Abstract: To accurately determine the surrounding rock grade of a tunnel, it is essential to conduct real-time, rapid, and objective evaluations of the tunnel excavation face and perform proactive rock mass risk assessments. Based on the Jianhe-Liping Expressway project in Guizhou, a rapid classification system for surrounding rock is established using basic rock mass quality indicators. During construction, image recognition, target detection, and image threshold segmentation technologies are used to quickly capture the joint information and weathering degree of the tunnel excavation face. Combined with seismic wave velocity and waveform diagrams from advanced geological prediction,surrounding rock parameters, rock integrity, and the development of joints and fractures are obtained. A Mamdani fuzzy inference system (FIS) is introduced, with both qualitative descriptions and quantitative parameters of surrounding rock as inputs to the evaluation information. This system is used to build a rapid real-time dynamic classification model for surrounding rock during the construction phase. The study shows that the model can integrate realtime tunnel excavation face images and advanced geological prediction data to monitor the surrounding rock condition at the tunnel excavation face in real-time, quickly responding to geological changes. The model classification results can provide a basis for adjusting construction strategies in a timely manner.