CN116183468A - Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration - Google Patents
Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration Download PDFInfo
- Publication number
- CN116183468A CN116183468A CN202211392957.6A CN202211392957A CN116183468A CN 116183468 A CN116183468 A CN 116183468A CN 202211392957 A CN202211392957 A CN 202211392957A CN 116183468 A CN116183468 A CN 116183468A
- Authority
- CN
- China
- Prior art keywords
- gob
- contour
- real
- points
- new image
- 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.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 239000007788 liquid Substances 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000012545 processing Methods 0.000 claims abstract description 6
- 230000000007 visual effect Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000005303 weighing Methods 0.000 claims description 8
- 238000012937 correction Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 7
- 238000004519 manufacturing process Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 238000003908 quality control method Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000010008 shearing Methods 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920003023 plastic Polymers 0.000 description 1
- 239000004033 plastic Substances 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
- G01N2015/1029—Particle size
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
- G01N2015/103—Particle shape
Landscapes
- Chemical & Material Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Dispersion Chemistry (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a real-time monitoring algorithm of liquid gob characteristics based on vision and single pixel calibration, which is characterized in that a group of images of gobs are respectively acquired in real time through cameras, contours are identified, the identified contours are scaled and vertically corrected, a three-dimensional model is built by adopting a fitting algorithm, the number of three-dimensional pixels is calculated, the calibration quality of single three-dimensional pixels is calculated, the real-time quality of gobs is calculated and monitored, and the like. The invention belongs to the technical field of liquid gob monitoring, in particular to an image processing and quality acquisition algorithm of liquid height Wen Liaodi, which is suitable for monitoring the shape, volume and quality characteristics of liquid gobs. The three-dimensional modeling and single-pixel mass calibration algorithm based on three-phase machine vision can acquire the quality of a gob in real time in the gob dropping process, can acquire the characteristics of the three-dimensional model, the quality and the like of each gob in real time, has high precision and accurate and reliable data, and provides data support for monitoring of liquid gobs.
Description
Technical Field
The invention belongs to the technical field of liquid gob monitoring, in particular to an image processing and quality acquisition algorithm of liquid height Wen Liaodi, which is suitable for monitoring the shape, volume and quality characteristics of liquid gobs, and particularly relates to a real-time monitoring algorithm of liquid gob characteristics based on vision and single-pixel calibration.
Background
In the manufacturing industries of steel, glass, plastics and the like, the quality control of products is realized by controlling the quality of solution gobs, the quality of the gobs is realized by the mechanical cooperation of a material passage punch or a pair of scissors, and the quality of the gobs is difficult to accurately obtain in the process. Thus, the quality control of the product is difficult to be accurate to the target value, and the quality control becomes an industry barrier. In addition, the liquid gob cannot be weighed and the like in the falling process, and only how to monitor the liquid gob by a non-contact indirect means can be considered.
A high-speed camera is one of industrial cameras, and generally refers to a digital industrial camera, which is generally installed on a machine pipeline to replace human eyes to make measurement and judgment, and is converted into an image signal by a digital image capturing object, and is transmitted to a special image processing system. High-speed camera visual analysis provides a solution for industry, sampling of hundreds of frames per second and high resolution, and can acquire a complete image of drop of each gob in real time. The gob is subject to tilting and surface irregularities during free fall due to the aerodynamic forces and initial shear stresses of the shears.
Disclosure of Invention
Problems to be solved by the invention:
aiming at the situation, in order to overcome the defects of the prior art, the invention aims to provide a three-dimensional modeling and single-pixel quality calibration algorithm based on three-phase machine vision, which can acquire the quality of a gob in real time in the gob dropping process; aiming at the situation that gobs can incline and surface irregularities due to the action of aeromechanics and initial shearing stress of scissors in the free falling process, the gobs are corrected through an algorithm, then three-dimensional modeling is carried out by adopting multiple cameras and multiple visual angles, more surface irregularities are obtained by improving the visual area, the accuracy of observation data and the algorithm is improved, and the calculated volume is comprehensively ensured not to be influenced by technical conditions.
The method comprises the steps of marking a gob, weighing the gob to obtain the volume and the mass of the gob, calculating the calibrated average mass of a single voxel, and taking the calculated calibrated average mass as a common variable for monitoring the gob, so that the mass of all gobs in the production process can be obtained in real time, and the calibrated average mass of the single voxel can be calibrated and corrected in time in the production process.
Technical means for solving the problems:
the invention provides a real-time monitoring algorithm of liquid gob characteristics based on vision and single-pixel calibration, which comprises the following steps:
firstly, respectively acquiring a group of images of a gob in real time through a camera, and carrying out contour recognition on the images to obtain a recognition contour;
step two, scaling and vertically correcting the identification profile in the step one; a group of three new images are obtained again;
thirdly, establishing a three-dimensional model by adopting a fitting algorithm according to the center lines formed by the three contour curves of the new image and the center points of the nth contour point of the three new image, and calculating the number of the three-dimensional pixels;
weighing the calibration gob;
step five, calculating the calibration quality of a single stereoscopic pixel according to the number and the quality of the calibration gob, wherein the number and the quality of the stereoscopic pixel are contained;
and step six, calculating and monitoring the real-time quality of the gob according to the volume of the gob measured later and the calibration quality of a single three-dimensional pixel.
Further, the cameras in the first step are three cameras, one group of images is three images, and the recognition contour is three recognition contours.
Preferably, the method of contour recognition in the first step is to extract the contour after binarizing the three images, specifically extracting the coordinate values of the contour points of the respective contour boundary lines.
As a further preferred aspect of the present invention, the coordinate value of the contour point is a row-column value of a continuous and closed boundary pixel string.
As a further preferred aspect of the present invention, the method of scaling and vertical correction described in the second step is to obtain tangent points at the top and bottom in each gob identification outline, and calculate the height value (i.e. the row number difference between the top tangent point and the bottom tangent point) and width value (the column number difference between the leftmost tangent point and the rightmost tangent point) of the gob shape, scale the other two images according to the height value of the first image, and retrieve a set of three new images, the new image being new image a, new image B, and new image C, respectively.
Further, the specific steps of establishing a stereoscopic model by using a fitting algorithm and calculating the number of the stereoscopic pixels in the third step include the following operations:
1) Mobile center hierarchical matching
In order to form a gob three-dimensional contour by using a contour vector framework of three new images with equal heights, carrying out two-dimensional center matching on six contour points with equal line numbers, wherein coordinate values of the contour points of the three new images corresponding to an nth line of pixel rows in respective images are as follows: in the nth row, two contour point coordinates in the new image a are (n, x) 1 )、(n,x 2 ) The center point is (n, x) (1,2) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points in the new image B are (n, x) 3 )、(n,x 4 ) The center point is (n, x) (3,4) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points of the new image C are (n, x) 5 )、(n,x 6 ) The center point is (n, x) (5,6) ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein n represents the number of rows and x represents the number of columns of contour points;
the three cameras are a camera A, a camera B and a camera C, the new image B and the new image C are translated to the center point of the new image A according to the coordinates of the center point, and simultaneously included angles of the camera B, the camera C and the camera A are counted into the two-dimensional coordinates of the new image point, wherein the included angle of the picture obtained by the camera A is alpha 1 =0; the included angle of the picture obtained by the camera B is alpha 2 The method comprises the steps of carrying out a first treatment on the surface of the The included angle of the picture obtained by the camera C is alpha 3 Then the two-dimensional section of the nth row is obtained, and two contour points in the new image A are expressed as follows: (n, 0, (x) 2 -x 1 )/2)、(n,180,(x 2 -x 1 ) 2); the two contour points in the new image B are expressed as: (n, alpha) 2 ,(x 4 -x 3 )/2)、(n,α 2 +180,(x 4 -x 3 ) 2); the two contour points in the new image C are expressed as: (n, alpha) 3 ,(x 6 -x 5 )/2)、(n,α 3 +180,(x 6 -x 5 ) 2); wherein alpha is 3 Not necessarily alpha 2 Is twice as many as when three cameras are respectively clamped by 120 degrees, alpha 3 ＝120、α 2 ＝60；
2) Error data is removed based on error detection removal method in double
Before the contour points of the new image in 1) are adopted to construct ellipses, the points with error data or larger errors need to be removed;
the rejecting method comprises the following steps:
(1) Calculating the distances from all points to the center point;
(2) Calculating the average value of all the distances;
(3) Calculating the deviation of the relative mean value of each distance, and calculating the standard deviation;
(4) Eliminating the points with the distance deviation larger than two times of standard deviation;
the standard deviation calculation formula is as follows:
sigma is the standard deviation, d is the distance from each point to the center point;
n represents the number of rows, which is the average value of the distance of each point from the center point;
3) Least squares fit ellipse parameters
The nth layer can be regarded as a two-dimensional plane, and six points and the center position are known to fit an ellipse;
curve equation:
Ax 2 +Bxy+Cy 2 +Dx+Ey+F＝0 (2)
the coordinates of six points are known to be the coordinate values of the contour points of three new images corresponding to the nth row of pixel rows in the respective images, six vector points pass through six plane points with angle calculation and are brought into a curve normal equation, and the objective function is as follows:
least square adjustment processing, namely minimizing F (A, B, C, D, E and F), and finally calculating the values of equation coefficients A, B, C, D, E and F;
according to the values of A, B, C, D, E and F, the ellipse parameters can be calculated according to an ellipse calculation formula:
where a is the length of the major half axis of the ellipse, b is the length of the minor half axis of the ellipse, x e Is a parameter, y e Is a parameter, and theta represents an included angle between a connecting line of an origin and a point on the ellipse and an x positive half axis or is called an elevation angle;
then, the area of the pixel row where the ellipse is located (i.e. the number of pixels contained) can be calculated:
s n ＝π*a n *b n (i＝0～N) (9)
wherein s is n Represents the area of the pixel row where the ellipse is located, pi is the circumference ratio, a n Length of major half axis of n-row ellipse b n The length of a short half shaft is n rows of ellipses;
4) Three-dimensional gob volume was calculated by stacking and accumulating layers by layer:
wherein V is the volume of the stereoscopic gob.
Preferably, the calculating the calibration quality of the single voxel in the fifth step includes the following steps:
firstly, weighing the calibrated gob product to obtain the mass G 0 The single voxel calibration quality is:
v 0 ＝G 0 /V i (11)
wherein v is 0 Calibrating the quality for a single stereoscopic pixel; v (V) i For each volume of stereo gob.
As a further preferred aspect of the invention, the calculation and monitoring of the real-time mass of the gobs described in step six, wherein the real-time mass of each gob is calculated as follows:
G j ＝V j *v 0 (12)
wherein G is j For the real-time mass of each gob, V j For the volume of the gob measured subsequently, v 0 The mass is calibrated for a single voxel.
Preferably, the three cameras are three cameras with the same resolution and intersecting any angle respectively.
By adopting the scheme, the beneficial effects obtained by the invention are as follows:
according to the invention, three cameras with the same resolution ratio and at any angle are adopted, falling gobs at the liquid melting furnace charge ports on the production line are photographed in real time, three identical images are matched, and the three identical images are subjected to contour recognition, scaling, vertical correction and coaxial stereo fitting, so that the number of stereo pixels contained in a stereo manner is obtained; meanwhile, after precisely weighing one marking material drop, calculating the marking quality of a single three-dimensional pixel by adopting the weight of the marking material drop and the number of the three-dimensional pixels; the quality of the subsequent gob can be calculated by using the marking quality of a single voxel and the number of the voxels in real time. The algorithm can realize real-time shape and quality acquisition and monitoring of all the gobs on the production line.
The invention provides a three-dimensional modeling and single-pixel quality calibration algorithm based on three-phase machine vision, which can acquire the quality of a gob in real time in the gob dropping process; aiming at the situation that gobs can incline and surface irregularities due to the action of aeromechanics and initial shearing stress of scissors in the free falling process, the gobs are corrected through an algorithm, then three-dimensional modeling is carried out by adopting multiple cameras and multiple visual angles, more surface irregularities are obtained by improving the visual area, the accuracy of observation data and the algorithm is improved, and the calculated volume is comprehensively ensured not to be influenced by technical conditions.
The method comprises the steps of marking a gob, weighing the gob to obtain the volume and the mass of the gob, calculating the calibrated average mass of a single voxel, and taking the calculated calibrated average mass as a common variable for monitoring the gob, so that the mass of all gobs in the production process can be obtained in real time, and the calibrated average mass of the single voxel can be calibrated and corrected in time in the production process.
The method and the device can acquire the characteristics of the three-dimensional model, the quality and the like of each gob in real time, are high in precision and accurate and reliable in data, and provide data support for monitoring of liquid gobs.
Drawings
FIG. 1 is a diagram of a data acquisition task for acquiring images with three cameras according to the present invention;
FIG. 2 is three images acquired by a camera during the falling of a liquid drop;
FIG. 3 is three new images after scaling and vertical correction;
FIG. 4 is a diagram showing the extracted contours after binarization;
FIG. 5 is a contour view of three images acquired by a camera during the falling of a liquid drop;
FIG. 6 is a profile of three new images after scaling and vertical correction;
FIG. 7 is a diagram showing six contour vector points and center point coordinate values corresponding to the nth row of pixels;
FIG. 8 is a center matching graph of the nth row of contour points;
FIG. 9 is a two-dimensional planar ellipsometric representation of a three-dimensional gob cross-section;
FIG. 10 is a perspective model fitted from a three-phase machine;
FIG. 11 is a graph of real-time quality monitoring of gob quality.
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention; all other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," "outer," and the like indicate orientation or positional relationships based on those shown in the drawings, merely to facilitate description of the invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus should not be construed as limiting the invention.
As shown in fig. 1 to 11, the invention provides a real-time monitoring algorithm of liquid gob characteristics based on vision and single pixel calibration, which comprises the following steps:
step one, referring to a data acquisition working diagram of images acquired by three cameras in fig. 1, respectively acquiring a group of images of a gob in real time by the cameras (referring to fig. 2 and 3), and carrying out contour recognition on the images to obtain a recognition contour;
step two, scaling and vertically correcting the identification profile in the step one; a group of three new images are obtained again;
thirdly, establishing a three-dimensional model by adopting a fitting algorithm according to the center lines formed by the three contour curves of the new image and the center points of the nth contour point of the three new image, and calculating the number of the three-dimensional pixels;
weighing the calibration gob;
step five, calculating the calibration quality of a single stereoscopic pixel according to the number and the quality of the calibration gob, wherein the number and the quality of the stereoscopic pixel are contained;
and step six, calculating and monitoring the real-time quality of the gob according to the volume of the gob measured later and the calibration quality of a single three-dimensional pixel.
As an embodiment of the present solution, as shown in fig. 1, the cameras in the step one are three cameras, and the three cameras are three cameras with the same resolution and intersecting any angle respectively. One group of images is three images; the recognition profiles are three recognition profiles.
As an embodiment of the present disclosure, as shown in fig. 4, the method of contour recognition described in the first step is to extract the contour after binarizing the three images, and as shown in fig. 5, specifically, extract the coordinate values of the contour points of the respective contour boundary lines.
Preferably, the coordinate value of the contour point is a row-column value of a continuous and closed boundary pixel string.
With the above embodiment in mind, as another embodiment of the present disclosure, as shown in fig. 5, the scaling and vertical correction method in the second step is to obtain tangent points at the top and bottom in each gob identification outline, and calculate the height value (i.e. the row number difference between the top tangent point and the bottom tangent point) and the width value (the column number difference between the leftmost tangent point and the rightmost tangent point) of the gob shape, and scale the other two images according to the height value of the first image, to obtain a set of three new images, i.e. new image a, new image B, and new image C, respectively.
The specific steps of establishing a stereoscopic model by adopting a fitting algorithm and calculating the number of the stereoscopic pixels in the third step include the following operations:
1) Mobile center hierarchical matching
In order to form a gob three-dimensional contour from a contour vector architecture of three new images with equal height, two-dimensional center matching is performed on six contour points with equal line numbers, as shown in fig. 7, coordinate values of the contour points of the three new images corresponding to the nth row of pixel rows in respective images are as follows: in the nth row, two contour point coordinates in the new image a are (n, x) 1 )、(n,x 2 ) The center point is (n, x) (1,2) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points in the new image B are (n, x) 3 )、(n,x 4 ) The center point is (n, x) (3,4) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points of the new image C are (n, x) 5 )、(n,x 6 ) The center point is (n, x) (5,6) ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein n represents the number of rows and x represents the number of columns of contour points;
the three cameras are a camera A, a camera B and a camera C, the new image B and the new image C are translated to the center point of the new image A according to the coordinates of the center point, and simultaneously included angles of the camera B, the camera C and the camera A are counted into the two-dimensional coordinates of the new image point, wherein the included angle of the picture obtained by the camera A is alpha 1 =0; the included angle of the picture obtained by the camera B is alpha 2 The method comprises the steps of carrying out a first treatment on the surface of the The included angle of the picture obtained by the camera C is alpha 3 Then a two-dimensional section of the nth row is obtained, as shown in fig. 8 (fig. 8 is a schematic diagram): the two contour points in the new image a are expressed as: (n, 0, (x) 2 -x 1 )/2)、(n,180,(x 2 -x 1 ) 2); the two contour points in the new image B are expressed as: (n, alpha) 2 ,(x 4 -x 3 )/2)、(n,α 2 +180,(x 4 -x 3 ) 2); the two contour points in the new image C are expressed as: (n, alpha) 3 ,(x 6 -x 5 )/2)、(n,α 3 +180,(x 6 -x 5 ) 2); wherein alpha is 3 Not necessarily alpha 2 Is twice as many as when three cameras are respectively clamped by 120 degrees, alpha 3 ＝120、α 2 ＝60。
2) Error data is removed based on error detection removal method in double
Before the contour points of the new image in 1) are adopted to construct ellipses, the points with error data or larger errors need to be removed;
the rejecting method comprises the following steps:
(1) Calculating the distances from all points to the center point;
(2) Calculating the average value of all the distances;
(3) Calculating the deviation of the relative mean value of each distance, and calculating the standard deviation;
(4) Eliminating the points with the distance deviation larger than two times of standard deviation;
the standard deviation calculation formula is as follows:
sigma is the standard deviation, d is the distance from each point to the center point;
n represents the number of rows, which is the average of the distances of each point from the center point.
3) Least squares fit ellipse parameters
The nth layer can be regarded as a two-dimensional plane, and six points and the center position are known to fit an ellipse;
curve equation:
Ax 2 +Bxy+Cy 2 +Dx+Ey+F＝0 (2)
the coordinates of six points are known to be the coordinate values of the contour points of three new images corresponding to the nth row of pixel rows in the respective images, six vector points pass through six plane points with angle calculation and are brought into a curve normal equation, and the objective function is as follows:
least square adjustment processing, namely minimizing F (A, B, C, D, E and F), and finally calculating the values of equation coefficients A, B, C, D, E and F;
from the values of a, B, C, D, E, F, the ellipse parameters can be calculated according to the ellipse calculation formula and the equation shown in fig. 9:
where a is the length of the major half axis of the ellipse, b is the length of the minor half axis of the ellipse, x e Is a parameter, y e Is a parameter, and theta represents an included angle between a connecting line of an origin and a point on the ellipse and an x positive half axis or is called an elevation angle;
then, the area of the pixel row where the ellipse is located (i.e. the number of pixels contained) can be calculated:
s n ＝π*a n *b n (i＝0～N) (9)
wherein s is n Represents the area of the pixel row where the ellipse is located, pi is the circumference ratio, a n Length of major half axis of n-row ellipse b n The length of a short half shaft is n rows of ellipses;
4) Three-dimensional gob volume was calculated by stacking and accumulating layers by layer:
wherein V is the volume of the stereoscopic gob.
As a preferred embodiment of the present solution, the calculating the calibration quality of the single voxel in the fifth step includes the following steps:
first, the fourth label is processedWeighing the fixed material drop product to obtain the mass G 0 The single voxel calibration quality is:
v 0 ＝G 0 /V i (11)
wherein v is 0 Calibrating the quality for a single stereoscopic pixel; v (V) i For each volume of stereo gob.
Calculating and monitoring the real-time mass of the gobs as described in step six, wherein the real-time mass of each gob is calculated as follows:
G j ＝V j *v 0 (12)
wherein G is j For the real-time mass of each gob, V j For the volume of the gob measured subsequently, v 0 The mass is calibrated for a single voxel.
The above is the overall operation flow of the present invention, and as an embodiment of the present invention, referring to the drawings, as shown in fig. 10, a three-phase machine fitting stereo model for gobs proposed by the present invention is shown; FIG. 11 is a graph of the present invention for real-time quality monitoring of gob quality.
The three-dimensional image three-dimensional calculation method is a brand new method in the industry, and no disclosure exists in the industry at present.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
The invention and its embodiments have been described above with no limitation, and the actual construction is not limited to the embodiments of the invention as shown in the drawings. In summary, if one of ordinary skill in the art is informed by this disclosure, a structural manner and an embodiment similar to the technical solution should not be creatively devised without departing from the gist of the present invention.
Claims (9)
1. The real-time monitoring algorithm for the characteristics of the liquid gob based on vision and single-pixel calibration is characterized by comprising the following steps of:
firstly, respectively acquiring a group of images of a gob in real time through a camera, and carrying out contour recognition on the images to obtain a recognition contour;
step two, scaling and vertically correcting the identification profile in the step one; a group of three new images are obtained again;
thirdly, establishing a three-dimensional model by adopting a fitting algorithm according to the center lines formed by the three contour curves of the new image and the center points of the nth contour point of the three new image, and calculating the number of the three-dimensional pixels;
weighing the calibration gob;
step five, calculating the calibration quality of a single stereoscopic pixel according to the number and the quality of the calibration gob, wherein the number and the quality of the stereoscopic pixel are contained;
and step six, calculating and monitoring the real-time quality of the gob according to the volume of the gob measured later and the calibration quality of a single three-dimensional pixel.
2. The real-time monitoring algorithm of liquid gob characteristics based on visual and single pixel calibration according to claim 1, wherein the cameras in the step one are three cameras, one group of images is three images, and the recognition contour is three recognition contours.
3. The real-time monitoring algorithm for liquid gob characteristics based on visual and single-pixel calibration according to claim 2, wherein the contour recognition method in the step one is to extract contours after binarizing three images, in particular to extract coordinate values of contour points of respective contour boundary lines.
4. A real time monitoring algorithm for liquid gob characteristics based on visual and single pixel calibration according to claim 3, wherein said contour point coordinate values are row and column values of a continuous and closed border pixel string.
5. The method for real-time monitoring of liquid gob characteristics based on visual and single-pixel calibration according to claim 4, wherein the scaling and vertical correction method in the second step is to obtain tangential tangent points of the top and bottom in each gob identification outline, calculate the height value and width value of the gob shape, scale the other two images according to the height value of the first image, and retrieve a set of three new images, the new image being new image a, new image B, and new image C, respectively.
6. The method for real-time monitoring of liquid gob characteristics based on visual and single pixel calibration according to claim 5, wherein the step three of creating a three-dimensional model by using a fitting algorithm and calculating the number of pixels comprising three-dimensional model comprises the following steps:
1) Mobile center hierarchical matching
In order to form a gob three-dimensional contour by using a contour vector framework of three new images with equal heights, carrying out two-dimensional center matching on six contour points with equal line numbers, wherein coordinate values of the contour points of the three new images corresponding to an nth line of pixel rows in respective images are as follows: in the nth row, two contour point coordinates in the new image a are (n, x) 1 )、(n，x 2 ) The center point is (n, x) (1，2) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points in the new image B are (n, x) 3 )、(n，x 4 ) The center point is (n, x) (3，4) ) The method comprises the steps of carrying out a first treatment on the surface of the The coordinates of two contour points of the new image C are (n, x) 5 )、(n，x 6 ) The center point is (n, x) (5，6) ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein n represents the number of rows and x represents the number of columns of contour points;
the three cameras are camera A, camera B and camera C, the new image B and the new image C are translated to the center point of the new image A according to the coordinates of the center point, and the cameras are used for simultaneouslyB. The included angle between the camera C and the camera A is counted in the two-dimensional coordinates of the new image point, wherein the included angle of the picture obtained by the camera A is alpha 1 =0; the included angle of the picture obtained by the camera B is alpha 2 The method comprises the steps of carrying out a first treatment on the surface of the The included angle of the picture obtained by the camera C is alpha 3 Then the two-dimensional section of the nth row is obtained, and two contour points in the new image A are expressed as follows: (n, 0, (x) 2 -x 1 )/2)、(n，180，(x 2 -x 1 ) 2); the two contour points in the new image B are expressed as: (n, alpha) 2 ，(x 4 -x 3 )/2)、(n，α 2 +180，(x 4 -x 3 ) 2); the two contour points in the new image C are expressed as: (n, alpha) 3 ，(x 6 -x 5 )/2)、(n，α 3 +180，(x 6 -x 5 ) 2); wherein alpha is 3 Not necessarily alpha 2 Is twice as many as when three cameras are respectively clamped by 120 degrees, alpha 3 ＝120、α 2 ＝60；
2) Error data is removed based on error detection removal method in double
Before the contour points of the new image in 1) are adopted to construct ellipses, the points with error data or larger errors need to be removed;
the rejecting method comprises the following steps:
(1) Calculating the distances from all points to the center point;
(2) Calculating the average value of all the distances;
(3) Calculating the deviation of the relative mean value of each distance, and calculating the standard deviation;
(4) Eliminating the points with the distance deviation larger than two times of standard deviation;
the standard deviation calculation formula is as follows:
sigma is the standard deviation, d is the distance from each point to the center point;
n represents the number of rows, which is the average value of the distance of each point from the center point;
3) Least squares fit ellipse parameters
The nth layer can be regarded as a two-dimensional plane, and six points and the center position are known to fit an ellipse;
curve equation:
Ax 2 +Bxy+Cy 2 +Dx+Ey+F＝0 (2)
the coordinates of six points are known to be the coordinate values of the contour points of three new images corresponding to the nth row of pixel rows in the respective images, six vector points pass through six plane points with angle calculation and are brought into a curve normal equation, and the objective function is as follows:
least square adjustment processing, namely minimizing F (A, B, C, D, E and F), and finally calculating the values of equation coefficients A, B, C, D, E and F;
according to the values of A, B, C, D, E and F, the ellipse parameters can be calculated according to an ellipse calculation formula:
where a is the length of the major half axis of the ellipse, b is the length of the minor half axis of the ellipse, x e Is a parameter, y e Is a parameter, and theta represents an included angle between a connecting line of an origin and a point on the ellipse and an x positive half axis or is called an elevation angle;
then, the area of the pixel row where the ellipse is located can be calculated:
s n ＝π*a n *b n (i＝0～N) (9)
wherein s is n Represents the area of the pixel row where the ellipse is located, pi is the circumference ratio, a n Length of major half axis of n-row ellipse b n The length of a short half shaft is n rows of ellipses;
4) Three-dimensional gob volume was calculated by stacking and accumulating layers by layer:
wherein V is the volume of the stereoscopic gob.
7. The real-time monitoring algorithm for liquid gob characteristics based on visual and single pixel calibration of claim 6, wherein said calculating the calibration quality of a single voxel in step five comprises the steps of:
firstly, weighing the calibrated gob product to obtain the mass G 0 The single voxel calibration quality is:
v 0 ＝G 0 /V i (11)
wherein v is 0 Calibrating the quality for a single stereoscopic pixel; v (V) i For each volume of stereo gob.
8. The real-time monitoring algorithm for liquid gob characteristics based on visual and single pixel calibration of claim 7, wherein said calculating and monitoring the real-time mass of the gob in step six is performed as follows:
G j ＝V j *v 0 (12)
wherein G is j For the real-time mass of each gob, V j For the volume of the gob measured subsequently, v 0 The mass is calibrated for a single voxel.
9. The real-time monitoring algorithm of liquid gob characteristics based on visual and single pixel calibration according to claim 8, wherein the three cameras are three cameras with the same resolution crossing any angle respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211392957.6A CN116183468A (en) | 2022-11-08 | 2022-11-08 | Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211392957.6A CN116183468A (en) | 2022-11-08 | 2022-11-08 | Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116183468A true CN116183468A (en) | 2023-05-30 |
Family
ID=86449575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211392957.6A Pending CN116183468A (en) | 2022-11-08 | 2022-11-08 | Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116183468A (en) |
-
2022
- 2022-11-08 CN CN202211392957.6A patent/CN116183468A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7175312B2 (en) | Method and apparatus for controlling molding processes | |
| CN103471531B (en) | The online non-contact measurement method of axial workpiece linearity | |
| CN102095733B (en) | Machine vision-based intelligent detection method for surface defect of bottle cap | |
| CN109190628A (en) | A kind of plate camber detection method based on machine vision | |
| CN109520436B (en) | Butterfly spring three-dimensional size automatic measurement system based on machine vision and measurement method thereof | |
| CN108613630B (en) | Two-wire tube level bubble offset measurement method based on image processing technology | |
| CN105764860A (en) | Methods for determining shape of substantially cylindrical specular reflective surface | |
| CN102441581A (en) | Machine vision-based device and method for online detection of structural steel section size | |
| CN108534802B (en) | Tube level bubble deviation measurement method based on image processing technology | |
| CN206981462U (en) | Stamping parts surface defect detection apparatus based on 3D vision | |
| CN102589516A (en) | Dynamic distance measuring system based on binocular line scan cameras | |
| CN106056603A (en) | Stereoscopic vision-based welding execution parameter on-line detection method | |
| CN102798349A (en) | Three-dimensional surface extraction method based on equal-gray line search | |
| CN111421001A (en) | Online accurate centering system of high-speed wire rod mill and centering method thereof | |
| CN117649404A (en) | Medicine packaging box quality detection method and system based on image data analysis | |
| CN114383505A (en) | Automatic detection device for dimension of short shaft type part | |
| IL263097B2 (en) | Optimizing a set-up stage in an automatic visual inspection process | |
| CN112893172A (en) | Gasket size detection system and method based on machine vision, processing terminal and medium | |
| CN116183468A (en) | Real-time monitoring algorithm for liquid gob characteristics based on vision and single-pixel calibration | |
| CN108053485B (en) | Potato three-dimensional modeling method based on contour image | |
| CN110586491B (en) | Method, device and storage medium for detecting posture and quality of pomelo | |
| CN114964024B (en) | Steel pipe measuring device | |
| CN107578417B (en) | Liquid drop volume real-time tracking measurement method based on optical profile curve extraction | |
| CN116091617A (en) | Tire pose detection method and system based on machine vision | |
| CN114581385A (en) | Welding seam defect area mapping algorithm based on circle positioning |
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 |