Abstract
In the process of X-ray weld defect detection, a deep learning network structure based on the principle of simulated visual perception is constructed. The size and number of layers of convolutional neural network template, and the influence of different activation functions, are analyzed with an improved method proposed, which may avoid the characteristic steps for the extraction of defect images, and can be used to directly determine the presence of any defect. Experiments on 200 images show that the proposed method for SDR images has an identification rate of more than 98%, which is better than other methods, and has highly practically in pipeline defect detection.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.