A laser stripe sensor has limited application when a point cloud of geometric samples on the surface of the object needs to be collected, so a galvanometric laser scanner is designed by using a one-mirror galvanometer element as its mechanical device to drive the laser stripe to sweep along the object. A novel mathematical model is derived for the proposed galvanometer laser scanner without any position assumptions and then a model-driven calibration procedure is proposed. Compared with available model-driven approaches, the influence of machining and assembly errors is considered in the proposed model. Meanwhile, a plane-constraint-based approach is proposed to extract a large number of calibration points effectively and accurately to calibrate the galvanometric laser scanner. Repeatability and accuracy of the galvanometric laser scanner are evaluated on the automobile production line to verify the efficiency and accuracy of the proposed calibration method. Experimental results show that the proposed calibration approach yields similar measurement performance compared with a look-up table calibration method.

Large-scale surfaces are prevalent in advanced manufacturing industries, and 3D profilometry of these surfaces plays a pivotal role for quality control. This paper proposes a novel and flexible large-scale 3D scanning system assembled by combining a robot, a binocular structured light scanner and a laser tracker. The measurement principle and system construction of the integrated system are introduced. A mathematical model is established for the global data fusion. Subsequently, a robust method is introduced for the establishment of the end coordinate system. As for hand-eye calibration, the calibration ball is observed by the scanner and the laser tracker simultaneously. With this data, the hand-eye relationship is solved, and then an algorithm is built to get the transformation matrix between the end coordinate system and the world coordinate system. A validation experiment is designed to verify the proposed algorithms. Firstly, a hand-eye calibration experiment is implemented and the computation of the transformation matrix is done. Then a car body rear is measured 22 times in order to verify the global data fusion algorithm. The 3D shape of the rear is reconstructed successfully. To evaluate the precision of the proposed method, a metric tool is built and the results are presented.

Defect detection for specular surfaces plays a vital role in precision manufacturing. However, traditional defect detection methods are unsuitable for specular surfaces because of their specular reflection property. The defect detection on specular surfaces is usually performed by inspectors, which makes the defect detection a time-consuming and unstable task. Deflectometry has been widely used in defect detection for specular surfaces combined with machine learning. Nevertheless, conventional deflectometry methods use the local curvature deviation map based on the unwrapped phase, which can only detect geometrical defects. Moreover, hand-crafted features need to be defined for each specific task. We present a method based on deflectometry and deep learning. Deflectometry provides the input images for the network, and the deep learning network completes the identification and location of defects. In deflectometry, the proposed method uses the light intensity contrast map to replace the local curvature map, which can detect both geometrical and textural defects. Based on conventional networks, depthwise separable convolution kernel is applied to reduce parameters, and residual convolution block is utilized to alleviate vanishing or exploding gradients. A subnet for feature aggregation is used to obtain multiscale information of defect features. Performance evaluation based on experiment results proved the effectiveness of the proposed method.

This paper presents a novel camera calibration method using a planar target with pure translation and a known translation distance. It only requires the straightness and position accuracy of the one-axis translational platform. There are theoretically no constraints among the normal direction of the planar target, the optical axis direction of the camera, and the moving direction of the translational platform. The paper analyzes the closed-form solution, followed by a nonlinear refinement based on the maximum likelihood criterion. Both computer simulation and real data are implemented to verify the effectiveness of the proposed method. Compared with Zhang’s method, the proposed method realizes the camera calibration process automatically and evaluates the calibration process via the measurement accuracy of the calibrated camera, so it is a key factor to advance 3D computer vision one more step from expert to novice use.

3-D local feature extraction and matching is a key step in point cloud registration. However, this process commonly draws into false correspondences caused by noise, occlusion, incomplete surface, and so on. To estimate correct transformation based on these corrupted correspondences, numerous transformation estimation techniques have been proposed. However, no comprehensive study comparing their accuracy, robustness, and efficiency performance under different nuisances has been conducted. This article evaluates thirteen popular transformation estimation proposals on both descriptor-based and synthetic correspondences. On descriptor-based correspondences, comprehensive evaluation items (e.g., combining with iterative closest point (ICP) and different local features) of these methods are tested on five popular datasets acquired with different devices (e.g., Minolta vivid scanner, Microsoft Kinect, and Space Time Stereo). On synthetic correspondences, the robustness of these methods to varying percentages of correct correspondences (PCCs) is evaluated. In addition, their efficiency is also evaluated. The results present some valuable findings that may provide a supplement to existing evaluations of transformation estimation techniques. A summary of merits, demerits, and application guidance of these tested methods is finally presented to guide real-world applications and new transformation estimation techniques crafting.

Point cloud registration plays an important role in three-dimensional (3-D) computer vision. Local feature-based registration as a kind of effective and robust method has two critical steps: descriptor generation and transformation estimation. This paper proposes a novel local feature descriptor termed Local Normal Deviation Statistic Histogram (LNDSH) and an accurate transformation estimation method named 2-point based SAmple Consensus with Compatibility Ranking (2SAC-CR). Our LNDSH is generated on a local reference axis (LRA), and fully encodes geometric and spatial information by six attributes in which a new attribute named Mean Normal Deviation Value ( mndv ) is proposed. mndv encodes mean normal deviation at each point, which is not influenced by the error of an LRA. In 2SAC-CR, an effective compatibility ranking is firstly conducted to increase the possibility of capturing correct correspondences. Then an LRA-based hypothesis generation and a novel hypothesis verification strategy are alternately implemented based on the ranking to ensure an accurate and efficient “hypothesis generation and verification”. Finally, the maximum consensus is used to generate the output transformation, further reducing the error of the result. Extensive experiments conducted on six standard datasets verify that LNDSH has high descriptiveness and strong robustness, and 2SAC-CR possesses high accuracy and strong robustness. Rigorous comparisons with the state-of-the-arts show the overall superiority of our methods.