RTCNet: A novel real-time triple branch network for pavement crack semantic segmentation

Bin Liu, Jian Kang, Haiyan Guan, Xiaodong Zhi, Yongtao Yu, Lingfei Ma, Daifeng Peng, Linlin Xu, Dongchuan Wang

Abstract: Highlights•A novel real-time triple-branch network was proposed for pavement crack detection under complex scenarios.•A transformer module for enlarging the global receptive field while preserving local details of cracks.•A boundary refinement module was designed to refine crack boundaries.•A manually annotated highway pavement crack dataset with complex scenarios was built for pavement crack detection.