Abstract: To evaluate the segment assembly quality during tunnel boring, a LiDAR-based method for measuring and calculating segment roundness is proposed. First, a two-dimensional LiDAR is used for circumferential scanning to obtain three-dimensional point cloud data of segments, and wavelet transform is applied to decompose the point cloud signal into low-frequency and high-frequency components to filter out random noise and preserve the characteristics of the main structure. Then, a normal differential algorithm is used to calculate the normal vector of each point, and outliers are identified and eliminated according to normal changes to enhance geometric features. Furthermore, a region-growing algorithm is adopted to expand the region from selected seed points based on point cloud similarity criteria, realizing effective point cloud segmentation and removing residual noise. In the feature extraction stage, principal component analysis is used to calculate the covariance matrix of the point cloud data, extract main features and perform dimensionality reduction to suppress noise and strengthen the representation of shape features. Finally, the processed point cloud is fitted using the Radim ellipse-fitting algorithm to obtain the geometric parameters of the ellipse, and the roundness of tunnel segments is quantitatively evaluated by analyzing the deviation between the fitted ellipse and the theoretically designed circle. The established segment roundness calculation system can collect segment roundness information in real time and efficiently, and significantly improves measurement accuracy while ensuring economic efficiency. The deviation between its calculation results and total station measurements can be controlled within 10 mm.