One Touch, Friend Forever
contact sales at SINOWON.
Latest products
optimization reconstruction method of object profile using flexible laser plane and bi-planar references
by：Sinowon
2020-02-02
Optimization method for reconstruction of object profile using flexible laser plane and double plane
Flat reference. The bi-
Plane reference is considered to be a flexible benchmark for realizing the conversion between two-world coordinate systems
Plane reference, camera coordinate system, and image coordinate system.
The laser plane is confirmed by the intersection between bi-
Plane reference and laser plane.
The 3D camera coordinates of the intersection of the laser plane with the measured object are initially reconstructed by the image coordinates of the intersection, the inherent parameter matrix, and the laser plane.
At the same time, using the parametric difference of the reconstruction distance of the target with eight marks and the parametric re-projection error of the feature points on the double mark, an optimization function is designed.
Flat reference.
Reconstruction method based on dual domain
The plane reference is evaluated by comparing the difference between the real distance and the standard distance.
The reconstruction error of the initial method is 1. 01u2009mm.
Moreover, the average reconstruction error of the optimization method is 0. 93u2009mm.
Therefore, the double optimization method
The planar references have great application prospects in contour reconstruction.
Visual measurement including structural light is an effective non-
Method of contact 3D measurement.
Since the laser projector provides stronger and narrower lighting than the ordinary digital projector, the laser projector is considered a flexible symbol for projecting the laser plane onto the object to be measured.
The laser curve between the object being measured and the laser plane contains information about the intersection position and the depth of the object being measured.
Therefore, the measurement system based on structured light has been favored by many researchers.
Visual measurement based on structural light has been widely studied due to its wide measurement range, reasonable test speed and high precision in visual measurement methods.
Huynh proposed a calibration method for structured light system based on projector.
The cross ratio in the contribution image of the three collineal world points is the same as the cross ratio calculated by the world points.
According to the projection from the structural light to the image, the recovery matrix of the stripe plane is provided.
Wei also reported a bi-crossover ratio approach to achieve structured-light-
Stripe vision sensor.
Wei uses vanishing points and vanishing lines derived from targets with parallel lines to improve laser plane calibration.
Li has designed a flexible laser scanning system including industrial robot arm and laser scanner.
The rotation and peaceful movement of the robot arm are considered in the scanning model.
Niola describes the calibration of laser scanners in robot applications.
Use a target of known motion to simulate the geometry of the laser transmitter and camera.
Through the fusion of structured lighting and data, the reconstruction of sharp edges of three dimensions is realized, and the automatic detection and reconstruction are realized.
In addition, an algorithm to reconstruct the contour of 3D objects using sharp edges is introduced.
Ma proposes a 3D contour reconstruction method based on Nonlinear Iterative optimization to reduce the error caused by lens distortion.
According to the shape of the projected light, the visual measurement method based on the structure light can be divided into four parts: point structure light, line structure light, grating structure light and coded structure light.
The point structured light method can obtain the 3D data of a point.
However, the speed is slow for large object measurements.
The method of grating or coding structured light is usually based on digital light processing (DLP)
Projector and camera.
Villa recommended reference-
A plane-based approach to reconstruct 3D data on an object.
Depth is achieved by moving a tablet on a linear stage.
Then, the cross grating pattern projected along and in the direction on the object under test.
And coordinates are determined by the stripe phase generated by the Fourier transform and the inverse Fourier transform.
Zhang proposed a method of using a digital projector as a camera.
Digital Micromirror device (DMD)
Like a map between the Republic of Uzbekistan (
Charge coupler
CCD pixels and dvr pixels.
The inherent parameter matrix of the digital projector is then calibrated as the matrix of the camera.
An external parameter matrix is derived from Zhang\'s method.
Finally, 3D coordinates are reconstructed by image coordinates and external parameter matrix.
Hu introduced a method to calibrate the projector. camera system.
The absolute phase diagram consists of three-step algorithm.
Then, a flat plate was designed on the linear stage and driven by a stepping motor.
By moving the plate and projected images, four unknown parameters were calibrated for the measurement system.
Finally, the system parameters are optimized by the coordinate measuring instrument (CMM)
, Plate with holes and iterative algorithm. The digital-projector-
The method has the advantages of high efficiency and rich data.
However, since the pattern of the digital projector is generated by the bulb, the coded light pattern of the digital projector is often affected by lighting from the environment (such as sunlight.
After calibration, the relative position between the digital projector and the camera is fixed.
The flat Laser Projector helps to narrow downband and high-
Density structured light, by avoiding environmental interference with lighting, it is appropriate to measure complex surfaces at medium test speeds.
Camera Calibration is the basis of the accuracy of the visual measurement system.
3D recovery depends on one or more
View the images taken by the camera.
Therefore, improving the accuracy of 3D reconstruction has also attracted the attention of researchers.
At present, the calibration methods are mainly divided into 1D calibration, 2D calibration and 3D calibration.
1D calibration reference is simple in structure and easy to manufacture, but the calibration accuracy is the lowest due to the lack of information on 1D reference.
Although the structure of the 2D calibration benchmark is a little more complicated than the structure of the 1D benchmark, it has the advantage of high calibration accuracy, which is easy to make and move to different places.
The 3D calibration reference provides the highest accuracy for camera calibration.
Nevertheless, the 3D calibration reference is complex and suitable for specific situations.
Although 2D reference can be used to calibrate the projected laser plane, many images of the plane reference at different positions should be captured in order to calibrate only one laser plane.
When the relative position between the camera and the projector is fixed in the previous work, a target can provide a sufficient number of features to solve the intersection.
We are concerned with another case where the intersection point is reconstructed from a flexible laser plane.
There is only one intersection between the flexible laser plane and the target.
Since the flexible laser plane is determined by at least two lines on the plane, a double
In order to perform laser plane calibration in the camera coordinate system, a plane reference is proposed. The bi-
The plane method with two plane references has four main advantages belonging to 2D reference and 3D reference.
When the two plane references are located on a different plane, the feature points on the two planes
The flat reference is similar to the reference on the 3D reference.
Therefore, it should be noted that calibration and reconstruction are only achieved through one image of the camera.
In addition, the calibration reference consists of two plane references, which are easier to manufacture than the 3D reference.
Large 3D or 2D references are then complex for contour measurement in a large field of view.
Reconstruction Method of Bi
Graphic references can easily be extended-
Flat reference. The -
The plane reference method extends effective measurements in a larger field of view.
However, large 3D or 2D references with larger horizons are replaced by small 2D references that are easy to implement in reconstruction.
Finally, the laser plane is flexible. g. hand-
Held as a laser plane projected to bi-
Plane reference can only be solved in one image.
An object contour optimization reconstruction method for flexible plane laser and dual plane laser vision measurement is proposed.
Flat reference.
Two plane references are distributed on the left and right sides of the object under measurement.
Two plane reference non
Space in total.
First, the laser plane generated by the laser projector is projected onto the object under test.
Therefore, the laser plane intersects the object and the plane reference that provides the intersection curve and two intersection lines.
Then, in the camera coordinate system, the flexible laser plane is modeled by the projection of the feature points on the two plane references and the image coordinates of the laser intersection.
Finally, the 3D coordinates of the object under test are determined by the projection of the intersection curve, and further enhanced by the optimization function.
The optimization method is compared with the initial method by reconstruction distance error, and the accuracy of reconstruction method is verified.
Flat reference. The bi-
Plane reference is considered to be a flexible benchmark for realizing the conversion between two-world coordinate systems
Plane reference, camera coordinate system, and image coordinate system.
The laser plane is confirmed by the intersection between bi-
Plane reference and laser plane.
The 3D camera coordinates of the intersection of the laser plane with the measured object are initially reconstructed by the image coordinates of the intersection, the inherent parameter matrix, and the laser plane.
At the same time, using the parametric difference of the reconstruction distance of the target with eight marks and the parametric re-projection error of the feature points on the double mark, an optimization function is designed.
Flat reference.
Reconstruction method based on dual domain
The plane reference is evaluated by comparing the difference between the real distance and the standard distance.
The reconstruction error of the initial method is 1. 01u2009mm.
Moreover, the average reconstruction error of the optimization method is 0. 93u2009mm.
Therefore, the double optimization method
The planar references have great application prospects in contour reconstruction.
Visual measurement including structural light is an effective non-
Method of contact 3D measurement.
Since the laser projector provides stronger and narrower lighting than the ordinary digital projector, the laser projector is considered a flexible symbol for projecting the laser plane onto the object to be measured.
The laser curve between the object being measured and the laser plane contains information about the intersection position and the depth of the object being measured.
Therefore, the measurement system based on structured light has been favored by many researchers.
Visual measurement based on structural light has been widely studied due to its wide measurement range, reasonable test speed and high precision in visual measurement methods.
Huynh proposed a calibration method for structured light system based on projector.
The cross ratio in the contribution image of the three collineal world points is the same as the cross ratio calculated by the world points.
According to the projection from the structural light to the image, the recovery matrix of the stripe plane is provided.
Wei also reported a bi-crossover ratio approach to achieve structured-light-
Stripe vision sensor.
Wei uses vanishing points and vanishing lines derived from targets with parallel lines to improve laser plane calibration.
Li has designed a flexible laser scanning system including industrial robot arm and laser scanner.
The rotation and peaceful movement of the robot arm are considered in the scanning model.
Niola describes the calibration of laser scanners in robot applications.
Use a target of known motion to simulate the geometry of the laser transmitter and camera.
Through the fusion of structured lighting and data, the reconstruction of sharp edges of three dimensions is realized, and the automatic detection and reconstruction are realized.
In addition, an algorithm to reconstruct the contour of 3D objects using sharp edges is introduced.
Ma proposes a 3D contour reconstruction method based on Nonlinear Iterative optimization to reduce the error caused by lens distortion.
According to the shape of the projected light, the visual measurement method based on the structure light can be divided into four parts: point structure light, line structure light, grating structure light and coded structure light.
The point structured light method can obtain the 3D data of a point.
However, the speed is slow for large object measurements.
The method of grating or coding structured light is usually based on digital light processing (DLP)
Projector and camera.
Villa recommended reference-
A plane-based approach to reconstruct 3D data on an object.
Depth is achieved by moving a tablet on a linear stage.
Then, the cross grating pattern projected along and in the direction on the object under test.
And coordinates are determined by the stripe phase generated by the Fourier transform and the inverse Fourier transform.
Zhang proposed a method of using a digital projector as a camera.
Digital Micromirror device (DMD)
Like a map between the Republic of Uzbekistan (
Charge coupler
CCD pixels and dvr pixels.
The inherent parameter matrix of the digital projector is then calibrated as the matrix of the camera.
An external parameter matrix is derived from Zhang\'s method.
Finally, 3D coordinates are reconstructed by image coordinates and external parameter matrix.
Hu introduced a method to calibrate the projector. camera system.
The absolute phase diagram consists of three-step algorithm.
Then, a flat plate was designed on the linear stage and driven by a stepping motor.
By moving the plate and projected images, four unknown parameters were calibrated for the measurement system.
Finally, the system parameters are optimized by the coordinate measuring instrument (CMM)
, Plate with holes and iterative algorithm. The digital-projector-
The method has the advantages of high efficiency and rich data.
However, since the pattern of the digital projector is generated by the bulb, the coded light pattern of the digital projector is often affected by lighting from the environment (such as sunlight.
After calibration, the relative position between the digital projector and the camera is fixed.
The flat Laser Projector helps to narrow downband and high-
Density structured light, by avoiding environmental interference with lighting, it is appropriate to measure complex surfaces at medium test speeds.
Camera Calibration is the basis of the accuracy of the visual measurement system.
3D recovery depends on one or more
View the images taken by the camera.
Therefore, improving the accuracy of 3D reconstruction has also attracted the attention of researchers.
At present, the calibration methods are mainly divided into 1D calibration, 2D calibration and 3D calibration.
1D calibration reference is simple in structure and easy to manufacture, but the calibration accuracy is the lowest due to the lack of information on 1D reference.
Although the structure of the 2D calibration benchmark is a little more complicated than the structure of the 1D benchmark, it has the advantage of high calibration accuracy, which is easy to make and move to different places.
The 3D calibration reference provides the highest accuracy for camera calibration.
Nevertheless, the 3D calibration reference is complex and suitable for specific situations.
Although 2D reference can be used to calibrate the projected laser plane, many images of the plane reference at different positions should be captured in order to calibrate only one laser plane.
When the relative position between the camera and the projector is fixed in the previous work, a target can provide a sufficient number of features to solve the intersection.
We are concerned with another case where the intersection point is reconstructed from a flexible laser plane.
There is only one intersection between the flexible laser plane and the target.
Since the flexible laser plane is determined by at least two lines on the plane, a double
In order to perform laser plane calibration in the camera coordinate system, a plane reference is proposed. The bi-
The plane method with two plane references has four main advantages belonging to 2D reference and 3D reference.
When the two plane references are located on a different plane, the feature points on the two planes
The flat reference is similar to the reference on the 3D reference.
Therefore, it should be noted that calibration and reconstruction are only achieved through one image of the camera.
In addition, the calibration reference consists of two plane references, which are easier to manufacture than the 3D reference.
Large 3D or 2D references are then complex for contour measurement in a large field of view.
Reconstruction Method of Bi
Graphic references can easily be extended-
Flat reference. The -
The plane reference method extends effective measurements in a larger field of view.
However, large 3D or 2D references with larger horizons are replaced by small 2D references that are easy to implement in reconstruction.
Finally, the laser plane is flexible. g. hand-
Held as a laser plane projected to bi-
Plane reference can only be solved in one image.
An object contour optimization reconstruction method for flexible plane laser and dual plane laser vision measurement is proposed.
Flat reference.
Two plane references are distributed on the left and right sides of the object under measurement.
Two plane reference non
Space in total.
First, the laser plane generated by the laser projector is projected onto the object under test.
Therefore, the laser plane intersects the object and the plane reference that provides the intersection curve and two intersection lines.
Then, in the camera coordinate system, the flexible laser plane is modeled by the projection of the feature points on the two plane references and the image coordinates of the laser intersection.
Finally, the 3D coordinates of the object under test are determined by the projection of the intersection curve, and further enhanced by the optimization function.
The optimization method is compared with the initial method by reconstruction distance error, and the accuracy of reconstruction method is verified.
Custom message
