Vision Tasks in Structural Images

There has been an increasing trend of applying CNNConvolutional neural networks models in structural damage classification, especially in concrete crack detection and steel corrosion detection. Soukup and Huber-Mörk [1] applied CNNConvolutional neural networks to detect steel surface defects of the railway, which is a binary classification problem. Cha et al. [2] used a deep CNNConvolutional neural networks to detect concrete cracks as a binary classification without calculating the defect features.