Abstract:Purpose: The structure of the rock mass determines its strength and load-bearing capacity, which directly impacts the permissible exposure and exposure time in underground engineering. Borehole cameras play a crucial role in obtaining detailed information about the structure of the rock mass. This paper focuses on the challenges associated with automatically extracting structural parameters of the rock mass from camera images, including issues related to accuracy. Method: It analyzes the characteristics of structural surfaces in the images and examines the relationship between the structural surfaces in the three-dimensional model and their representation in planar form. The paper presents a process and method for deriving parameters and reconstructing the three-dimensional space and structural surfaces based on camera image information. Significant progress has been made in three key technologies for extracting joint information from borehole images, including coordinate transformation, fitting structural surfaces, and extracting parameters such as the orientation, dip angle, and trace width of the structural surfaces. The software for automatically extracting structural surface parameters from borehole images was developed using the Python programming language, with PyCharm as the integrated development environment. It utilizes PyQt5 graphical controls, the Numpy numerical computing library, and the SQLite database. The software has been successfully applied to extract structural surface parameters of underground borehole rock masses in a specific mine. Result: The application results demonstrate that the proposed method and developed system offer fast extraction speed and high accuracy. Conclusion: This research provides a valuable approach for rapidly parameterizing structural surfaces in boreholes.