Automated pixel-level pavement distress detection based on stereo vision and deep learning

Automated pavement distress detection based on 2D images is facing various challenges. To efficiently complete the crack and pothole segmentation in a practical environment, an automated pixel-level pavement distress detection framework integrating stereo vision and deep learning is developed in this study. Based on the multi-view stereo imaging system, multi-feature pavement image datasets containing color images, depth images and color-depth overlapped images are established, providing a new perspective for deep learning. To alleviate computational burden, a modified U-net deep learning architecture introducing depthwise separable convolution is proposed for crack and pothole segmentation. These methods are tested in asphalt roads with different cir-cumstances. The results show that the 3D pavement image achieves millimeter-level accuracy. The enhanced 3D crack segmentation model outperforms other models in terms of segmentation accuracy and inference speed. After obtaining the high-resolution pothole segmentation map, the automated pothole volume measurement is realized with high accuracy.

Automated summary

An automated pixel-level pavement distress detection framework integrating stereo vision and deep learning is developed in this study, which establishes multi-feature pavement image datasets based on a multi-view stereo imaging system and proposes a modified U-net deep learning architecture introducing depthwise separable convolution for efficient crack and pothole segmentation. The results show that the 3D pavement image achieves millimeter-level accuracy, and the enhanced 3D crack segmentation model outperforms other models in terms of accuracy and speed, enabling high-precision automated pothole volume measurement.