3d reconstruction of laser projective point with projection invariant generated from five points on 2d target - projector

Visual measurement based on structural light is of great significance in optical detection research.
A 2D target fixed with a line laser projector is designed to realize the conversion between the world coordinate system, the camera coordinate system and the image coordinate system.
Laser projection points and five non-
A common line point randomly selected from the target is used to construct the projection invariant.
The closed-form solution of the 3D laser point is solved by the homogeneous linear equation generated by the projection invariant equation.
The optimization function is re-by parametric-
Projection error of laser point and target point in image coordinate system.
In addition, by minimizing the optimization function, the nonlinear optimization solution of the world coordinates, camera parameters and lens distortion coefficient of the projection point has been contributed.
By comparing the displacement of the reconstructed laser point with the actual displacement, the accuracy of 3D reconstruction is evaluated.
In experiments, the effects of image volume, lens distortion and noise are studied, and the results show that this reconstruction method is effective and helps to carry out accurate testing in the measurement system.
In recent years, the 3D reconstruction technology based on camera has been widely concerned by researchers and has been applied in various research fields. E. g.
Mian extracts features from face images from different angles obtained using computer screens as coded lighting and uses a new efficient algorithm to reconstruct 3D contour models.
Kim proposed a linear hierarchical approach, which includes affinity recovery and metric recovery.
Camera Calibration and 3D reconstruction using scene geometry.
Visual measurement based on structural light is a very important method to reconstruct the surface of 3D objects.
Contact test, simple operation, high efficiency, good accuracy.
Nguyen presents a real
The time measurement method of object shape has the advantages of fast speed and high precision.
The multi-threaded parallel algorithm is used to process the images collected by cameras and programmable projectors.
Qin provides a process to detect ground changes at the street level by combining points cloud and ground images.
It corrects and re-checks the consistency of cloud and stereo images
The stereo projection part of the Earth picture.
In order to automatically obtain the information of the object, Dipanda proposed a method of 3D shape reconstruction using structural light.
Real Communication Program
Time is realized by cell algorithm.
Xu describes the solution of plane structure light in vision
Based on the inspection, the plüker matrix is adopted, which shows the accurate calibration of plane structural light under the influence factors of image quantity and noise.
Saeed introduces a method for extracting weld surface information from images.
Charge after calibration
The coupled sensor is used to capture the laser beam of the mirror image to make up the image.
By using the information of the charge, the depth of the surface of the welding pool can be calculated
Position of coupler sensor and laser.
Chen proposed a method to reconstruct the wheel surface using visual sensor data.
The wheel surface is structured from two-light-
The sensor is calibrated through the iteration of the nearest point.
Li outlined a method to improve the contour technique of Fourier transform (FTP).
According to geometric conditions, the ash of the light phase recovery on the map
Scale image of stripes.
Cloud proposed a framework to improve FTP method by retrieving absolute phase pixelsby-pixel.
Two images with different frequencies are used in this method. The high-
Generating frequency phase from low levelFrequency phase.
Accurate and effective camera calibration technology is particularly important for the 3D reconstruction of objects.
The three calibration systems of the camera were recorded in the previous work: 3D calibration system, 2D plane calibration system and 1D linear calibration system.
The 3D calibration system only obtains the parameters of the camera through one captured image.
At the same time, high precision is achieved by 3D calibration target.
However, in order to extract enough feature points, the 3D target needs to be made into a certain volume, and the three sides of the volume must be precisely perpendicular to each other.
Therefore, the process and transportation of the 3D target is time-
Consumption and inconvenience.
Compared with the three-dimensional target, the two-dimensional plane target system has many advantages.
2D target calibration system is easy to manufacture for on-
Calibration on site.
In addition, the 2D plane calibration system provides sufficient calibration information in a convenient way.
1D targets are also easier to produce than 2D targets and 3D targets, however, the measurement accuracy of 1D targets is lower than 2D calibration systems and 3D calibration systems due to lack of information.
Therefore, the two-dimensional plane calibration system is adopted in this study.
A reconstruction method of 3D laser projection point is proposed for the use of line-structured-light.
The 3D reconstruction method of laser projection points is divided into four parts.
First, the laser line of the projector is located on a plane that coincides with the 2D calibration target plane.
Therefore, on the same plane
The common line point can be extracted to construct the projection variant.
Secondly, one laser projection point and five non-
Randomly extract the common line points on the target from the captured image.
Similarly, the same point is calculated in the camera coordinate system.
For the same point, the invariant item in the image coordinate system is equal to the invariant item in the camera coordinate system.
According to the above idea, the homogeneous linear equation is constructed and the solution of the closed form is determined by the decomposition of singular values.
Thirdly, the optimization function is designed to improve the reconstruction accuracy by minimizing the parametric reconstruction accuracy.
Projection error of laser point and target point.
Reconstruct 3D laser points in the world coordinate system.
Finally, by comparing the difference between the reconstructed displacement and the real displacement of the laser point, the effects of image volume, lens distortion and noise were evaluated experimentally.
The following paper is summarized as follows: Section 2, method, describes the reconstruction method consisting of closed formal solutions and optimized solutions.
Section 3, results, provides an experiment to verify the reconstruction method.
Factors, lens distortion, noise and application cases are considered in this section.
Section 4, discussion, evaluation of the reconstruction method based on absolute error and relative error.
The fifth part summarizes and summarizes the paper.