3D InspectionNet: a deep 3D convolutional neural networks based approach for 3D defect detection on concrete columns

Deep learning-based defect feature recognition from 2D image datasets, has recently been a very active research area and deep Convolutional Neural Networks have brought breakthroughs toward object detection and recognition. Due to CNN’s outstanding performance, several recent studies applied it for defect detection in either routine or post-earthquake infrastructure inspections and have reported competitive performance and potential toward automating infrastructure safety assessment. Despite their benefits, the majority of 2D approaches do not leverage or provide 3D depth information directly from the content of the images and as a result do not enable the 3D measurement of severity of these defects. With the increased popularity of 3D scanning and reconstruction technologies, there is pressing need for defect recognition models that operate on 3D data. In this paper, a novel framework using Deep 3D Convolutional Neural networks (3DCNNs) termed 3D InspectionNet is introduced to learn 3D defects features from an artificially generated 3D dataset, intended to mimic defects on the surface of concrete columns such as either cracks or spalls. InspectionNet has the capability of learning the distribution of complex defect features from a large 3D dataset, and distinguishing defects features. For training 3D InspectionNet, a large simulated 3D defect dataset of 3D CAD models was automatically constructed with labeled defect features. The proposed framework can distinguish defect features from the geometric data such as voxels with a high accuracy. The results of this preliminary work demonstrate and emphasize the feasibility and potentials of this approach for 3D defect detection in automated inspection applications.