Drone-Based Wall Crack Detection Using Model-Agnostic Meta-Learning

With the urbanization process and aging of buildings, wall crack detection plays a crucial role in the maintenance and safety of building structures. Due to the inherent characteristics of defects, however, cracks in the wall are relatively sparse, compared with the normal area. High-altitude regions are also difficult to access. As a result, large wall crack data sets are rare. This issue impairs the training of deep networks and may lead to the suboptimal detection result. To address this issue, we first capture a set of pure wall crack images using a drone, which are comprised of a new wall crack data set, namely, Ocean University of China Wall Crack Data Set (OUC-Crack). In contrast to existing crack data sets, OUC-Crack only contains wall crack images and hence it is particularly useful for wall crack detection. Then we propose a drone-based wall crack detection system, which consists of a drone platform and a crack detection network referred to as the Model-Agnostic Meta-Learning Based Segmentation Network (MAML-SegNet). Therefore, only a small number of training images are required. To fulfill the real-time detection task on drones, we further develop an efficient MAML-SegNet (EFF-MAML-SegNet). The proposed systems are tested on the OUC-Crack and two publicly available crack data sets, including Volker and Crack500. Experimental results show that our MAML-SegNet outperforms 14 baselines in terms of Dice Coefficient, with improvements of at least 1.26%, 13.19% and 7.12% on the three data sets, respectively. Given that the EFF-MAML-SegNet is deployed on a drone, real-time detection can be achieved with the Dice Coefficient values of 66.83%, 55.42% and 48.03% on the three data sets, respectively. These promising results should be due to the few-shot learning ability of meta-learning and the efficient network design. (Code, data and models are available at https://indtlab.github.io/projects/Wall-Crack-Detection) Note to Practitioners—The proposed system leverages the MAML that we deliberately design to perform the wall crack detection task using only a small number of training images. This system addresses the issue of insufficient training data in the scenario of wall crack detection. As a result, the challenge of data scarcity that practitioners in the community of wall crack detection have to encounter can be alleviated. In addition, the MAML-SegNet is built using a lightweight network design. This network enables the accurate crack detection with the less computational burden and a small number of training images, which increases the practical applications of the system. Furthermore, the deployment of the EFF-MAML-SegNet on the drone is able to perform the real-time wall crack detection task, due to the efficient design of the network. Compared with manual high-altitude inspection operations, the proposed system reduces both the cost and the risk while greatly improving the work efficiency. This system potentially facilitates the early detection and treatment of wall cracks, enabling the formulation of the effective maintenance and repair strategies. Moreover, the OUC-Crack data can be used by the researchers and engineers in the community for further research.