Advances in deep learning methods for pavement surface crack detection and identification with visible light visual images

Cracks inevitably exist widely in buildings, structures, parts or products. Compared to contact detection methods such as nondestructive test (NDT) and health monitoring, surface crack detection or identification with visible light visual images is a kind of non-contact method, which is not limited by the material of the tested object and is easy to achieve online real-time fully automation, thus, has the advantages of fast detection speed, low cost and high precision. Firstly, typical pavement (concrete also) crack public data sets for classification, location or segmentation were collected, and the characteristics of sample images and the random variable factors, including environmental, noise and interference, were summarized. Subsequently, the advantages and shortcomings of the three main crack identification methods, i.e., Hand-crafted Feature Engineering, Machine Learning, Deep Learning, were compared. Finally, from the aspects of model architecture, testing performance and predicting effectiveness, the development and progress of typical deep learning models, namely self-built CNN, transfer learning (TL) and encoder-decoder (ED), which can be easily deployed on embedded platform, were reviewed. Meanwhile, from this, we can see the evolution of CNN model architecture, as well as the obvious improvement of performance and effect because of computing power enhancement and algorithm optimization. The benchmark test shows that: 1) It has been able to realize real-time pixel-level crack identification on embedded platform: the entire crack detection average time cost of an image sample is less than 100ms, either using the ED method (i.e., FPCNet) or the TL method based on InceptionV3. It can be reduced to less than 10ms with TL method based on MobileNet (a lightweight backbone base network). 2) In terms of accuracy, it can reach over 99.8% on CCIC which is easily identified by human eyes. On SDNET2018, some samples of which are difficult to be identified, FPCNet can reach 97.5%, while TL method is close to 96.1%. To the best of our knowledge, this paper for the first time comprehensively summarizes the pavement crack public data sets, and the performance and effectiveness of surface crack detection and identification deep learning methods for embedded platform, are reviewed and evaluated.