The laser triangulation method is one of the most advanced methods for large inner diameter measurement. Our research group proposed a kind of inner diameter measuring device that is principally composed of three laser displacement sensors known to be fixed in the same plane measurement position. It is necessary to calibrate the direction of the laser beams that are emitted by laser displacement sensors because they do not meet the theoretical model accurately. For the purpose of calibrating the direction of laser beams, a calibration method and mathematical model were proposed. The inner diameter measuring device is equipped with the spindle of the machine tool. The laser beams rotate and translate in the plane and constitute the rotary rays which are driven to scan the inner surface of the ring gauge. The direction calibration of the laser beams can be completed by the sensors’ distance information and corresponding data processing method. The corresponding error sources are analyzed and the validity of the method is verified. After the calibration, the measurement error of the inner diameter measuring device reduced from $\pm 25\mathsf{\mu }\mathrm{m}$ to $\pm 15\mathsf{\mu }\mathrm{m}$ and the relative error was not more than 0.011%.

With the rapid development of high-speed image sensors and optical imaging technology, these have effectively promoted the improvement of non-contact 3D shape measurement. Among them, striped structured-light technology has been widely used because of its high measurement accuracy. Compared with classical methods such as Fourier transform profilometry, many deep neural networks are utilized to restore 3D shape from single-shot structured light. In actual engineering deployments, the number of learnable parameters of convolution neural network (CNN) is huge, especially for high-resolution structured-light patterns. To this end, we proposed a dual-path hybrid network based on UNet, which eliminates the deepest convolution layers to reduce the number of learnable parameters, and a swin transformer path is additionally built on the decoder to improve the global perception of this network. The experimental results show that the learnable parameters of the model are reduced by 60% compared with the UNet, and the measurement accuracy is not degraded at the same time. The proposed dual-path hybrid network provides an effective solution for structured-light 3D reconstruction and its practice in engineering.

Structured light profilometry (SLP) is now widely utilized in noncontact three-dimensional (3D) reconstruction due to its convenience in dynamic measurements. Compared with classic fringe projection profilometries, multiple deep neural networks are proposed to demodulate or unwrap the fringe phase, and these networks utilize convolution layers to extract local features while omitting global characteristics. In this paper, we propose SwinConvUNet, a deep neural network for single-shot SLP that can extract local and global features simultaneously. In the network structure design, convolution layers are applied in shallow layers to extract local features, whereas transformer layers extract global features in deep layers, and an improved loss function by combining gradient-based structural similarity is employed to improve reconstruction details. The experimental results demonstrate that SwinConvUNet is more effective than the U-net model at decreasing learnable parameters while maintaining 3D reconstruction accuracy.

A new method for fast diameter measurement of coaxial holes is studied. The paper describes a multi-layer measuring rod that installs a single laser displacement sensor (LDS) on each layer. This method is easy to implement by rotating the measuring rod, and immune from detecting the measuring rod’s rotation angles, so all diameters of coaxial holes can be calculated by sensors’ values. While revolving, the changing angles of each sensor’s laser beams are approximately equal in the rod’s radial direction so that the over-determined nonlinear equations of multi-layer holes for fitting circles can be established. The mathematical model of the measuring rod is established, all parameters that affect the accuracy of measurement are analyzed and simulated. In the experiment, the validity of the method is verified, the inner diameter measuring precision of 28 μm is achieved by 20 μm linearity LDS. The measuring rod has advantages of convenient operation and easy manufacture, according to the actual diameters of coaxial holes, and also the varying number of holes, LDS’s mounting location can be adjusted for different parts. It is convenient for rapid diameter measurement in industrial use.