Search

The measurement of shiny objects is a major challenge for structured light 3D measurement. Polarization systems can be used to measure shiny objects. However, the polarizer may attenuate the light intensity of the captured image, reducing the SNR of the image, and the measurement accuracy will be affected. To improve the measurement accuracy and efficiency, this paper proposes a 3D measurement method based on polarization optimal projection intensity. By establishing the camera response function under the polarization system, the desired optimal projection intensity can be directly estimated and used to compensate for the image light intensity attenuated by the additional polarizer. This method does not require rotating polarizers and multiple exposure times. In addition, the optimal fringe images are obtained by applying image fusion algorithms. The experimental results show that the method improves pixels in the image that are too dark, while significantly reducing the saturated pixels on the shiny surface. This method can effectively improve the image contrast, get better fringe images, extract more effective information, and reconstruct a more complete 3D point cloud.

Nonlinear and nonconvex optimization problem (NNOP) is a challenging problem in control theory and applications. In this paper, a novel mixture varying-gain dynamic learning network (MVG-DLN) is proposed to solve NNOP with inequality constraints. To do so, first, this NNOP is transformed into some equations through Karush–Kuhn–Tucker (KKT) conditions and projection theorem, and the neuro-dynamics function can be obtained. Second, the time varying convergence parameter is utilized to obtain a faster convergence speed. Third, an integral term is used to strengthen the robustness. Theoretical analysis proves that the proposed MVG-DLN has global convergence and good robustness. Three numerical simulation comparisons between FT-FP-CDNN and MVG-DLN substantiate the faster convergence performance and greater robustness of the MVG-DLN in solving the nonlinear and nonconvex optimization problems.

In complex field of view (FOV) environments, a single camera’s FOV measurement range is limited and cannot cover the entire object under test for global calibration. Multiple cameras are used mostly for large FOV environment measurements, but the traditional one- and two-dimensional targets used for global calibration in large FOV environments are prone to overlapping FOV. Furthermore, other large-sized targets are difficult to produce and process, and the laser projection method and plane mirror calibration methods are easily affected by the outdoor environment. To solve this problem, a non-common FOV binocular calibration method based on rigidly connected stereo targets is proposed. The calibration process is as follows: First, the rigidly connected target, which is composed of two plane targets with a checkerboard, is placed in front of the two cameras, and the vision sensor captures the corresponding sub-target image; then, the target is moved multiple times, and the transformation relationship between multiple vision sensors is obtained according to the spatial constraint characteristics of the rigidly connected target. Hence, the method overcomes the limitation of the non-overlapping FOV calibration method that relies on large measuring instruments. The experimental results show that the RMS error of the 13 mm distance is 0.16 mm. The proposed method is effective, simpler to operate than other methods, and does not rely on the constraint of complex targets. More importantly, this measurement method solves the difficult problem of measurement in non-public FOV, meeting the requirements of large FOV measurement ranges.

A specular suppression model on reflective material figures based on polarization information and figure grayscale information is built. The model can optimize the imaging effect of reflective material in a strong light environment to restore the information characteristics of the figure. Taking the workpiece surface of reflective material as the experimental object, figure surfaced information restoration and visual measurement effect analysis were carried out. The experimental results show that the method offered can achieve good suppression of the large-area light spot phenomenon in the imaging of reflective material. The texture contours of the figures are significantly improved, and the imaging indicators in all aspects are further optimized, which verifies the feasibility of the offered method to a certain extent.

Line-structured light 3D measurement is often used for 3D contour reconstruction of objects in complex industrial environments, where light plane calibration is a key step. In this paper, we propose a calibration method for a line-structured optical system based on a hinge-connected double-checkerboards stereo target. First, the target is moved randomly in multiple positions at any angle within the camera measurement space. Then, by acquiring any one image of the target with line-structured light, the 3D coordinates of the light stripes feature points are solved with the help of the external parameter matrix of the target plane and the camera coordinate system. Finally, the coordinate point cloud is denoised and used to quadratically fit the light plane. Compared with the traditional line-structured measurement system, the proposed method can acquire two calibration images at once; thus, only one image of line-structured light is needed to complete the light plane calibration. There is no strict requirement for the target pinch angle and placement, which improve system calibration speed with high accuracy. The experimental results show that the maximum RMS error of this method is 0.075 mm, and the operation is simpler and more effective to meet the technical requirements of industrial 3D measurement.

Image indexing plays a significant role in approximate K -nearest neighbor image retrieval. In the past decades, the most commonly used methods in image indexing are unsupervised-based. We recognize that unsupervised-based methods cannot benefit from supervised information. In this paper, supervised learning to index method is proposed to realize approximate K -nearest neighbor image retrieval. For designing the supervised learning to index algorithm, we first need to establish the supervision information of an image that should be quantified to the codeword in the codebook. Thus, we propose an image relabeling algorithm to relabel the database images so that the supervised codebook learning can be realized. Second, compared with only distance-based unsupervised image indexing method, a better idea of supervised learning to index is incorporating the distance-based similarities with the classification probabilities of samples at the image indexing stage. In order to achieve this goal, a classifier is also trained based on the relabeled images, and we finally build a constraint optimization objective function to learn the codebook and the classifier simultaneously. The proposed method is verified on the MNIST, the CIFAR-10, and the ImageNet data sets. Experimental results demonstrate the effectiveness of the proposed method.

Stray light from line-structured light vision sensors easily interferes with the center extraction accuracy of a structured light strip. In this paper, a high-accuracy calibration technique using polarized beam splitting simultaneous imaging is proposed. First, two polarization component images with different polarization directions are obtained through a single measurement. The method overcomes problems related to stray light and errors caused by the shaking of the instruments. Therefore, false polarization information can be avoided in this way. Second, an image fusion algorithm is used to improve the imaging quality of the light bar. Experiments are performed to show that the light strips no longer have local overexposure, and the calibration accuracy of the system is improved.

Multicamera systems are commonly applied in the large field-of-view (FOV) measurements. However, two cameras may have a non-overlapping FOV in some scenes, and traditional binocular camera calibration methods cannot be used to directly calibrate non-overlapping cameras. To solve this problem, our study proposes a calibration method for binocular cameras with non-overlapping FOVs on the basis of planar mirrors. According to the reflection characteristics of an optical planar mirror, the proposed method can use the same target to calibrate non-overlapping cameras. Hence, the method overcomes the limitation of non-overlapping cameras being unable to observe common targets. Experiments results show that the maximum RMS error does not exceed 0.53 mm. Hence, the proposed method is effective, and its measurement technique is simpler and more universal than those of other methods. In addition, it is applicable to a wide range of measurements.

Fringe projection technology is often used for three-dimensional (3D) measurement, but it is difficult to measure highlight surfaces. Polarization systems are usually used to remove highlight surfaces. Polarizing filters can be used to eliminate the highlights of the image, but they may also cause the image to be too dark and affect the measurement accuracy. Otherwise, to ensure measurement accuracy, the complexity of the operation of the polarization system will be increased. A method of polarization-based camera intensity response function for 3D measurement is proposed. The intensity response function of the camera under the polarization system is established. It can avoid the complicated polarized bidirectional reflectance distribution function model and directly and quantitatively calculate the required angle between the transmission axes of the two polarizing filters. Then it is combined with the image fusion algorithm to generate the optimal fringe pattern. Experimental results demonstrate that this method significantly eliminates the effects of highlights in the image. The fuzzy transition area between the black and white fringes is effectively reduced, and the edge information of the fringes is correctly restored. Moreover, the high signal-to-noise ratio and contrast of the image are retained when the polarization filters are added.

The accurate estimation of camera internal and external parameters is one of the keys to high-precision camera calibration. At present, the mainstream Zhang's calibration method requires the use of a checkerboard to achieve high-precision calibration result. Therefore, the precise extraction of the pixel coordinates at the corner points on the checkerboard is one of the key technologies. The traditional methods are to improve the corner extraction algorithm, but ignore the imaging process of the calibration images. It is difficult to detect and eliminate the high-light regions in the process of calibrating image, and the existence of high-light regions will cause low accuracy of camera calibration. Therefore, a camera calibration method based on optimal polarization information is proposed in this paper. The method first uses a Stokes variable linear feedback through a CCD camera equipped with a polarizer to obtain a checkerboard calibration image at an optimal polarization angle in different spatial positions. Then use the sub-pixel level detection algorithm and Gaussian fitting method to locate the corner points in the image and to solve the sub-pixel coordinates of the corner points precisely. At last, the two-dimensional pixel coordinates of the corner points in each checkerboard image are extracted to perform camera calibration. The experimental results show that the coordinates of corner points of calibration plate image based on the optimal polarization angle have about 1 pixel difference compared with the coordinates of the calibration plate corner points under natural illumination, and the slopes between the parallel lines of the fitted are closer, meet the high precision calibration requirement of camera.

This paper presents a new, to the best of our knowledge, calibration method for line-structured light vision sensors. The laser stripe intersecting the cylindrical target and the line-structured light is captured by the camera, and the light plane parameters of the line-structured light are obtained by combining the cross-sectional characteristics of the ellipse. The nonlinear optimization of the light plane parameters use multiple locations to get the optimal solution of the light plane equation. This method requires only a single cylindrical target, which in turn greatly simplifies the calibration process. The results of simulation experiments and physical experiments show that the proposed calibration method can achieve higher calibration accuracy and measurement accuracy. The effectiveness of the new calibration method is verified.