CN114049324B - Rapid calibration method for correlated reference telecentric measurement under super-view field scale - Google Patents
Rapid calibration method for correlated reference telecentric measurement under super-view field scale Download PDFInfo
- Publication number
- CN114049324B CN114049324B CN202111344665.0A CN202111344665A CN114049324B CN 114049324 B CN114049324 B CN 114049324B CN 202111344665 A CN202111344665 A CN 202111344665A CN 114049324 B CN114049324 B CN 114049324B
- Authority
- CN
- China
- Prior art keywords
- measurement
- camera
- telecentric
- calibration
- coordinates
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000005259 measurement Methods 0.000 title claims abstract description 108
- 238000000034 method Methods 0.000 title claims abstract description 57
- 230000002596 correlated effect Effects 0.000 title description 4
- 230000005540 biological transmission Effects 0.000 claims abstract description 30
- 230000007246 mechanism Effects 0.000 claims abstract description 26
- 238000003384 imaging method Methods 0.000 claims abstract description 18
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000011065 in-situ storage Methods 0.000 claims abstract description 8
- 238000004364 calculation method Methods 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 42
- 230000009466 transformation Effects 0.000 claims description 24
- 238000013519 translation Methods 0.000 claims description 15
- 239000013598 vector Substances 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 239000003550 marker Substances 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000005286 illumination Methods 0.000 claims description 5
- 230000010349 pulsation Effects 0.000 claims description 5
- 238000006073 displacement reaction Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 2
- 238000013178 mathematical model Methods 0.000 claims description 2
- 230000003287 optical effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000013461 design Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention provides a quick calibration method of association reference telecentric measurement under the scale of super view field, which is based on telecentric vision system to realize the calibration of association reference telecentric measurement under the scale of view field by an in-situ calibration method and a dimension chain association method from a measurement area to a measurement reference axis, and comprises the following steps: the device comprises a workbench, a precise pose adjusting mechanism, a measuring camera, a workpiece to be measured, a background light source, a triggerable photoelectric sensing mechanism, a transmission target and a three-dimensional correlation target; the invention realizes the image feature extraction and the position accurate measurement of the local element to be measured of the longer-scale workpiece, then establishes a measurement transmission chain of the local element to be measured and the axis reference element outside the imaging view field of the camera, and converts the local element into the global radial dimension through calculation, thereby realizing the rapid and accurate measurement exceeding the scale of the imaging view field.
Description
Technical Field
The invention belongs to the technical field of quick high-precision online visual calibration measurement in industrial detection of production lines, and particularly relates to a quick calibration method for associated reference telecentric measurement under the super-view field scale.
Background
In industrial detection application, the topological structure of a part is shown in fig. 1, the part has the actual requirement of on-line measurement, the distance between the tip edge element extending out of the cross section and the axis of the long rod is a key geometric quantity for measurement, the rotation precision after assembly is related, and the structural characteristics of the part bring certain difficulty to on-line measurement. The parts are similar to a precise shafting, have long rod technical characteristics and have rotation technical characteristics; the shaft parts are various in form and can be thin-wall saw blades, outer tubes and the like, the workpiece is damaged by contact measurement, and the key problem is how to realize quick and high-precision non-contact measurement.
Because of the specificity of the workpiece structure, the existing non-contact measurement method and equipment are limited in application, such as grating projection measurement, are sensitive to reflective materials on the metal surface, and elements to be measured are tip corner points, so that accurate extraction of the elements is limited; when a three-coordinate measuring machine laser triangulation method is adopted for measurement, the similar problems are faced, and the precise acquisition of the tip corner elements is difficult; the more common image measurement mode has the possibility of measurement when the epaxial part is in stock both sides terminal surface position department, but when epaxial part is located the middle section footpath, how to realize the distance of edge point element to axis has become the difficult problem, and the mode that adopts at present is mostly artifical image decomposition measurement, but inefficiency, assembly precision can't objectively evaluate. In addition, due to the radial dimension expansion characteristics of the on-axis parts, high-precision optical measurement systems generally cannot image both the axis end face and the part tip element at one large optical aperture, because the large optical aperture affects measurement precision, and in addition, the large mechanical aperture is covered with the large optical aperture, which means an increase in cost and a relatively expensive system design. Therefore, the research of a non-contact quick and high-precision calibration method suitable for the workpieces has very important significance.
Disclosure of Invention
The application relates to a quick calibration method of associated reference telecentric measurement under the super-view field scale, which belongs to an indirect calibration method of space scale images, firstly, the method utilizes telecentric imaging characteristics to realize image feature extraction and position accurate measurement of local to-be-measured elements of a longer scale workpiece, and then, a measurement transmission chain of the local element to be measured and an axis reference element outside the imaging view field of the camera is established, and the local element is converted into a global radial dimension through calculation, so that the rapid and precise measurement exceeding the scale of the imaging view field is realized. The operation of the measuring device is shown in fig. 2.
In order to achieve the above purpose, the invention adopts the following technical scheme:
The invention relates to a quick calibration method for correlated reference telecentric measurement under the scale of a super-view field, which is based on a telecentric vision system, and realizes the calibration of correlated reference telecentric measurement under the scale of the view field through an in-situ calibration method and a dimension chain correlation method from a measurement area to a measurement reference axis, wherein the telecentric vision system comprises: the device comprises a workbench, a precise pose adjusting mechanism, a measuring camera, a workpiece to be measured, a background light source, a triggerable photoelectric sensing mechanism, a transmission target and a three-dimensional correlation target; wherein:
The in-situ calibration method is characterized in that a transmission target is placed in a guide groove between a measuring camera and a background light source, so that a checkerboard of the target surface is full of the field of the measuring camera, and the axial displacement of the transmission target is finely adjusted to be positioned in a depth-of-field range of the measuring camera to calibrate an image sequence; extracting checkerboard corner coordinates of a calibration image sequence by using a sub-pixel corner extraction algorithm, and combining known corner world coordinates to perform measurement camera joint calibration;
the dimension chain association method from the measurement area to the measurement reference axis is characterized in that the dimension chain association method from the measurement area to the measurement reference axis is placed on the precise pose adjusting mechanism through a three-dimensional association target, the cross section overhanging target mark point rotates into the field of view of a measurement camera, the three-dimensional association target and the standard axis are integrally manufactured, the three-dimensional association target and the standard axis are the overhanging target structure of the long axis cross section, and the target surface is a mark lattice.
Further: the measuring camera can adopt a 0.5-time high-resolution low-distortion telecentric lens; the background light source 7 can adopt a telecentric parallel light source and a blue light cold light source for illumination; a checkerboard pattern with a positional accuracy of 1 μm can be used for penetration of the target 9.
The method for calibrating the position of the telecentric vision system comprises the following steps: the telecentric vision system is composed of a measuring camera, a transmission target and a background illumination light source, and because the requirements of quick and high-precision measurement are simultaneously met, the vision measuring system is designed into a narrow-view-field high-precision telecentric vision measuring system, and a monitoring view field only covers an overhanging tip element of the cross section of a workpiece to be measured so as to ensure high-precision measurement of geometric elements in a local measuring area of the workpiece to be measured, and the specific calibration process comprises the following steps:
According to the technical parameters of each component, fine adjustment of the working distance, the angle and the horizontal state of each camera and each light source is carried out through a workbench surface guide mechanism and a precise pose adjusting mechanism, so that the imaging effect is in an optimal state, and each component is locked after adjustment;
The transmission target is placed in a guide groove between the camera and the light source, the height and the posture are finely adjusted, so that a checkerboard of the target surface is full of the field of view of the camera, the axial displacement of the target is finely adjusted, and the target is located in the depth of field range of the camera and clearly imaged;
The telecentric measurement camera acquires calibration images of the target in such a way that the transmission target axially moves a plurality of positions within the depth of field of the camera, and images of the target at the plurality of positions are shot to form a plurality of groups of calibration image sequences within the depth of field;
Extracting checkerboard corner coordinates of a plurality of groups of calibration image sequences by using a subpixel corner extraction algorithm, combining the known corner world coordinates, carrying out telecentric lens joint calibration, and calculating to obtain calibration parameters such as magnification, distortion coefficients and the like;
randomly moving the target to any position in the depth of field, calculating a reprojection error, and completing the calibration of the telecentric lens after meeting the precision requirement.
Further, when the transmission target is used for measuring camera calibration, the plane where the target surface checkerboard pattern is located is the X-Y plane of the world coordinate system, the plane is also the Z=0 direction, and the external parameter matrix of the parallel projection model can be reduced by one dimension, as shown in the formula (1):
Firstly, each parameter in the transformation matrix is subjected to preliminary reduction solution according to an ideal parallel projection model. After the deformation of the formula (1), the formula (2) is as follows:
converting the formula (2) into a least square structural form, as shown in the formula (3):
Wherein, record The standard form al=b, which can be written as least squares.
Substituting the image coordinates and world coordinates of the checkerboard corner into formula (3) to obtain a matrix A, B, and obtaining a coefficient matrix L through a least square algorithm. Considering the property that the rotation matrix is a unit orthogonal matrix, there is a relationship as shown in equation (4) between the vectors in the matrix:
the above formulae can be combined to give formula (5):
By combining the formulas (3) and (5), the calculation of the translation parameter T y can be realized, and the solution of other parameters in the coefficient matrix can be further realized, so that the rotation matrix R and the translation vector T are finally obtained.
According to the formula (2), the rotation matrix and the translation vector are combined, and the magnification m of the double telecentric lens can be obtained by inverse solution.
In the above case, the distortion effect of the lens is not considered, the distortion of the telecentric lens is extremely small, and the effect of the distortion coefficient of more than three terms can be ignored in the distortion model. In order to improve the optimization efficiency, the radial distortion and tangential distortion existing in the lens are considered during calibration by combining the distortion model, and a distortion model of the double telecentric lens in the measurement system is established, as shown in a formula (6):
Wherein, delta u、δv is the distortion of the computer image in the u and v directions respectively; k 1、p1、p2 is the radial and tangential distortion coefficient. For the imaging transformation model, the following relationship exists between the ideal image coordinates (u u,vu) and the actual image coordinates (u d,vd) of any point, as shown in equation (7):
After the internal and external parameters under the linear imaging model are calculated, the distortion influence of the lens is considered. Because the distortion of the double telecentric lens is small, in order to improve the calibration efficiency, the initial value of the distortion coefficient can be set to 0, the internal and external parameters which are solved previously are combined to be unified as an initial value, a target equation is established, and the problem is converted into a problem of nonlinear optimization to solve the optimal solution. The link adopts LM algorithm to find the optimal solution of the target equation, and the difference between the coordinates of the calculated corner point and the coordinates of the actual corner point is minimized by establishing a mathematical model, so that a system equation is established, as shown in the formula (8):
Wherein n represents the number of corner points; p di is the actual image coordinates of corner p i; p ui is the image coordinates obtained by the telecentric imaging model; where p ui can be denoted as p ui(m,R,T,k1,p1,p2), m is the magnification of the double telecentric lens, R is the rotation matrix, T is the translation vector, and k 1、p1、p2 is the distortion coefficient. And substituting the initial values into the target equation, and minimizing the target equation to obtain the optimal solution of each parameter.
Further, the dimension chain association method from the measurement area to the measurement reference axis can associate the geometric elements of the local measurement area to the rotation axis measurement reference outside the camera field of view, so as to realize the dimension association of the super field of view, and the specific calibration process comprises the following steps:
and (3) placing the three-dimensional associated targets on two-end precise V-shaped supports, and rotating the cross section overhanging target mark points into the field of view of the camera.
The three-dimensional correlation target has been subjected to mapping of spatial elements, fitting of axes and conversion and generation of a coordinate conversion matrix from the mark point to the axes in place, namely, a spatial conversion relation matrix from the mark point to the axes is established.
When the three-dimensional associated target is in the field of view of the camera, the world coordinates (defined at the upper left corner of the target surface) and the pixel coordinates of the marker point are recorded at the same time, namely the coordinates of the marker point under the coordinate system of the camera.
Establishing a size chain: knowing the coordinates of the target surface mark points under the camera coordinate system and knowing the coordinates of the points under the axis reference space rectangular coordinate system, obtaining the transformation relation from the camera coordinate system to the axis space rectangular coordinate system through calibration model calculation, establishing a space transmission size chain, realizing the super-view field geometric quantity measurement of the local view field associated reference axis, and obtaining the radial size information of the tip element to be measured.
Further, the transformation relation process from the camera coordinate system to the axis space rectangular coordinate system is obtained through the calculation of the calibration model by the space dimension chain:
converting the solution of the transformation relationship between the world coordinates in the telecentric imaging model and the world coordinates in the reference correlation model into the solution of the matrix H in the formula (9):
B＝HA (9)
wherein A (X 1、Y1、Z1) is the world coordinate of each mark point under the coordinate system of the camera; b (X 2、Y2、Z2) is the three-dimensional world coordinate of a in the reference associated axis coordinate system. The further solution to the transformation matrix H is the solution to the rotation matrix R and the translation vector T, R and T being transformation relation matrices between two coordinate systems.
And solving the rotation matrix and the translation vector by utilizing a singular value decomposition method. In this method, the construction and decomposition of the matrix is mainly involved; a and B are obtained under the camera coordinate system and the axis reference coordinate system respectively, and a transformation matrix H is constructed according to the A and the B.
Where M' i、P′i represents the world coordinates of the centroid in each of the A and B coordinate systems, respectively.
Singular value decomposition of H:
H＝U×S×VT (11)
Finally, through solving The maximum value of (2) can be obtained as the optimal solution of the rotation matrix R:
R＝VUT (12)
finally substituting the value of R into a relational expression, and correspondingly, solving the value of the translation vector T;
Further, the position calibration method and the dimension chain association method acquire the world coordinates into a transformation relation process through rotation and translation: after the transformation relation between the two coordinates is measured by a telecentric imaging system, the world coordinate C (X 3、Y3、Z3) of the geometric element can be obtained by the transformation relation between the established reference relations, and the three-dimensional world coordinate D (X 4、Y4、Z4) of the coordinate C (X 3、Y3、Z3) in the reference coordinate system can be obtained, so that the geometric quantity to be measured can be calculated.
Further, the telecentric vision system further comprises a triggerable photo-sensing mechanism, wherein: the telecentric vision system obtains radial geometric quantity by adopting a continuous measurement mode and a pulsating multi-point measurement mode respectively, and comprises the following steps:
When the measured workpiece rotates for a certain stroke on the precise pose adjusting mechanism, a continuous image sequence of the tip element of the workpiece to be measured is obtained through the measuring camera, a continuous tip motion track is extracted from the continuous image sequence, the radial geometric quantity to be measured can be calculated by utilizing the arc fitting method, and a continuous measuring mode is realized;
When the overhanging tip of the cross section of the workpiece to be measured moves to a designated position, the photoelectric sensing device can be triggered to synchronously shoot the measuring camera, the extraction of a single angular point is instantly completed, and the radial geometric quantity of the single-point element to be measured is calculated; and if the tips of the workpiece to be measured are symmetrically distributed in the circumferential direction, a multipoint pulsation radial geometry measuring result is formed, and a pulsation multipoint measuring mode is realized.
Advantageous effects
1. The telecentric calibration system is easy to realize small-view-field high-precision measurement due to the parallel projection measurement characteristic, and has wider application in the field of plane geometry measurement application. The method relates to the occasion of measuring the space dimension of the workpiece, generally adopts telecentric camera and guide rail executing mechanism combination measurement and is realized by manual operation, and also belongs to another application of plane measurement, and the time consumption is long; compared with the prior art, the method is characterized in that a telecentric calibration method is adopted, the real-time measurement of the spatial dimension of the stereoscopic workpiece can be realized, the requirements of quick and high-precision measurement of the dimension can be met, and the practical problems that special non-contact measurement equipment is absent for the parts, manual operation is needed under the environment of an image measuring instrument and the like, and step-by-step inspection are solved; in the production application of the developed equipment, the improvement on the detection efficiency of the workpiece is obvious, and the effectiveness of the related reference telecentric measurement quick calibration method under the super-view field scale is verified.
2. According to the invention, the long rod piece to be measured is placed on the precise pose adjusting mechanisms at the two ends, and an accurate rotation axis reference is provided by means of the rotation characteristics of the long rod piece. The lower end of the precise pose adjusting mechanism is provided with a lifting part which is used for compensating the height difference formed by different shaft diameter dimensions and can be subjected to leveling treatment, so that the axis is horizontal when the long rod piece is in a measuring state. The measuring camera is positioned on one side of the long rod piece, a telecentric lens with small optical aperture and higher multiplying power is adopted, and the view field only covers the local area of the tip element to be measured. The background light source is positioned on the other side of the workpiece to be measured, and forms correlation illumination layout with the measuring camera, so as to improve imaging contrast of the tip element. The working table surface of the mounting substrate of the measuring device is provided with a dovetail guide mechanism for ensuring the coaxiality of the coaxial optical component tool on the table.
4. The invention aims to realize the measurement, when the element to be measured rotates to the field of view of the measurement camera, the measurement camera can perform position measurement on the tip corner element in the local area on one hand, and on the other hand, the measurement position can be related to an axis measurement reference to realize global dimension measurement. The functions are realized by the related reference telecentric measurement rapid calibration method under the super-view field scale.
5. The invention adopts a telecentric calibration measurement method, but can realize the real-time measurement of the spatial dimension of the stereoscopic workpiece, namely, can simultaneously meet the requirements of quick and high-precision measurement of the dimension.
Drawings
FIG. 1 is a schematic diagram of the topology of a workpiece to be measured;
FIG. 2 is a schematic diagram (axial view) of the operation of the measuring device;
FIG. 3 is a schematic diagram (radial view) of the operation of the measuring device;
FIG. 4 is a schematic diagram of a telecentric vision system in-situ calibration method;
FIG. 5 is a schematic diagram of a dimension chain correlation method from a measurement region to a measurement reference axis.
Detailed Description
An embodiment of the present invention is shown in fig. 2 to 5.
As shown in fig. 2 and 3, the telecentric vision system includes: the visual measurement system is designed into a narrow-view-field high-precision telecentric visual measurement system, and a monitoring view field only covers overhanging tip elements of the cross section of the workpiece to be measured so as to ensure high-precision measurement of geometric elements in a local measurement area of the workpiece to be measured, because of the requirements of both rapid and high-precision measurement.
According to the technical parameters of each component, the working distance, the angle and the horizontal state of each of the measuring camera 5 and the background light source 7 are finely adjusted through the embedded guide mechanism of the working table surface 1 and the precise pose adjusting mechanism 4, so that the imaging effect is in an optimal state, and each component is locked after adjustment.
As shown in fig. 4, the transmission target 9 is placed in the guide slot between the measurement camera 5 and the background light source 7, the height and the posture are finely adjusted, so that the checkerboard of the target surface is full of the field of the measurement camera 5, and the axial displacement of the transmission target 9 is finely adjusted, so that the transmission target 9 is in the depth of field of the measurement camera 5 and is clearly imaged.
The measurement camera 5 acquires the calibration image of the transmission target 9 in such a way that the transmission target 9 axially moves by 3 positions within the depth of field range of the measurement camera 5, images of the transmission target 9 at the 3 positions are shot, and 3 groups of calibration image sequences within the depth of field range are formed.
And (3) extracting checkerboard corner coordinates of 3 groups of calibration image sequences by using a subpixel corner extraction algorithm, combining the known corner world coordinates, carrying out joint calibration of the measurement camera 5, and calculating to obtain calibration parameters such as magnification, distortion coefficients and the like according to the processes of the formulas (1) - (8).
Randomly moving the transmission target 9 to any position in the depth of field range, checking and calculating the re-projection error, and completing the calibration of the measurement camera 5 after the accuracy requirement is met.
As shown in fig. 5, the three-dimensional correlation target 10 is placed on the precise pose adjustment mechanism 3, and the cross section of the three-dimensional correlation target 10 extends out to the target mark point and rotates into the field of view of the measurement camera 5. Wherein, the three-dimensional associated target 10 and the standard shaft are made into a whole, which is also a long-axis cross section overhanging target structure, and the target surface is a mark lattice.
The three-dimensional correlation target 10 has been mapped on site, fitted to the axis, and converted and generated from the marker point to axis coordinate transformation matrix, i.e., the marker point to axis spatial transformation relationship matrix is established.
When the three-dimensional correlation target 10 is within the field of view of the measurement camera 5, the world coordinates (defined at the upper left corner of the target surface) of the marker point and the pixel coordinates, i.e. the coordinates of the marker point in the camera coordinate system, are recorded simultaneously.
Knowing the coordinates of the target surface mark points under the camera coordinate system and knowing the coordinates of the points under the axis reference space rectangular coordinate system, and calculating through a calibration model, the above process formulas (9) - (12) can obtain the transformation relationship from the camera coordinate system to the axis space rectangular coordinate system, establish a space transmission size chain, realize the hyperfield geometric measurement of the local field of view associated reference axis, and obtain the radial size information of the tip element to be measured.
The related reference telecentric measurement rapid calibration method under the super-view field scale comprises the following implementation aspects of a telecentric vision system: the measuring camera 5 can adopt a 0.5-time high-resolution low-distortion telecentric lens; the background light source 7 can adopt a telecentric parallel light source and a blue light cold light source for illumination; the transmissive target 9 may employ a checkerboard pattern with a positional accuracy of 1 μm.
The related reference telecentric measurement rapid calibration method under the super-view field scale can adopt a guide mechanism design, an automatic centering mechanism design and a precise fine tuning mechanism design to ensure the consistency of the tool, as shown in fig. 2.
The method can be applied to a plurality of measurement modes, as follows:
Continuous measurement mode. As shown in fig. 3, when the workpiece 6 rotates on the precise pose adjusting mechanism 3 for a certain stroke, a continuous image sequence of the tip element of the workpiece 6 to be measured is obtained by the measuring camera 5, a continuous tip motion track is extracted from the continuous image sequence, and the radial geometry to be measured can be calculated by using an arc fitting method.
Pulsed multipoint measurement mode. As shown in fig. 3, when the overhanging tip of the cross section of the workpiece 6 to be measured moves to a designated position, the photoelectric sensing device 8 can be triggered to synchronously shoot the measuring camera 5, so that the extraction of a single angular point is instantly completed, and the radial geometric quantity of the single-point element to be measured is calculated; if the tips of the workpiece 6 to be measured are symmetrically distributed in the circumferential direction, a multipoint pulsation radial geometry measuring result is formed.
The application fully analyzes and synthesizes the characteristics of measurement requirements, such as high measurement accuracy but smaller field of view of a telecentric measurement system; the axial and radial dimensions of the workpiece to be measured are unequal, and the structural shape is not beneficial to direct measurement of the same optical view field; the three-dimensional projection measurement is not facilitated by the metal reflective material and the sharp points, and various practical problems such as difficulty in touch of the thin wall are also included. In view of the problems, the method for rapidly calibrating the associated reference telecentric measurement under the super-view field scale is designed and researched, the validity of the method is verified on researched equipment, and the problem that no universal detection equipment is currently available for detecting the workpieces on line is solved.
Claims (2)
1. The quick calibration method for the associated reference telecentric measurement under the super-view field scale is characterized in that the calibration method is based on a telecentric vision system, and the associated reference telecentric measurement calibration under the view field scale is realized through an in-situ calibration method and a dimension chain association method from a measurement area to a measurement reference axis, and the telecentric vision system comprises: the device comprises a workbench, a precise pose adjusting mechanism, a measuring camera, a workpiece to be measured, a background light source, a triggerable photoelectric sensing mechanism, a transmission target and a three-dimensional correlation target; wherein:
the in-situ calibration method of the telecentric vision system comprises the following steps:
According to the technical parameters of each component, the working distance, the angle and the horizontal state of each of the measuring camera and the background light source are finely adjusted through the embedded guide mechanism of the working table surface and the precise pose adjusting mechanism, and each component is locked after adjustment;
the transmission target is placed in a guide groove between the measuring camera and the background light source, the height and the gesture are finely adjusted, so that a checkerboard of the target surface is full of the field of the measuring camera, the axial displacement of the transmission target is finely adjusted, and the transmission target is located in the depth of field of the measuring camera and clearly imaged;
the method comprises the steps that a measurement camera obtains calibration images of a transmission target in a mode that the transmission target axially moves a plurality of positions within the depth of field of the measurement camera, and images of the transmission target at the plurality of positions are shot to form a plurality of groups of calibration image sequences within the depth of field;
Extracting checkerboard corner coordinates of a plurality of groups of calibration image sequences by using a subpixel corner extraction algorithm, combining the known corner world coordinates, carrying out measurement camera joint calibration, and calculating to obtain magnification and distortion coefficient calibration parameters;
Randomly moving the transmission target to any position in the depth of field, checking and calculating a re-projection error, and completing calibration of a measurement camera after the accuracy requirement is met;
The dimension chain association method from the measuring area to the measuring reference axis comprises the following steps:
Placing the three-dimensional associated targets on precise pose adjusting mechanisms at two ends, and rotating the cross section overhanging target mark points into the field of view of the camera;
The three-dimensional correlation target has been mapped by in-situ calibration, fitted to the axis, converted and generated from the marker point to the axis coordinate conversion matrix, i.e. the spatial conversion relation matrix from the marker point to the axis is established;
When the three-dimensional associated target is in the field of view of the camera, coordinates of the mark points under a camera coordinate system are obtained, and world coordinates and pixel coordinates of the mark points are recorded;
establishing a space dimension chain: knowing the coordinates of the target surface mark points under a camera coordinate system and knowing the coordinates of the points under an axis reference space rectangular coordinate system, and calculating through a calibration model; the transformation relation from the camera coordinate system to the axis space rectangular coordinate system is obtained, a space transmission size chain is established, the super-view field geometric quantity measurement of the local view field associated reference axis is realized, and the radial size information of the tip element to be measured is obtained;
the process for obtaining the plurality of groups of calibration image sequences comprises the following steps:
when a transmission target is used for measuring camera calibration, the plane where the target surface checkerboard pattern is located is an X-Y plane of a world coordinate system, the plane is also in the Z=0 direction, and the external parameter matrix of the parallel projection model can be reduced by one dimension, as shown in the formula (1):
carrying out preliminary simplified solution on each parameter in the transformation matrix according to the parallel projection model; after the deformation of the formula (1), the formula (2) is as follows:
converting the formula (2) into a least square structural form, as shown in the formula (3):
Wherein, record The standard form al=b of formula (3) denoted as least squares;
Substituting the image coordinates and world coordinates of the checkerboard corner points into the formula (3) to obtain a matrix A, B, and obtaining a coefficient matrix L through a least square algorithm; the rotation matrix is a unit orthogonal matrix, and then the relation shown in the formula (4) exists between the vectors in the matrix:
simultaneously with the above formulae, formula (5) is obtained:
Combining the formulas (3) and (5), solving the translation parameter T y, and further solving other parameters in the coefficient matrix to finally obtain a rotation matrix R and a translation vector T;
According to the formula (2), combining the rotation matrix and the translation vector, and reversely solving to obtain the magnification m of the double telecentric lens;
in the above situation, the distortion influence of the lens is not considered, the distortion of the telecentric lens is extremely small, and the influence of distortion coefficients above three times is ignored in the distortion model; in order to improve the optimization efficiency, the distortion model is combined, radial distortion and tangential distortion existing in the lens are considered in calibration, and a distortion model of the double telecentric lens in the measurement system is built, as shown in a formula (6):
Wherein, delta u、δv is the distortion of the computer image in the u and v directions respectively; k 1 is the radial distortion coefficient and p 1、p2 is the tangential distortion coefficient; for the imaging transformation model, the following relationship exists between the ideal image coordinates (u u,vu) and the actual image coordinates (u d,vd) of any point, as shown in equation (7):
After the internal and external parameters under the linear imaging model are calculated, the distortion influence of the lens is considered, and as the distortion of the double telecentric lens is small, the initial value of the distortion coefficient is set to 0, the internal and external parameters which are solved previously are combined to be unified as initial values, a target equation is established, and the problem is converted into the problem of nonlinear optimization to calculate the optimal solution; the link adopts an LM algorithm to find an optimal solution of a target equation, a mathematical model is established to minimize the difference between the coordinates of the calculated corner point and the coordinates of the actual corner point, and the target equation is established as shown in a formula (8):
Wherein n represents the number of corner points; p di is the actual image coordinates of corner p i; p ui is the image coordinates obtained by the telecentric imaging model; where p ui is denoted as p ui(m,R,T,k1,p1,p2), m is the magnification of the double telecentric lens, R is the rotation matrix, T is the translation vector, and k 1、p1、p2 is the distortion coefficient; after substituting the initial values into the target equation, minimizing the target equation, namely solving the optimal solution of each parameter;
the transformation relation process from the camera coordinate system to the axis space rectangular coordinate system is obtained through the calculation of the calibration model by the space dimension chain:
converting the solution of the transformation relationship between the world coordinates in the telecentric imaging model and the world coordinates in the reference correlation model into the solution of the matrix H in the formula (9):
B＝HA (9)
Wherein A (X 1、Y1、Z1) is the world coordinate of each mark point under the coordinate system of the camera; b (X 2、Y2、Z2) is the three-dimensional world coordinate of A in a reference associated axis coordinate system; the further solving of the transformation matrix H is the solving of the rotation matrix R and the translation vector T, wherein R and T are transformation relation matrixes between two coordinate systems;
solving a rotation matrix and a translation vector by utilizing a singular value decomposition method; wherein: the construction and decomposition of the matrix is as follows:
Respectively obtaining A and B under a camera coordinate system and an axis reference coordinate system, and constructing a transformation matrix H according to the A and B;
Wherein M' i、P′i represents the world coordinates of centroid in each of the coordinate systems A and B;
Singular value decomposition of H:
H＝U×S×VT (11)
Finally, through solving The maximum value of (2) can be obtained as the optimal solution of the rotation matrix R:
R＝VUT (12)
substituting the value of R into a relational expression P' i＝RM′i +T, and obtaining the value of a translation vector T;
The in-situ calibration method and the dimension chain association method establish a reference association transformation relationship between two coordinates; after the world coordinates C (X 3、Y3、Z3) of the geometric elements are measured by a telecentric imaging system, the three-dimensional world coordinates D (X 4、Y4、Z4) of the coordinates C (X 3、Y3、Z3) in a reference coordinate system are obtained by the established transformation relation between the reference correlations, so that the geometric quantity to be measured is calculated;
the telecentric vision system obtains radial geometric quantity by adopting a continuous measurement mode and a pulsating multi-point measurement mode respectively, and comprises the following steps:
When a measured workpiece rotates for a certain stroke on a precise pose adjusting mechanism, a continuous image sequence of a tip element of the workpiece to be measured is obtained through a measuring camera, a continuous tip motion track is extracted from the continuous image sequence, a radial geometric quantity to be measured is obtained through calculation by using an arc fitting method, and a continuous measuring mode is realized;
When the overhanging tip of the cross section of the workpiece to be measured moves to a designated position, triggering the photoelectric sensing device to enable the measuring camera to synchronously shoot, instantly completing the extraction of a single angular point and calculating the radial geometric quantity of the single-point element to be measured; when the tips of the workpiece to be measured are symmetrically distributed in the circumferential direction, a multipoint pulsation radial geometry measuring result is formed, and a pulsation multipoint measuring mode is realized.
2. The method for rapidly calibrating the associated reference telecentric measurement under the super-field scale according to claim 1, wherein the measuring camera adopts a 0.5-time high-resolution low-distortion telecentric lens; the background light source adopts a telecentric parallel light source and a blue light cold light source for illumination; the transmission target adopts a checkerboard pattern with a position accuracy of 1 μm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111344665.0A CN114049324B (en) | 2021-11-15 | 2021-11-15 | Rapid calibration method for correlated reference telecentric measurement under super-view field scale |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111344665.0A CN114049324B (en) | 2021-11-15 | 2021-11-15 | Rapid calibration method for correlated reference telecentric measurement under super-view field scale |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114049324A CN114049324A (en) | 2022-02-15 |
| CN114049324B true CN114049324B (en) | 2024-08-23 |
Family
ID=80208784
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111344665.0A Active CN114049324B (en) | 2021-11-15 | 2021-11-15 | Rapid calibration method for correlated reference telecentric measurement under super-view field scale |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114049324B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116485918B (en) * | 2023-06-25 | 2023-09-08 | 天府兴隆湖实验室 | Calibration method, calibration system and computer readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034238A (en) * | 2010-12-13 | 2011-04-27 | 西安交通大学 | Multi-camera system calibrating method based on optical imaging test head and visual graph structure |
| CN102062578A (en) * | 2010-12-13 | 2011-05-18 | 西安交通大学 | Handheld optical target for measuring visual coordinate and measuring method thereof |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103499302B (en) * | 2013-09-27 | 2015-11-25 | 吉林大学 | The camshaft diameter dimension On-line Measuring Method of structure based light Vision imaging system |
| CN103778640B (en) * | 2014-03-07 | 2017-09-22 | 中国工程物理研究院激光聚变研究中心 | A kind of object space telecentricity micro-vision system scaling method based on microballoon target |
| CN112432693B (en) * | 2020-10-22 | 2022-08-26 | 中国计量科学研究院 | Tracing method and device for machine vision low-frequency vibration measurement |
-
2021
- 2021-11-15 CN CN202111344665.0A patent/CN114049324B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102034238A (en) * | 2010-12-13 | 2011-04-27 | 西安交通大学 | Multi-camera system calibrating method based on optical imaging test head and visual graph structure |
| CN102062578A (en) * | 2010-12-13 | 2011-05-18 | 西安交通大学 | Handheld optical target for measuring visual coordinate and measuring method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114049324A (en) | 2022-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104034263B (en) | A kind of non-contact measurement method of forging's block dimension | |
| CN109870125B (en) | Hole-shaft coaxiality measuring device and method for hollow shaft | |
| Lu et al. | On-line measurement of the straightness of seamless steel pipes using machine vision technique | |
| CN104990515B (en) | Large-sized object three-dimensional shape measure system and its measuring method | |
| CN111366070B (en) | Multi-axis space coordinate system calibration method for combined type line laser measurement system | |
| CN111811395B (en) | Monocular vision-based dynamic plane pose measurement method | |
| CN108827187B (en) | A kind of measuring three-dimensional profile system | |
| CN112648934B (en) | Automatic elbow geometric form detection method | |
| CN102607457A (en) | Measuring device and measuring method for large three-dimensional morphology based on inertial navigation technology | |
| CN111964694A (en) | Laser range finder calibration method for three-dimensional measurement | |
| CN1971206A (en) | Calibration method for binocular vision sensor based on one-dimension target | |
| Liu et al. | Binocular-vision-based error detection system and identification method for PIGEs of rotary axis in five-axis machine tool | |
| CN111721198A (en) | Multi-axis space coordinate system calibration method for combined type line laser measurement system | |
| CN110686595A (en) | Laser beam space pose calibration method of non-orthogonal axis system laser total station | |
| CN110695982A (en) | Mechanical arm hand-eye calibration method and device based on three-dimensional vision | |
| CN113119129A (en) | Monocular distance measurement positioning method based on standard ball | |
| CN105046715A (en) | Space analytic geometry-based line-scan camera calibration method | |
| CN104036518B (en) | Camera calibration method based on vector method and three collinear points | |
| CN114049324B (en) | Rapid calibration method for correlated reference telecentric measurement under super-view field scale | |
| CN110260817B (en) | Complex surface deflection measurement self-positioning method based on virtual mark points | |
| Kumar et al. | An optical triangulation method for non-contact profile measurement | |
| Zou et al. | Extrinsic calibration method for 3D scanning system with four coplanar laser profilers | |
| Zheng et al. | Calibration of linear structured light system by planar checkerboard | |
| CN110458894B (en) | Calibration method for camera and contact type measuring head of measuring machine | |
| CN113409395B (en) | High-precision detection and positioning method for catheter end |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |