APLCNet: Automatic Pixel-Level Crack Detection Network Based on Instance Segmentation

Abstract

The accurate and automatic detection of pavement cracks is essential for pavement maintenance. However, automatic crack detection remains a challenging problem due to the inconspicuous visual features of cracks in complex pavement backgrounds, the complicated shapes and structures of cracks, and the influences of weather changes and noise. In recent years, with the development of artificial intelligence technology, crack detection methods based on classification and semantic segmentation have laid a good foundation for the automation of pavement crack detection. However, there remain shortcomings in the comprehensive acquisition of pavement crack attribute information and detection accuracy. To solve these problems, this paper proposes an instance segmentation network for pavement crack detection. The network can simultaneously obtain the crack category, position, and mask, and can realize end-to-end pixel-level crack detection. A semantic segmentation branch is first added to Mask R-CNN. This branch can extract the bottom-level detail information of the cracks and ultimately improves the accuracy of crack mask prediction. An adaptive feature fusion module is then designed. During feature fusion, this module highlights the attribute information and location information of cracks according to the channel attention mechanism and the spatial attention mechanism. Finally, these two modules are integrated to form an automatic pixel-level crack detection network, namely APLCNet. Without any embellishment, APLCNet achieves a precision of 92.21%, a recall of 94.89%, and an F1-score of 93.53% on the challenging public CFD dataset, thereby outperforming CrackForest and MFCD for pixel-wise crack detection. Moreover, APLCNet achieves a 16.5% mask AP on the self-captured GDPH dataset, thereby surpassing Mask R-CNN and PANet.

A deep learning‐based method for pixel‐level crack detection on concrete bridges

“…Through the comprehensive evaluation of the two‐stage network performance in this paper, it is proved that each stage of the network has better detection performance. The proposed method is compared with other excellent crack detection models based on depth learning (Deepcrack [26], FPHBN [29], APLCNet [32], ANet‐FSM [35], HDCB‐Net [36]), the performance and output of the different methods are shown in Figures 14 and 15.…”

“…Zhang et al. [32] modify the Mask‐Rcnn [33] network to solve the problem that Mask branches in Mask‐Rcnn cannot predict the details of cracks accurately. In addition to the research on the deep network structure, Li et al.…”

“…Beckman et al [31] separate the crack parts from the point cloud images of depth camera, and quantify the crack volume directly. Zhang et al [32] modify the Mask-Rcnn [33] network to solve the problem that Mask branches in Mask-Rcnn cannot predict the details of cracks accurately. In addition to the research on the deep network structure, Li et al [34] propose a pixel-based adaptive weighted cross-entropy (WCE) loss in conjunction with Jaccard distance to facilitate high-quality pixel-level road crack detection by analyzing the crack images.…”

Anisotropic Diffusion-Based Enhancement of Scene Segmentation with Instance Labels

“…The authors of [17] enforce consistency between semantic and instance segmentation masks prediction through a specialized loss construction. In [30], the authors add a semantic segmentation head in Mask R-CNN to uncover fine details in a crack detection setting. All the aforementioned works combine semantic and instance segmentation in a fusion setting, where it is assumed that the combination of the two tasks will facilitate the overall mask predictions.…”

Automated Ground Truth Generation for Learning-Based Crack Detection on Concrete Surfaces

“…Furthermore, they verified that the U-Net outperformed other existing deep CNNs in terms of robustness, effectiveness, and detection accuracy. Hitherto, both FCNs and U-Nets have been investigated extensively, e.g., the automatic pixel-level crack detection network [26] and the convolutional encoderdecoder network [27].…”