Towards Automatic Crack Detection by Deep Learning and Active Thermography

Metal joining processes are crucial in current technological devices. To grant the quality of the weldings is the key to ensure a long life cycle of a component. This work faces crack detection in Electron-Bean Welding (EBW) and Tungsten Inert Gas (TIG) weldings using Inductive Thermography with the aim to substitute traditional Non-Destructive Testing (NDT) inspection techniques. The novel method presented in this work can be divided up into two main phases. The first one corresponds to the thermographic inspection, where the thermographic recordings are reconstructed and processed, whereas the second one deals with cracks detection. Last phase is a Convolutional Neural Network inspired in the well-known VGG model which segments the thermographic information, detecting accurately where the cracks are. The thermographic inspection has been complemented with measurements in an optical microscope, showing a good correlation between the experimental and the prediction of this novel solution.

[1]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[2]  I. Gorosmendi,et al.  A novel Automatic Defect Detection Method for Electron Beam Welded Inconel 718 Components using Inductive Thermography , 2018 .

[3]  Ting Liu,et al.  Recent advances in convolutional neural networks , 2015, Pattern Recognit..

[4]  Beate Oswald-Tranta,et al.  Thermo-Inductive Surface Crack Detection in Metallic Materials , 2006 .

[5]  Junhao Wen,et al.  Fundus Image Classification Using VGG-19 Architecture with PCA and SVD , 2018, Symmetry.

[6]  U. Netzelmann,et al.  Induction thermography: principle, applications and first steps towards standardisation , 2016 .

[7]  Kaushik Roy,et al.  Going Deeper in Spiking Neural Networks: VGG and Residual Architectures , 2018, Front. Neurosci..

[8]  Beata Oswald-Tranta,et al.  Thermoinductive investigations of magnetic materials for surface cracks , 2004 .

[9]  Krzysztof Krawiec,et al.  Segmenting Retinal Blood Vessels With Deep Neural Networks , 2016, IEEE Transactions on Medical Imaging.

[10]  Andrew J. R. Simpson.  Over-Sampling in a Deep Neural Network , 2015, ArXiv.

[11]  Beate Oswald-Tranta,et al.  Localizing surface cracks with inductive thermographical inspection: from measurement to image processing , 2011 .

[12]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[13]  Chang Liu,et al.  Regression with Support Vector Machines and VGG Neural Networks , 2019, AMLTA.

[14]  Mitsunori Matsushita,et al.  Estimating Comic Content from the Book Cover Information Using Fine-Tuned VGG Model for Comic Search , 2019, MMM.

[15]  Vladimír Dekýš,et al.  Nondestructive Testing of Metal Parts by Using Infrared Camera , 2017 .

[16]  José García Rodríguez,et al.  A survey on deep learning techniques for image and video semantic segmentation , 2018, Appl. Soft Comput..