A K-Dimensional Tree–Iterative Closest Point Algorithm for Overbreak and Underbreak Assessment of Mountain Tunnels

With the increasing scale of mountain tunnel construction, the control of tunnelling quality is becoming a major concern. The efficient and accurate assessment of overbreak and underbreak is vital to the evaluation and optimization of tunnelling quality, but remains a challenge. Thus, this paper proposes an assessment method for overbreak and underbreak based on the K-dimensional (KD) tree and Iterative Closest Point (ICP) algorithm. Firstly, point clouds are acquired using laser scanning during tunnelling and 3D modeling is performed. Secondly, the as-designed 3D models are converted into point clouds and registered with the acquired as-built point clouds using the improved ICP algorithm with KD tree searching. Thirdly, through registration, the deviation between the as-designed and as-built point clouds is calculated, providing an assessment of overbreak and underbreak during tunnelling. Finally, the effectiveness of the proposed algorithm is validated by data from an ultra-long mountain tunnel. Compared with other methods, the merits of the proposed method include the following: (a) detailed and comprehensive data can be acquired efficiently and (b) a promising assessment accuracy (over 90%) can be obtained.