Real-time semantic segmentation for underground mine tunnel

Jiawen Wang, Dewei Li, Qihang Long, Zhongqi Zhao, Xuan Gao, Jingchuan Chen, Kehu Yang

Abstract: Highlights•A real-time semantic segmentation algorithm for underground mines is proposed.•A semantic segmentation dataset of underground mines is constructed and utilized.•FAINM aims to enhance the robustness of network in varying lighting conditions.•SRCM is proposed to enhance the performance of network in underground mines.•Extensive comparative experiments have demonstrated the efficacy of FA-DDRNet.