

This paper applied computer aided technique to do the image recognition work of the welding-line's defects. And this paper uses aspect ratio, roundness, compactness, symmetry, steepness, gray contrast of defect and the background, position of the defect as the defects' eigenvalues. And this paper applied BP neural network to recognize the defects. And experiments are used to get the best values of the eigenvalues.