Corrosion-like Defect Severity Estimation in Pipelines Using Convolutional Neural Networks

Evaluation of the nominal thickness of pipelines using neural networks through the dispersion curves drawn with the SAFE method

“…Being different from traditional ANN, CNN can directly capture the spatial features from images to improve both prediction accuracy and efficiency [31]. The CNN has been widely employed in diverse fields such as computer vision including image classification [32], object tracking, visual salience detection, action recognition; natural language processing [33] and time series classification and forecasting [22], [30], [34].…”

“…The fully connected layer classifies the extracted features from the previous layers [22]. It focuses on the transformation of the pooled features into a suitable representation for regression.…”

“…The progress in machine learning techniques allows the reduction of human intervention in the analysis of detected defects [17]- [20]. Among the various techniques that have been developed a Convolutional Neural Network (CNN) technique is a genre of neural networks that have already demonstrated excellent capabilities in tasks such as pattern recognition, computer vision [21], and time series classification [22], proving its ability to deal with complex and nonlinear problems [23], [24]. It is designed to automatically learn and extract features from images, allowing it to perform tasks such as object recognition, segmentation, and classification [25]- [27].…”