Abstract: With the development of tunnel construction, the detection, control, and treatment of all kinds of tunnel lining diseases have received increased attention. Therefore, tunnel nondestructive testing technology is widely used as an intelligent emerging technology. Due to the cost limitations of industrial applications, nondestructive testing technology of medium- and small-shield tunnels experiences serious losses in accuracy and speed when transferred to large-diameter shield tunnels. A large shield tunnel forming quality inspection vehicle is developed by integrating a two-dimensional (2D) laser scanner with a wheel shaft encoder and an industrial computer. Based on this equipment, an asymmetric shield quality inspection method based on digital images is proposed. By analyzing the construction environment of a large-diameter shield tunnel, the ground and vehicle body point clouds are filtered, the central axis is extracted using the neighborhood vector method, and the central coordinate system of the tunnel is established. The inspection route is arranged eccentrically, and the sampling point clouds are divided into dense and sparse side point clouds according to their tunnel density. Moreover, different methods are used to pick up the joint features of the segment. The dense side point cloud is expanded around the central axis into a 2D gray image, and the joint image is segmented by scaling, normalization, and gradient threshold segmentation. According to the classification of joints based on the linear equation, the linear distribution formula of sparse and dense side joints is deduced by combining the segment structure and the point placement, with the sparse side joints picked up indirectly. According to the joint characteristic points, the edge point clusters on both sides of the segment and the segment misalignment are calculated. The least square method is used to fit the tunnel point cloud and calculate tunnel ellipticity. Tests are conducted in a large-diameter shield tunnel to verify the effectiveness of the inspection method, and the following conclusions are made from the test results: The inspection speed of the molding quality inspection vehicle in the 14-meter shield tunnel is 3 km·h⁻¹; compared with traditional detection methods, the average deviation of segment dislocation detection is less than 2 mm; the average deviation of tunnel contour maximum deformation detection is less than 2 mm, and the average deviation of ovality detection is less than 0.1%. This meets the high-speed, high-precision, and low-cost requirements of large-diameter shield tunnel molding quality inspections.