Tunnel Construction Face Geological Recognition Dataset

In tunnel construction, accurately recognizing the geological characteristics of the tunnel face is an important challenge to ensure construction safety and efficiency. Currently, traditional geological exploration techniques are costly and time-consuming, and they fall short in terms of recognition accuracy and operational continuity. This dataset aims to address the need for automated, real-time recognition results in existing tunnel geological recognition processes. The dataset collection process involves using drones equipped with professional photography equipment to capture at tunnel construction sites, combined with fixed cameras to collect images of geological changes at different times. The data undergoes multiple rounds of annotation and expert consistency checks to ensure annotation accuracy, managed by a team with a background in geology and engineering. Preprocessing includes denoising, enhancement, and image normalization. Data is stored in high-resolution JPG format and organized according to geological characteristic labels.