Abstract: The stability of tunnel chambers is predominantly governed by internally developed structural planes. Consequently, rapid, accurate and comprehensive acquisition of structural plane information is essential for engineering stability assessment and construction safety assurance. However, conventional contact-based measurement methods exhibit limitations such as low operational efficiency and high labor intensity, making them unsuitable for demands of efficient construction in long, large-scale tunnels. To address this challenge, this study proposes an intelligent identification and extraction method for tunnel geological structural planes, based on 3D laser scanning point cloud data. Initially, high-precision 3D laser scanning is employed to capture detailed point clouds of the tunnel surface, which are subsequently utilized to reconstruct a geometrically accurate 3D model. Subsequently, an enhanced region-growing algorithm is implemented, incorporating robust point cloud normal vector estimation and convex hull-guided scanning, which enables fully automatic extraction of structural plane geometric features and intelligent parameter interpretation. Field validation at adit No.7 of the Xulong Hydropower Station demonstrated the method′s efficacy in identifying geological structural planes from extensive point cloud datasets. The angular deviation between automated identification results and on-site manual measurements was controlled within 6°, meeting rigorous engineering accuracy requirements. Collectively, this work establishes an end-to-end intelligent workflow encompassing data acquisition, processing, and geological interpretation, thereby providing a robust technical foundation for the digital transformation and intelligent advancement of tunnel rock mass quality evaluation, stability analysis, and safe construction practices.