Abstract: The stability of tunnel chambers is primarily influenced by the numerous structural planes that develop within them. Consequently, accurately, efficiently, and comprehensively acquiring information about these structural planes is crucial for conducting a thorough stability analysis of the chambers. Three-dimensional laser scanning technology enables rapid collection of point cloud data from the tunnel surface and facilitates the construction of a detailed point cloud model of the tunnel. Building upon this foundation, an improved region-growing algorithm can be employed to extract geometric features associated with the structural planes. By incorporating methods for calculating point cloud normal vectors and utilizing convex hull scanning techniques, intelligent interpretation of structural plane parameters can be achieved. This methodology was successfully implemented in the No. 7 horizontal tunnel at Xulong Hydropower Station. The recognition results obtained were compared and analyzed against on-site geological measurement outcomes. The findings indicate that this approach effectively enables intelligent recognition of geological structural plane information from extensive point cloud datasets pertaining to tunnels, with an engineering rock mass recognition error maintained within 6°, thereby satisfying measurement accuracy requirements for structural planes. Ultimately, a comprehensive intelligent recognition framework has been established—from rock mass data collection to geological information interpretation—providing valuable theoretical and technical support for advancing digitalization and intelligence in rock mass quality assessment and safety evaluation.