CN115770731A - Method and system for eliminating bad workpieces based on laser vision - Google Patents
Method and system for eliminating bad workpieces based on laser vision Download PDFInfo
- Publication number
- CN115770731A CN115770731A CN202211420231.9A CN202211420231A CN115770731A CN 115770731 A CN115770731 A CN 115770731A CN 202211420231 A CN202211420231 A CN 202211420231A CN 115770731 A CN115770731 A CN 115770731A
- Authority
- CN
- China
- Prior art keywords
- workpiece
- detected
- workpieces
- laser
- flatness
- 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
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a method and a system for removing bad workpieces based on laser vision, wherein the method comprises the following steps: carrying out online laser visual detection on a plurality of detected workpieces conveyed along a calibration plate, and acquiring flatness information of each detected workpiece in the plurality of detected workpieces; marking the tested workpiece according to the flatness information of each tested workpiece to generate two-dimensional code information of the tested workpiece; the removing mechanism identifies the two-dimensional code information of each detected workpiece and judges whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information; and when the tested workpiece is judged to meet the rejection operation, rejecting the tested workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate. According to the invention, a laser scanning triangulation method is adopted to carry out 3D image data acquisition and spatial model construction, a detected workpiece which meets rejection is found, and an automatic rejection task is completed, so that the reject ratio of a product can be finely controlled.
Description
Technical Field
The invention relates to the technical field of industrial control, in particular to a method and a system for removing bad workpieces based on laser vision.
Background
The flatness measurement of plane workpieces usually adopts feeler gauge measurement, height gauge measurement, simple visual measurement and the like, the measurement mode has the problems of low precision or low detection efficiency and the like, the modern industrial production such as 3C shells, automobile parts, furniture products and the like needs large-batch rapid high-precision non-contact measurement, the current non-contact measurement precision is low, the products cannot be rapidly applied to the industrial automatic control process, the defective product rejection of the plane workpieces can also cause a large amount of non-controllability, and the rejection work of defective products is particularly important.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a method and a system for rejecting bad workpieces based on laser vision.
In order to solve the above problems, the present invention provides a method for removing defective workpieces based on laser vision, comprising the following steps:
carrying out online laser visual detection on a plurality of detected workpieces conveyed along a calibration plate, and acquiring flatness information of each detected workpiece in the plurality of detected workpieces;
marking identification processing is carried out on the tested workpieces according to the flatness information of each tested workpiece, and two-dimensional code information of the tested workpieces is generated, wherein the two-dimensional code information comprises a rejection instruction of each tested workpiece;
controlling a plurality of tested workpieces subjected to code marking processing to move towards the removing mechanism;
the removing mechanism identifies the two-dimensional code information of each detected workpiece and judges whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information;
and when the tested workpiece is judged to meet the rejection operation, rejecting the tested workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
The online laser visual inspection of a plurality of workpieces to be tested conveyed along a calibration plate comprises the following steps:
and when recognizing that the workpiece to be measured enters the measuring area, starting a laser to generate line laser to irradiate the workpiece to be measured, and triggering a 3D camera to acquire 3D image data of the workpiece to be measured.
The obtaining flatness information of each of the plurality of measured workpieces comprises:
carrying out surface data plane fitting on the 3D image data of the workpiece to be measured by a plane fitting method based on robust least squares to obtain workpiece surface data after plane fitting;
processing the fitted workpiece surface data based on an interested region screening operator to obtain concave-convex points on the workpiece surface;
and comparing the concave-convex point on the surface of the workpiece with the reference plane to calculate the flatness of the measured workpiece.
The method for performing surface data plane fitting on the 3D image data of the workpiece to be measured by the plane fitting method based on the robust least squares to obtain the workpiece surface data after the plane fitting comprises the following steps:
carrying out interference data point removing processing on point cloud data in the 3D image data;
substituting the point cloud data coordinate subjected to interference data point removal processing into an error equation;
and performing iterative computation fitting on the error equation by using an robust nonlinear least square principle to obtain a reference plane equation.
The step of processing the fitted workpiece surface data to obtain concave and convex points on the workpiece surface by the region-of-interest-based screening operator comprises the following steps:
performing median filtering on the image, filtering out salt and pepper noise, performing threshold segmentation processing, and performing opening operation, closing operation and corrosion processing;
and obtaining a stable concave-convex area by using the region-of-interest screening operator, and sorting the concave-convex area to screen out the highest point and the lowest point.
The step of comparing the concave-convex points on the surface of the workpiece with the reference plane to calculate the flatness of the measured workpiece comprises the following steps:
calculating the deviation value of the wave crest of the in-plane convex area from the reference plane;
calculating the deviation value of the wave trough of the plane concave region from the reference plane;
and calculating the mean square root value of the square sum of the deviation of each measuring point from the reference plane, and reflecting the deviation degree of each measuring point from the ideal plane.
The online laser visual inspection of a plurality of workpieces to be tested conveyed along a calibration plate comprises the following steps:
and controlling the feeding time interval of the plurality of measured workpieces, and controlling the conveyor belt to convey the plurality of measured workpieces according to the scanning speed required by the 3D camera.
The marking identification processing is carried out on the tested workpiece according to the flatness of each tested workpiece, and the generation of the two-dimensional code information of the tested workpiece comprises the following steps:
when the detected workpiece is identified to enter the coding identification area based on the position sensor, acquiring flatness information of the detected workpiece;
generating two-dimensional code information for the workpiece to be detected according to the flatness information of the workpiece to be detected;
and coding and identifying the two-dimensional code information on the surface of the workpiece to be detected.
Correspondingly, the invention also provides a system for removing the bad workpieces based on the laser vision, which comprises:
the laser vision module is used for carrying out online laser vision detection on a plurality of detected workpieces conveyed along the calibration plate and acquiring flatness information of each detected workpiece in the plurality of detected workpieces;
the code marking module is used for carrying out code marking processing on the detected workpiece according to the flatness information of each detected workpiece to generate two-dimensional code information of the detected workpiece, wherein the two-dimensional code information comprises a rejection instruction of each detected workpiece;
the motion control module is used for controlling the plurality of tested workpieces subjected to the coding identification processing to move towards the removing mechanism;
the removing mechanism is used for identifying the two-dimensional code information of each detected workpiece and judging whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information; and when the tested workpiece is judged to meet the rejection operation, rejecting the tested workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
When recognizing that the workpiece to be measured enters the measuring area, the laser vision module starts a laser to generate line laser to irradiate the workpiece to be measured and triggers the 3D camera to acquire 3D image data of the workpiece to be measured.
According to the invention, the transmission speed of the workpiece to be detected is controlled to be suitable for the scanning process of the 3D camera, the corresponding relation between the workpiece to be detected and the measuring area is detected through the position sensor, so that the line laser is started to generate laser to trigger the 3D camera to acquire 3D image data, and through the control mode, the line laser and the 3D camera can accurately realize the data acquisition process of the workpiece to be detected, so that the energy consumption is reduced, and the accuracy of the control process is improved. The method comprises the steps of carrying out data processing based on 3D image data, carrying out 3D image data acquisition and spatial model construction by adopting a laser scanning triangulation method, fitting the surface of a workpiece into a spatial plane by applying a least square method and a new visual algorithm, and calculating the surface flatness of the workpiece by the distance from a convex-concave point to the plane. The flatness of the surface of the workpiece is identified by adopting a laser scanning triangulation method to carry out 3D image data acquisition and spatial model construction, the workpiece to be detected which is in accordance with rejection is found, and an automatic rejection task is completed, so that the reject ratio of the product can be finely controlled, and the industrial quality detection scale is accelerated.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic diagram of a system for removing defective workpieces based on laser vision according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a laser vision module in an example of the invention;
fig. 3 is a flowchart of a method for rejecting defective workpieces based on laser vision in an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic structural diagram of a system for rejecting a defective workpiece based on laser vision in an embodiment of the present invention, where the system includes:
the laser vision module is used for carrying out online laser vision detection on a plurality of detected workpieces conveyed along the calibration plate and acquiring flatness information of each detected workpiece in the plurality of detected workpieces;
the code marking module is used for carrying out code marking processing on the detected workpiece according to the flatness information of each detected workpiece to generate two-dimensional code information of the detected workpiece, wherein the two-dimensional code information comprises a rejection instruction of each detected workpiece;
the motion control module is used for controlling the plurality of detected workpieces subjected to the coding identification processing to move towards the removing mechanism;
the removing mechanism is used for identifying the two-dimensional code information of each detected workpiece and judging whether each detected workpiece is subjected to removing operation or not according to a removing instruction in the two-dimensional code information; and when the detected workpiece is judged to meet the rejection operation, rejecting the detected workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
When recognizing that the workpiece to be measured enters the measurement area, the laser vision module starts a laser to generate line laser to irradiate the workpiece to be measured and triggers a 3D camera to acquire 3D image data of the workpiece to be measured.
Specifically, fig. 2 shows a schematic structural diagram of a laser vision module in an embodiment of the present invention, where the laser vision module includes:
the control unit is used for starting a laser to generate line laser to irradiate the workpiece to be measured and triggering the 3D camera to acquire 3D image data of the workpiece to be measured when the fact that the workpiece to be measured enters the measurement area is identified; performing data processing on the 3D image data to acquire flatness information of each of a plurality of measured workpieces;
the 3D camera is used for acquiring 3D image data of the workpiece to be detected;
and the laser is used for controlling the control unit to start the linear laser to irradiate the workpiece to be detected.
The control unit is used for carrying out surface data plane fitting on the 3D image data of the workpiece to be measured based on a plane fitting method of robust least squares to obtain workpiece surface data after plane fitting; processing the fitted workpiece surface data based on the region-of-interest screening operator to obtain concave and convex points on the workpiece surface; and comparing the concave-convex points on the surface of the workpiece with the reference plane to calculate the flatness of the workpiece to be measured.
The control unit is also used for controlling the feeding time interval of the plurality of measured workpieces and controlling the conveyor belt to convey the plurality of measured workpieces according to the speed required by the scanning of the 3D camera.
This control unit makes its scanning process that can be adapted to the 3D camera through the transfer rate of control quilt survey work piece, detect the corresponding relation of quilt survey work piece and measurement area through position sensor, thereby open line laser instrument and produce laser and trigger 3D camera and gather 3D image data, through this kind of control mode, make line laser and 3D camera can accurate realization to the data acquisition process of quilt survey work piece, reduce the energy consumption, promote control process's precision. The method comprises the steps of carrying out data processing based on 3D image data, carrying out 3D image data acquisition and spatial model construction by adopting a laser scanning triangulation method, fitting the surface of a workpiece into a spatial plane by applying a least square method and a new visual algorithm, and calculating the surface flatness of the workpiece by the distance from a convex-concave point to the plane.
The system provided by the embodiment of the invention can adapt to the scanning process of the 3D camera by controlling the transmission speed of the workpiece to be detected, and the corresponding relation between the workpiece to be detected and the measurement area is detected by the position sensor, so that the line laser is started to generate laser to trigger the 3D camera to acquire 3D image data. The method comprises the steps of carrying out data processing based on 3D image data, carrying out 3D image data acquisition and space model construction by adopting a laser scanning triangulation method, fitting the surface of a workpiece into a space plane by applying a least square method and a new visual algorithm, and calculating the surface flatness of the workpiece by the distance from a convex-concave point to the plane. The flatness of the surface of the workpiece is identified by adopting a laser scanning triangulation method to carry out 3D image data acquisition and spatial model construction, the workpiece to be detected which is in accordance with rejection is found, and an automatic rejection task is completed, so that the reject ratio of the product can be finely controlled, and the industrial quality detection scale is accelerated.
The method for eliminating the bad workpieces based on the laser vision, which is related by the embodiment of the invention, carries out online laser vision detection on a plurality of detected workpieces conveyed along a calibration plate, and acquires the flatness information of each detected workpiece in the plurality of detected workpieces; according to the flatness information of each detected workpiece, carrying out coding identification processing on the detected workpiece to generate two-dimensional code information of the detected workpiece, wherein the two-dimensional code information comprises a rejection instruction of each detected workpiece; controlling a plurality of tested workpieces subjected to code marking processing to move towards the removing mechanism; the removing mechanism identifies the two-dimensional code information of each detected workpiece and judges whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information; and when the detected workpiece is judged to meet the rejection operation, rejecting the detected workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
Specifically, fig. 3 shows a flowchart of a method for rejecting a defective workpiece based on laser vision in an embodiment of the present invention, where the method includes:
s301, carrying out online laser visual detection on a plurality of detected workpieces conveyed along a calibration plate;
in the specific implementation process, the feeding time interval of a plurality of tested workpieces is controlled, and the conveyor belt is controlled to convey the plurality of tested workpieces according to the scanning speed required by the 3D camera.
When the fact that the workpiece to be measured enters the measuring area is recognized, the laser is started to generate line laser to irradiate the workpiece to be measured, and the 3D camera is triggered to collect 3D image data of the workpiece to be measured, so that flatness information of each of the plurality of workpieces to be measured can be obtained through S302-S304.
The 3D camera collects 3D image data of the workpiece to be detected as surface workpiece data of the line laser acting on the workpiece to be detected. The line laser is matched with the 3D camera, the triangulation principle is adopted, laser line information projected on the surface of an object by the laser generator is captured through the image sensor, the surface contour information of the object can be reconstructed, and the characteristic measurement of a planar object with high precision requirement can be realized.
The speed of the conveyor belt needs to be controlled, when the measured workpiece reaches a measurement area, if the speed of the conveyor belt is too high, the image collected by the 3D camera is easy to distort, the measured workpiece needs to be controlled to transmit the measured workpiece at a reasonable speed, and therefore the image collected by the 3D camera can be ensured to accord with workpiece surface flatness identification processing.
S302, performing surface data plane fitting on the 3D image data of the workpiece to be measured by a plane fitting method based on robust least squares to obtain workpiece surface data subjected to plane fitting;
it should be noted that, here, performing surface data plane fitting on the 3D image data of the workpiece to be measured by the plane fitting method based on robust least squares to obtain the workpiece surface data after plane fitting includes: carrying out interference data point removing processing on point cloud data in the 3D image data; substituting the point cloud data coordinate subjected to interference data point removal processing into an error equation; and performing iterative calculation fitting on the error equation by using an robust nonlinear least square principle to obtain a reference plane equation.
Compared with the traditional least square method, the robust estimation least square method Huber method is adopted, the influence of outliers can be reduced, the equivalence weight processing can be carried out on the measurement, and when the measured data have errors, the Huber method can obtain a more accurate estimation result. The accuracy of plane fitting can directly influence the accuracy of workpiece surface flatness detection, so the accurate plane fitting is the key point of workpiece surface quality detection, the point cloud data is firstly processed by removing interference data points and the like before the plane fitting is carried out, and the workpiece surface data is fitted after the data processing is finished.
Here, the plane model is set as:
F(x,y,z)=a*x+b*y+c*z+d;
in the formula, (x, y, z) is a data point obtained by measurement in an experiment, and a, b, c and d are parameters to be fitted.
In order to make the data points fall on a plane, i.e. satisfy the equation, an error equation is constructed:
wherein M is the total amount of the collected point cloud data, (x) m ,y m ,z m ) And the coordinates of the collected plane point cloud data are obtained. By selecting an iteration initial value x =[a0,b0,c0,d0]And substituting the processed point cloud data coordinates, performing iterative calculation by using an robust nonlinear least square principle, and finally fitting to obtain a reference plane equation.
S303, processing the fitted workpiece surface data based on an interested region screening operator to obtain concave and convex points on the workpiece surface;
the step of processing the fitted workpiece surface data to obtain concave and convex points on the workpiece surface by the region-of-interest-based screening operator comprises the following steps: performing median filtering on the image, filtering out salt and pepper noise, performing threshold segmentation processing, and performing opening operation, closing operation and corrosion processing; and obtaining a stable concave-convex area by using the region of interest screening operator, and sequencing the concave-convex area to screen out the highest point and the lowest point.
Here by finding the surface asperities of the workpiece. And carrying out median filtering on the image, filtering out salt-pepper noise, carrying out threshold segmentation processing, carrying out opening operation, closing operation and corrosion processing, and further removing interference. And (4) obtaining a stable concave-convex area by using the region-of-interest screening operator, and sorting the concave-convex area to screen out the highest point and the lowest point.
In the specific implementation process, an interested area in the surface data of the workpiece is obtained firstly, then the outline of the interested area is extracted, and the interested area of a rectangle can be intercepted by utilizing a rectangle intercepting function; the image of the region of interest is then subjected to enhancement processing.
In the specific implementation process, the gray gradient of the workpiece surface data of the region of interest is searched, and Sobel operators are used for processing the workpiece surface data of the region of interest; searching a local maximum value in the gradient direction by using a non-maximum value suppression algorithm, suppressing non-maximum value elements and thinning edges; and extracting and connecting edges by adopting a dual-threshold algorithm to obtain a stable concave-convex area.
S304, comparing the concave-convex points on the surface of the workpiece with a reference plane to calculate the flatness of the workpiece to be measured;
the step of comparing the concave-convex points on the surface of the workpiece with the reference plane to calculate the flatness of the measured workpiece comprises the following steps: calculating the deviation value of the wave crest of the in-plane convex area from the reference plane; calculating a deviation value of a wave trough of the plane concave area from a reference plane; and calculating the mean square root value of the sum of squares of the deviation of each measuring point from the reference plane, and reflecting the deviation degree of each measuring point from the ideal plane.
Specifically, the flatness of the workpiece to be measured is obtained by comparing the screened peak-valley value of the concave-convex part with a reference plane, wherein: (1) D1, a peak-base flatness error index, a deviation value of a wave crest of the in-plane convex region from a base plane;
(2) D2, a valley-base flatness error index, a deviation value of a valley of a plane concave area from a base plane;
(3) The root mean square flatness error index and the mean square root value of the sum of squares of the deviation of each measuring point from the reference plane reflect the deviation degree of each measuring point from the ideal plane.
RMSE=sqrt((D1^2+D2^2+……+Dn^2)/n);
Calculating a variance formula: avg = (D1 + D2+ D3)/3;
Var=((D1-Avg)^2+(D2-Avg)^2+(D3-Avg)^2)/3。
s305, performing coding identification processing on the workpiece to be detected according to the flatness information of each workpiece to be detected to generate two-dimensional code information of the workpiece to be detected;
it should be noted that the two-dimensional code information includes a rejection instruction of each detected workpiece.
The marking identification processing is carried out on the tested workpiece according to the flatness of each tested workpiece, and the generation of the two-dimensional code information of the tested workpiece comprises the following steps: when the detected workpiece is identified to enter the coding identification area based on the position sensor, acquiring flatness information of the detected workpiece; generating two-dimensional code information for the workpiece to be detected according to the flatness information of the workpiece to be detected; and coding and identifying the two-dimensional code information on the surface of the workpiece to be detected.
The two-dimension code information can establish a two-dimension code generating structure through a tested workpiece; generating a two-dimensional code generation page according to the two-dimensional code generation structure; the two-dimensional code generation page comprises a set number of page parts, and the page parts are used for displaying two-dimensional code associated information; the two-dimensional code generation page is provided with a related data receiving component; the data receiving component is used for establishing a pushing event link between the coding identification module and the rejecting mechanism; the pushing event link is used for acquiring the two-dimension code state information pushed by the eliminating mechanism in real time; the two-dimension code state information is used for switching the state of the two-dimension code by the code printing identification module; and generating each page component of the two-dimensional code generation page and the data receiving component, wherein the data receiving component does not display the two-dimensional code generation page.
S306, controlling the plurality of tested workpieces subjected to the coding identification processing to move towards the removing mechanism;
s307, the rejection mechanism identifies the two-dimensional code information of each detected workpiece, and judges whether each detected workpiece carries out rejection operation or not according to rejection instructions in the two-dimensional code information;
s308, when the tested workpiece is judged to meet the rejection operation, the tested workpiece which does not meet the flatness requirement is rejected from the conveyor belt of the calibration plate;
and S309, conveying the tested workpiece meeting the flatness requirement to a specified area.
The method provided by the embodiment of the invention can be suitable for the scanning process of the 3D camera by controlling the transmission speed of the workpiece to be detected, and the corresponding relation between the workpiece to be detected and the measurement area is detected by the position sensor, so that the line laser is started to generate laser to trigger the 3D camera to acquire 3D image data. The method comprises the steps of carrying out data processing based on 3D image data, carrying out 3D image data acquisition and space model construction by adopting a laser scanning triangulation method, fitting the surface of a workpiece into a space plane by applying a least square method and a new visual algorithm, and calculating the surface flatness of the workpiece by the distance from a convex-concave point to the plane. The flatness of the surface of the workpiece is identified by adopting a laser scanning triangulation method to carry out 3D image data acquisition and spatial model construction, the workpiece to be detected which is in accordance with rejection is found, and an automatic rejection task is completed, so that the reject ratio of the product can be finely controlled, and the industrial quality detection scale is accelerated.
The above embodiments of the present invention are described in detail, and the principle and the implementation of the present invention are described herein by using specific embodiments, and the description of the above embodiments is only used to help understanding the method of the present invention and the core idea thereof; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A method for rejecting bad workpieces based on laser vision is characterized by comprising the following steps:
carrying out online laser visual detection on a plurality of detected workpieces conveyed along a calibration plate, and acquiring flatness information of each detected workpiece in the plurality of detected workpieces;
according to the flatness information of each detected workpiece, carrying out coding identification processing on the detected workpiece to generate two-dimensional code information of the detected workpiece, wherein the two-dimensional code information comprises a rejection instruction of each detected workpiece;
controlling a plurality of detected workpieces processed by the code marking identification to move towards the removing mechanism;
the removing mechanism identifies the two-dimensional code information of each detected workpiece and judges whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information;
and when the tested workpiece is judged to meet the rejection operation, rejecting the tested workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
2. The method for laser-based visual rejection of defective workpieces as claimed in claim 1, wherein said on-line laser visual inspection of a plurality of workpieces under test conveyed along a calibration plate comprises:
and when recognizing that the workpiece to be measured enters the measuring area, starting a laser to generate line laser to irradiate the workpiece to be measured, and triggering a 3D camera to acquire 3D image data of the workpiece to be measured.
3. The method for eliminating the bad workpieces based on the laser vision as claimed in claim 2, wherein said obtaining the flatness information of each of a plurality of tested workpieces comprises:
performing surface data plane fitting on the 3D image data of the workpiece to be measured by a plane fitting method based on robust least squares to obtain workpiece surface data after plane fitting;
processing the fitted workpiece surface data based on the region-of-interest screening operator to obtain concave and convex points on the workpiece surface;
and comparing the concave-convex points on the surface of the workpiece with the reference plane to calculate the flatness of the workpiece to be measured.
4. The method for rejecting the bad workpiece based on the laser vision as claimed in claim 3, wherein the performing the surface data plane fitting on the 3D image data of the workpiece to be tested by the robust least squares based plane fitting method to obtain the surface data of the workpiece after the plane fitting comprises:
carrying out interference data point removing processing on point cloud data in the 3D image data;
substituting the point cloud data coordinate subjected to interference data point removal processing into an error equation;
and performing iterative computation fitting on the error equation by using an robust nonlinear least square principle to obtain a reference plane equation.
5. The method for removing the bad workpieces based on the laser vision as claimed in claim 3, wherein the processing the fitted workpiece surface data to obtain the concave-convex points on the workpiece surface by the region-of-interest screening operator comprises:
performing median filtering on the image, filtering out salt and pepper noise, performing threshold segmentation processing, and performing opening operation, closing operation and corrosion processing;
and obtaining a stable concave-convex area by using the region of interest screening operator, and sequencing the concave-convex area to screen out the highest point and the lowest point.
6. The method for rejecting the bad workpieces based on the laser vision as claimed in claim 3, wherein the comparing the concave-convex points on the surface of the workpiece with the reference plane to calculate the flatness of the measured workpiece comprises:
calculating the deviation value of the wave crest of the in-plane convex area from the reference plane;
calculating a deviation value of a wave trough of the plane concave area from a reference plane;
and calculating the mean square root value of the square sum of the deviation of each measuring point from the reference plane, and reflecting the deviation degree of each measuring point from the ideal plane.
7. The method for rejecting defective workpieces based on laser vision as claimed in any one of claims 1 to 6, wherein the online laser vision inspection of the plurality of workpieces to be tested conveyed along the calibration plate comprises:
and controlling the feeding time interval of the plurality of measured workpieces, and controlling the conveyor belt to convey the plurality of measured workpieces according to the scanning speed required by the 3D camera.
8. The method for eliminating the bad workpieces based on the laser vision as claimed in claim 6, wherein the step of marking the workpiece to be detected according to the flatness of each workpiece to be detected comprises the following steps:
when the detected workpiece is identified to enter the coding identification area based on the position sensor, acquiring flatness information of the detected workpiece;
generating two-dimensional code information for the workpiece to be detected according to the flatness information of the workpiece to be detected;
and coding and identifying the two-dimensional code information on the surface of the workpiece to be detected.
9. A system for rejecting defective workpieces based on laser vision, the system comprising:
the laser vision module is used for carrying out online laser vision detection on a plurality of detected workpieces conveyed along the calibration plate and acquiring flatness information of each detected workpiece in the plurality of detected workpieces;
the code marking module is used for carrying out code marking processing on the detected workpiece according to the flatness information of each detected workpiece to generate two-dimensional code information of the detected workpiece, wherein the two-dimensional code information comprises a rejection instruction of each detected workpiece;
the motion control module is used for controlling the plurality of detected workpieces subjected to the coding identification processing to move towards the removing mechanism;
the removing mechanism is used for identifying the two-dimensional code information of each detected workpiece and judging whether each detected workpiece is subjected to removing operation or not according to removing instructions in the two-dimensional code information; and when the detected workpiece is judged to meet the rejection operation, rejecting the detected workpiece which does not meet the flatness requirement from the conveyor belt of the calibration plate.
10. The system for eliminating the defective workpieces based on the laser vision as claimed in claim 9, wherein the laser vision module starts the laser to generate line laser to irradiate the workpiece to be detected and triggers the 3D camera to acquire the 3D image data of the workpiece to be detected when recognizing that the workpiece to be detected enters the measurement area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211420231.9A CN115770731A (en) | 2022-11-14 | 2022-11-14 | Method and system for eliminating bad workpieces based on laser vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211420231.9A CN115770731A (en) | 2022-11-14 | 2022-11-14 | Method and system for eliminating bad workpieces based on laser vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115770731A true CN115770731A (en) | 2023-03-10 |
Family
ID=85389064
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211420231.9A Pending CN115770731A (en) | 2022-11-14 | 2022-11-14 | Method and system for eliminating bad workpieces based on laser vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115770731A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117358622A (en) * | 2023-12-08 | 2024-01-09 | 格力大松(宿迁)生活电器有限公司 | Product detection method, device and system |
-
2022
- 2022-11-14 CN CN202211420231.9A patent/CN115770731A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117358622A (en) * | 2023-12-08 | 2024-01-09 | 格力大松(宿迁)生活电器有限公司 | Product detection method, device and system |
CN117358622B (en) * | 2023-12-08 | 2024-04-16 | 格力大松(宿迁)生活电器有限公司 | Method, device and system for detecting indoor and outdoor units of air conditioner |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105066915B (en) | Mould camber processing error and surface roughness On-machine Test device and detection method | |
CN109949303B (en) | Workpiece shape detection method and device | |
Medina et al. | Automated visual classification of frequent defects in flat steel coils | |
Bernal et al. | Performance evaluation of optical scanner based on blue LED structured light | |
CN110260818B (en) | Electronic connector robust detection method based on binocular vision | |
von Enzberg et al. | A multiresolution approach to model-based 3-D surface quality inspection | |
CN110146017A (en) | Industrial robot repetitive positioning accuracy measurement method | |
CN115770731A (en) | Method and system for eliminating bad workpieces based on laser vision | |
CN114092411A (en) | Efficient and rapid binocular 3D point cloud welding spot defect detection method | |
CN115752300A (en) | Method and system for detecting flatness of planar workpiece through online laser vision | |
CN111738971A (en) | Circuit board stereo scanning detection method based on line laser binocular stereo vision | |
Allard et al. | Differentiation of 3D scanners and their positioning method when applied to pipeline integrity | |
CN117367516A (en) | Aluminum-plastic composite board performance detection system based on multidimensional test | |
Armingol et al. | Statistical pattern modeling in vision-based quality control systems | |
CN112504184A (en) | Rapid online quality inspection system for three-dimensional size of steel plate | |
CN115578429B (en) | Mold online precision detection method based on point cloud data | |
Zhou et al. | Automated detection of surface defects on sphere parts using laser and CDD measurements | |
Sun et al. | Precision work-piece detection and measurement combining top-down and bottom-up saliency | |
Akhtar et al. | A robotics inspection system for detecting defects on semi-specular painted automotive surfaces | |
CN105783782B (en) | Surface curvature is mutated optical profilometry methodology | |
CN109063738A (en) | A kind of ceramic water valve plates automatic on-line detection method of compressed sensing | |
Klarák et al. | Processing 3D data from laser sensor into visual content using pattern recognition | |
Shao et al. | High-Speed and Accurate Method for the Gear Surface Integrity Detection Based on Visual Imaging | |
CN117670887B (en) | Tin soldering height and defect detection method based on machine vision | |
Sahoo et al. | An effective quality inspection system using image processing techniques |
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 |