CN102455171A - Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof - Google Patents
Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof Download PDFInfo
- Publication number
- CN102455171A CN102455171A CN2010105212997A CN201010521299A CN102455171A CN 102455171 A CN102455171 A CN 102455171A CN 2010105212997 A CN2010105212997 A CN 2010105212997A CN 201010521299 A CN201010521299 A CN 201010521299A CN 102455171 A CN102455171 A CN 102455171A
- Authority
- CN
- China
- Prior art keywords
- weld
- image
- seam
- laser
- weld seam
- 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.)
- Granted
Links
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Laser Beam Processing (AREA)
Abstract
The invention relates to a method for detecting the geometric shape of the back of a tailor-welding weld and an implementing device thereof. The method comprises the following steps that: a linear laser is used for projecting laser on the back of a tailor-welding weld so as to generate a plurality of laser streaks on the surface of the tailor-welding weld; a sensing unit acquires and obtains a weld surface gray level image including the laser streaks and transmits the acquired image to an image processing unit; the image processing unit processes the acquired weld image to obtain all shape parameters; and the weld quality is judged and thus the detection on the geometric shape of the back of the weld is finished. The implementing device of the method comprises a sensing unit, an image processing unit and a parameter setting and control unit, wherein the sensing unit is used for acquiring the image of a tailor-welding weld and transmitting the image to the image processing unit, the image processing unit is used for processing the acquired weld image, the parameter setting and control unit is used for setting all parameters and is respectively in communication with the sensing unit, the image processing unit and an upper computer. According to the invention, detection error caused by interference factors such as dust on the back of the weld, welding splatter, board scratch and the like in the welding process can be avoided, and the detection accuracy is improved greatly.
Description
Technical field
The present invention relates to welding quality Automatic Measurement Technique field, how much pattern detection methods of specifically a kind of laser assembly solder back of weld and implement device thereof.
Background technology
Laser assembly solder technology has a wide range of applications in industries such as automobile making, metallurgy owing to have characteristics such as high-level efficiency, high-speed, high precision, strong adaptability.Laser assembly solder plate butt welded seam quality requirements is strict especially; Especially back of weld melts the wide pattern parameters that wait how much; Be the important indicator of weighing quality of weld seam molding, yet because the laser assembly solder speed of welding is fast, and its back of weld is different from the weld seam front; Can adsorb flue dust in the welding process, make that the online detection of pattern of laser assembly solder back of weld is very difficult.
In order to understand the back of weld quality, adopt the mode of artificial vision's off-line sampling observation usually, because tester's fatigue and nonuniformity; Make a lot of detection tasks for manual work time-consuming and the effort; And because detection information can not be in time corresponding with production status, can not reflect the state of production line in real time, can not realize the closed loop size Control of producing effectively; Reduced the effective operation rate of laser assembly solder line; The also corresponding rejection rate that increases the welding plate has seriously influenced the robotization that laser assembly solder is produced, and has limited the further lifting of product quality.
By contrast; Computer Vision Detection method reliability is high, can guarantee the consistance that detects can in welding process, detect weldquality by real-time online simultaneously; Realize the deliver a child monitoring of quality in the product process of butt welding, can also improve the autonomy-oriented and the intelligent level of welding robot welding process.The appliance computer vision is carried out the canonical system of face of weld quality testing such as the face of weld quality detecting system of Canadian ServoRobot company; This system adopts structured light principle of triangulation identification weldquality; For back of weld, because the structured light of projection does not have the obvious characteristic point, and back of weld can adsorb flue dust in the welding process; Disturb the structure optical information, can't obtain the molten wide information of back of weld.The face of weld quality detecting system of Switzerland Soutec company, this system can obtain laser stripe and face of weld gray level image simultaneously, realize the detection of how much pattern parameters of back of weld, but price is more expensive, and can not carries out the production run closed-loop control.
Summary of the invention
High to weld seam detection cost in the prior art, can not carry out weak point such as production run closed-loop control, the technical matters that the present invention will solve provides that a kind of cost is lower, antijamming capability strong, helps to realize the how much pattern detection methods of laser assembly solder back of weld and the implement device thereof of production run closed-loop control.
For solving the problems of the technologies described above, the technical scheme that the present invention adopts is:
How much pattern detection methods of a kind of laser assembly solder back of weld of the present invention may further comprise the steps:
Utilize the word line laser instrument to be incident upon the back bead surface and produce laser stripe;
Obtain the face of weld gray level image that comprises laser stripe through the sensing unit collection, and send the image of being gathered to graphics processing unit;
Graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value;
Above-mentioned each pattern parameter value is compared with international welding quality standard, differentiate weldquality, accomplish how much patterns of back of weld and detect.
Said graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value and may further comprise the steps:
Extract weld bead feature points according to laser stripe and weld seam texture region common factor;
Detect how much patterns of face of weld according to weld bead feature points, obtain the pixel value of weld width and weld seam convex-concave degree;
Extract the laser stripe beyond the above-mentioned common factor, calculate the pixel value of weld seam mispairing;
Set up direct mapping relations according to the corresponding relation of image space and physical space, the sensing unit that comprises camera, laser instrument and additional source of light is demarcated, calculated, obtain the physical values of weld width, weld seam convex-concave degree and weld seam mispairing.
The said common factor according to laser stripe and weld seam texture region extracts weld bead feature points and is:
(1) to the sheet material of unlike material, different-thickness, carries out the parameter setting according to gathering the image that obtains;
(2) on the image after parameter is provided with, set the weld seam area-of-interest according to the laser stripe center;
(3) butt welded seam region of interest area image carries out pre-service, obtains weld seam area-of-interest texture image;
(4) according to pretreated gradation of image, set the image segmentation threshold value, cut apart weld seam area-of-interest texture image, obtain the relevant range that comprises the weld seam texture region;
(5) in the above-mentioned relevant range that comprises the weld seam texture region, set the row barycenter, extract the weld seam texture region as the characteristic of extracting the weld seam texture region;
(6) with regional major radius as characteristic, in above-mentioned weld seam texture region, remove the influence of the interference region that welding plate surface reflection, cut produced;
(7) according to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
(8) juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
(9) calculate laser stripe and weld seam texture region common factor extraction weld bead feature points according to the juncture area left and right edges.
Said parameter comprises additional source of light brightness, laser stripe brightness, area-of-interest window width and length of window, welded seam area segmentation threshold and laser stripe segmentation threshold.
Said butt welded seam region of interest area image carries out pre-service, comprises the filtering of area-of-interest image texture, image smoothing and grey level histogram linearization.
The pixel value of said weld width obtains through following method:
Introduce corrected parameter, the weld bead feature points of laser stripe and the extraction of weld seam texture region common factor is revised, remove the pixel deviation that is produced by pre-service, border, revised welded seam area common factor left and right sides difference is the pixel value of the current width of weld seam;
The pixel value of the current width of butt welded seam is checked, and obtains the pixel value of the weld width of zone of reasonableness.
The current width value of said butt welded seam is checked, and the pixel value that obtains the weld width of zone of reasonableness carries out through following method:
Calculate the weld width average of the preceding n two field picture of current weld image, do poorly, obtain difference with the current width value of weld seam;
Above-mentioned difference is compared with predefined weld width decision threshold,, think that then current weld width value is a reasonable value, be the weld width value of detection if above-mentioned difference does not exceed the weld width decision threshold; If above-mentioned difference exceeds the weld width decision threshold, then be unreasonable value, then the weld width average of preceding n two field picture is the weld width value of detection.
The pixel value of weld seam convex-concave degree obtains through following method:
Left and right sides unique point according to the weld width that detects is confirmed a line, and the point on the laser stripe center line in the middle of two unique points is convex value to the positive maximal value in the ultimate range of this line, and negative maximal value is the concavity value.
Laser stripe beyond the above-mentioned common factor of said extraction, the pixel value that calculates the weld seam mispairing obtains through following method:
According to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
The juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
Calculate the capable average of the outer left and right sides area pixel of laser stripe juncture area, the difference of the capable average of left and right sides area pixel is the pixel value of weld seam mispairing.
The implement device sensing unit 1 that how much patterns of a kind of laser assembly solder back of weld of the present invention detect produces laser stripe on welded seam area to be detected, gathers the laser assembly solder weld image, transfers to graphics processing unit 2;
Parameter is provided with and control module 3, each parameter of sensing unit 1 and graphics processing unit 2 is provided with, and carries out communication with sensing unit 1, graphics processing unit 2, host computer 5 respectively and be connected.
Said sensing unit 1 is made up of video camera 12, word line laser instrument 13, additional source of light 10 and reflective mirror 14; Wherein, Video camera 12 is located at the position that can take the laser assembly solder weld seam; Word line laser instrument 13 is installed on displacement and regulates on the seat, and its laser stripe that sends projects welded seam area to be detected through reflective mirror 14, and additional source of light 10 is located at the position that can illuminate whole welded seam area to be detected.
Apparatus of the present invention also comprise neutral light damping plate 11, are located at the front end of video camera 12 camera lenses.
Apparatus of the present invention also are provided with the coldplate that is used to reduce video camera 12 working temperatures, are installed on the side of video camera 12.
Said additional source of light 10 is two LED bar light, and the length direction of LED bar light is consistent with bead direction at the camera lens two ends, and symmetry is installed, and the light source incident light plane of LED bar light becomes more than or equal to 65 ° and less than 90 ° of angles with plane, weld seam place.
The present invention has following beneficial effect and advantage:
1. the present invention passes through based on image sequence information; The method that adopts laser stripe and weld seam texture to combine; The detection error of having avoided disturbing factors such as the flue dust that back of weld adsorbs in the welding process, spatter, sheet material cut to produce; Greatly improve accuracy of detection, guaranteed the validity of testing result.
2. the present invention can realize automatic, the online detection of laser assembly solder back of weld how much patterns, has that antijamming capability is strong, measuring speed fast, can in rugged surroundings, use, the measuring accuracy height, helps to realize the advantage of welding process parameter close-loop feedback control.
Description of drawings
Fig. 1 is the inventive method general flow chart;
The process flow diagram of Fig. 2 in the inventive method the weld image that collects being handled;
Fig. 3 (a)~(c) how much pattern testing result synoptic diagram ()~(three) of back of weld for utilizing the inventive method;
Fig. 4 is apparatus of the present invention structural representation.
Embodiment
Below in conjunction with accompanying drawing the present invention is made further detailed description.
As shown in Figure 4; The implement device that how much patterns of laser assembly solder back of weld of the present invention detect comprises: sensing unit 1, graphics processing unit 2 and parameter setting and control module 3; Wherein sensing unit 1 is used to produce laser stripe in welded seam area to be detected; Gather the laser assembly solder weld image, transfer to graphics processing unit 2; Graphics processing unit 2 is handled the weld image that collects under the control of parameter setting and control module 3; Parameter is provided with and control module 3, each parameter of sensing unit 1 and graphics processing unit 2 is provided with, and carries out communication with sensing unit 1, graphics processing unit 2, host computer 5 respectively and be connected, and host computer 5 comprises PLC or robot.
Also can comprise neutral light damping plate 11, be located at the front end of video camera 12 camera lenses, its effect is to reduce light intensity, and the brightness of image of avoiding photographing is too high; Seat is regulated in displacement, is the one-dimensional linear translation stage, can carry out the one-dimensional linear of word line laser instrument 13 and regulate, and realizes the fine position on the laser focus direction; Said additional source of light 10 is two LED bar light (present embodiment is blue strip source); LED bar light is the symmetry installation at the camera lens two ends; Length direction is consistent with bead direction; The light source incident light plane of LED bar light becomes more than or equal to 65 ° and less than 90 ° of angles with plane, weld seam place, and present embodiment adopts 70 ° of angles; Reflective mirror is installed on one and regulates on the seat, and integral installation can realize the fine position of laser stripe in the image-region of camera acquisition in word line laser instrument below.
The course of work of this implement device is following:
The laser beam that word line laser instrument 13 sends produces structured light face (laser stripe), shines the back side of welding work pieces 9, forms the laser stripe 7 of reflection weld seam 6 how much shape characteristics; Video camera 12 is gathered and is obtained the face of weld gray level image 8 that comprises laser stripe 7; And send the image of being gathered to graphics processing unit 2 through Cameralink cable or netting twine, and handle by 2 pairs of weld images that collect 8 of graphics processing unit, extract weld bead feature points, detect how much patterns of face of weld; The butt welded seam image is demarcated; Calculate and obtain each pattern parameter value, and compare, differentiate weldquality with international welding quality standard.Each parameter of 3 pairs of sensing units 1 of parameter setting and control module and graphics processing unit 2 is provided with; And carry out communication with sensing unit 1, graphics processing unit 2 and host computer 5 respectively and be connected; Accomplish the identification and the judgement of how much patterns of back of weld, realize that finally the quality of how much patterns of back of weld detects automatically.
As shown in Figure 1, the workflow synoptic diagram when the how much pattern detection methods of laser assembly solder back of weld based on structured light that propose for the present invention detect, this method may further comprise the steps:
Utilize the word line laser instrument to be incident upon the back bead surface and produce laser stripe;
Obtain the face of weld gray level image that comprises laser stripe through the sensing unit collection, and send the image of being gathered to graphics processing unit;
Graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value;
Above-mentioned each pattern parameter value is compared with international welding quality standard, differentiate weldquality, accomplish how much patterns of back of weld and detect.
As shown in Figure 2, said graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value and may further comprise the steps:
Extract weld bead feature points according to laser stripe and weld seam texture region common factor;
Detect how much patterns of face of weld according to weld bead feature points, obtain the pixel value of weld width and weld seam convex-concave degree;
Extract the laser stripe beyond the above-mentioned common factor, calculate the pixel value of weld seam mispairing;
Set up direct mapping relations according to the corresponding relation of image space and physical space, the sensing unit that comprises camera, laser instrument and additional source of light is demarcated, calculated, obtain the physical values of weld width, weld seam convex-concave degree and weld seam mispairing.
The said common factor according to laser stripe and weld seam texture region extracts weld bead feature points and is:
(1) to the sheet material of unlike material, different-thickness, carries out the parameter setting according to gathering the image that obtains; In step 101, gather and obtain laser assembly solder back of weld original image; Step 102 is carried out the parameter setting; Sheet material to unlike material, different-thickness; Carry out the initial parameter setting, comprise additional source of light brightness, laser stripe brightness, area-of-interest window width and length of window, welded seam area segmentation threshold and laser stripe segmentation threshold.
(2) on the image after parameter is provided with, set the weld seam area-of-interest according to the laser stripe center; The purpose of this step is to dwindle processing region, improves processing speed.
First width of cloth image to collection is obtained by click weld seam light line central area, is designated as a P
Center(i, j), as central point of area of interest, then 4 apex coordinates of window area are respectively P
1(i-Width/2, j-Length/2), P
2(i-Width/2, j+Length/2), P
3(i+Width/2, j-Length/2), P
4(i+Width/2 j+Length/2), and is applied to the area-of-interest setting of every width of cloth image afterwards.Wherein Width is a window width, and Length is a length of window,
(3) before pre-service, send beginning sense command (step 103).Detect beginning, butt welded seam region of interest area image carries out pre-service, to remove picture noise, improves picture contrast, obtains weld seam area-of-interest texture image.Pre-service comprises texture filtering, image smoothing, and the grey level histogram linearization, purpose is in order accurately to cut apart the weld seam texture region information of obtaining.
(4) according to pretreated gradation of image, set the image segmentation threshold value, cut apart weld seam area-of-interest texture image, obtain the relevant range that comprises the weld seam texture region;
(5) in the above-mentioned relevant range that comprises the weld seam texture region, set the row barycenter, extract the weld seam texture region as the characteristic of extracting the weld seam texture region;
(6) with regional major radius as characteristic, remove the influence of the interference region that welding plate surface reflection, cut produced;
Butt welded seam region of interest area image carries out the extraction of weld seam texture respectively and laser stripe extracts the processing of two steps: step 104-step 106 butt welded seam region of interest area image carries out texture filtering; Image smoothing; Pre-service such as grey level histogram linearization, Threshold Segmentation; To remove picture noise, improve picture contrast, obtain the relevant range that comprises the weld seam texture region; In step 107, in the above-mentioned relevant range that comprises the weld seam texture region, set the row barycenter as the characteristic of extracting the weld seam texture region, extract the weld seam texture region; In step 108, as characteristic, remove the influence of the interference region that welding plate surface reflection, cut produced with regional major radius, if extracted the weld seam texture region then carry out next step calculating, otherwise then refusal detects (step 109).
(7) according to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
In step 110, according to laser stripe gray scale setting threshold, the butt welded seam area-of-interest carries out Threshold Segmentation, and step 111 extracts the laser stripe zone.
(8) juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
In step 112, the weld seam texture region with cut apart after the juncture area of laser stripe, if there is juncture area, then carry out next step calculating, otherwise then refusal detects.
(9) calculate laser stripe and weld seam texture region common factor extraction weld bead feature points according to the juncture area left and right edges;
Juncture area according to obtaining can calculate and obtain weld width and weld seam convex-concave degree parameter: in step 113, extract the juncture area left and right boundary point, i.e. weld bead feature points.
The pixel value of weld width obtains through following method:
Introduce corrected parameter; Weld bead feature points to laser stripe and the extraction of weld seam texture region common factor is revised; The pixel deviation that removal is produced by pre-service, border, revised welded seam area common factor left and right sides difference is the pixel value (corresponding step 114) of the current width of weld seam.
If there is error in image, need be in the present frame of handling, the pixel value of the current width of weld seam is checked, and obtains the pixel value of the weld width in the zone of reasonableness; Step 115 is calculated the preceding n frame weld width average of current weld seam; Step 116 is calculated the difference Delta of current weld width and weld width average, if judge that difference Delta is unreasonable, and the weld width value of n frame weld width average for detecting before the step 120; Characteristic of correspondence point is a weld bead feature points; If instead difference Delta is reasonable, step 121 weld bead feature points is extracted correct, and current weld width value is detected weld width value.
The pixel value of weld seam convex-concave degree obtains through following method:
Left and right sides unique point according to the weld width that detects is confirmed a line, and the point on the laser stripe center line in the middle of two unique points is convex value to the positive maximal value in the ultimate range of this line, and negative maximal value is the concavity value.In step 122; Extract the corresponding light line center line left and right boundary point of corresponding weld width as weld seam left and right sides unique point; Confirm a line by weld seam left and right sides unique point; Point on the laser stripe center line in the middle of two unique points is the convex-concave degree value of weld seam to the ultimate range of line, and wherein just maximal value is a convex value, and negative maximal value is the concavity value.
Extract the laser stripe beyond the above-mentioned common factor, the pixel value that calculates the weld seam mispairing obtains through following method:
According to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
The juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
Calculate the capable average of the outer left and right sides area pixel of laser stripe juncture area, the difference of the capable average of left and right sides area pixel is the pixel value of weld seam mispairing.
Step 117-step 119; For the weld seam misfit parameter is obtained in calculating: zone about outside the step 117 calculating laser stripe juncture area; The capable average of step 118 calculating left and right sides laser stripe, laser stripe capable equal value difference in the step 119 calculating left and right sides is the pixel value of weld seam mispairing.
As shown in Figure 3, be how much pattern testing results of back of weld synoptic diagram, can extract geometry pattern parameters such as comprising weld width, mispairing, convex-concave degree.Wherein, the distance of two vertical parallel lines of Fig. 3 (a) is the weld width value; There is positive extreme value in the distance of Fig. 3 (b) laser stripe center line and two unique point lines, is the convexity value; The distance of the parallel lines of two levels of Fig. 3 (c) is the weld seam misfit value.
Claims (14)
1. how much pattern detection methods of a laser assembly solder back of weld is characterized in that may further comprise the steps:
Utilize the word line laser instrument to be incident upon the back bead surface and produce laser stripe;
Obtain the face of weld gray level image that comprises laser stripe through the sensing unit collection, and send the image of being gathered to graphics processing unit;
Graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value;
Above-mentioned each pattern parameter value is compared with international welding quality standard, differentiate weldquality, accomplish how much patterns of back of weld and detect.
2. by how much pattern detection methods of the described laser assembly solder back of weld of claim 1, it is characterized in that: said graphics processing unit is handled the weld image that collects, and obtains each pattern parameter value and may further comprise the steps:
Extract weld bead feature points according to laser stripe and weld seam texture region common factor;
Detect how much patterns of face of weld according to weld bead feature points, obtain the pixel value of weld width and weld seam convex-concave degree;
Extract the laser stripe beyond the above-mentioned common factor, calculate the pixel value of weld seam mispairing;
Set up direct mapping relations according to the corresponding relation of image space and physical space, the sensing unit that comprises camera, laser instrument and additional source of light is demarcated, calculated, obtain the physical values of weld width, weld seam convex-concave degree and weld seam mispairing.
3. by how much pattern detection methods of the described laser assembly solder back of weld of claim 2, it is characterized in that: saidly be according to laser stripe and the weld seam texture region extraction weld bead feature points of occuring simultaneously:
(1) to the sheet material of unlike material, different-thickness, carries out the parameter setting according to gathering the image that obtains;
(2) on the image after parameter is provided with, set the weld seam area-of-interest according to the laser stripe center;
(3) butt welded seam region of interest area image carries out pre-service, obtains weld seam area-of-interest texture image;
(4) according to pretreated gradation of image, set the image segmentation threshold value, cut apart weld seam area-of-interest texture image, obtain the relevant range that comprises the weld seam texture region;
(5) in the above-mentioned relevant range that comprises the weld seam texture region, set the row barycenter, extract the weld seam texture region as the characteristic of extracting the weld seam texture region;
(6) with regional major radius as characteristic, in above-mentioned weld seam texture region, remove the influence of the interference region that welding plate surface reflection, cut produced;
(7) according to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
(8) juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
(9) calculate laser stripe and weld seam texture region common factor extraction weld bead feature points according to the juncture area left and right edges.
4. by how much pattern detection methods of the described laser assembly solder back of weld of claim 3; It is characterized in that: said parameter comprises additional source of light brightness, laser stripe brightness, area-of-interest window width and length of window, welded seam area segmentation threshold and laser stripe segmentation threshold.
5. by how much pattern detection methods of the described laser assembly solder back of weld of claim 3, it is characterized in that: said butt welded seam region of interest area image carries out pre-service, comprises the filtering of area-of-interest image texture, image smoothing and grey level histogram linearization.
6. by how much pattern detection methods of the described laser assembly solder back of weld of claim 2, it is characterized in that: the pixel value of said weld width obtains through following method:
Introduce corrected parameter, the weld bead feature points of laser stripe and the extraction of weld seam texture region common factor is revised, remove the pixel deviation that is produced by pre-service, border, revised welded seam area common factor left and right sides difference is the pixel value of the current width of weld seam;
The pixel value of the current width of butt welded seam is checked, and obtains the pixel value of the weld width of zone of reasonableness.
7. by how much pattern detection methods of the described laser assembly solder back of weld of claim 6, it is characterized in that: the pixel value of the current width of said butt welded seam is checked, and the pixel value that obtains the weld width of zone of reasonableness carries out through following method:
Calculate the weld width average of the preceding n two field picture of current weld image, do poorly, obtain difference with the current width value of weld seam;
Above-mentioned difference is compared with predefined weld width decision threshold,, think that then current weld width value is a reasonable value, be the weld width value of detection if above-mentioned difference does not exceed the weld width decision threshold; If above-mentioned difference exceeds the weld width decision threshold, then be unreasonable value, then the weld width average of preceding n two field picture is the pixel value of the weld width of detection.
8. by how much pattern detection methods of the described laser assembly solder back of weld of claim 2, it is characterized in that: the pixel value of weld seam convex-concave degree obtains through following method:
Left and right sides unique point according to the weld width that detects is confirmed a line, and the point on the laser stripe center line in the middle of two unique points is convex value to the positive maximal value in the ultimate range of this line, and negative maximal value is the concavity value.
9. by how much pattern detection methods of the described laser assembly solder back of weld of claim 2, it is characterized in that: the laser stripe beyond the above-mentioned common factor of said extraction, the pixel value that calculates the weld seam mispairing obtains through following method:
According to predefined laser stripe segmentation threshold, butt welded seam region of interest area image carries out Threshold Segmentation again, the laser stripe after obtaining cutting apart;
The juncture area of the laser stripe after calculating the weld seam texture region and cutting apart;
Calculate the capable average of the outer left and right sides area pixel of laser stripe juncture area, the difference of the capable average of left and right sides area pixel is the pixel value of weld seam mispairing.
10. the implement device that detects of how much patterns of a laser assembly solder back of weld is characterized in that comprising:
Sensing unit (1) produces laser stripe on welded seam area to be detected, gathers the laser assembly solder weld image, transfers to graphics processing unit (2);
Graphics processing unit (2) is handled the weld image that collects under the control of parameter setting and control module (3);
Parameter is provided with and control module (3), each parameter of sensing unit (1) and graphics processing unit (2) is provided with, and carries out communication with sensing unit (1), graphics processing unit (2), host computer (5) respectively and be connected.
11. implement device by how much patterns detections of the described laser assembly solder back of weld of claim 10; It is characterized in that: said sensing unit (1) is made up of video camera (12), word line laser instrument (13), additional source of light (10) and reflective mirror (14); Wherein, Video camera (12) is located at the position that can take the laser assembly solder weld seam; Word line laser instrument (13) is installed on displacement and regulates on the seat, and its laser stripe that sends projects welded seam area to be detected through reflective mirror (14), and additional source of light (10) is located at the position that can illuminate whole welded seam area to be detected.
12. the implement device by how much patterns of the described laser assembly solder back of weld of claim 11 detect is characterized in that: also comprise neutral light damping plate (11), be located at the front end of video camera (12) camera lens.
13. the implement device by how much patterns of the described laser assembly solder back of weld of claim 11 detect is characterized in that: also be provided with the coldplate that is used to reduce video camera (12) working temperature, be installed on the side of video camera (12).
14. implement device by how much patterns detections of the described laser assembly solder back of weld of claim 11; It is characterized in that: said additional source of light (10) is two LED bar light; The length direction of LED bar light is consistent with bead direction at the camera lens two ends; Symmetry is installed, and the light source incident light plane of LED bar light becomes more than or equal to 65 ° and less than 90 ° of angles with plane, weld seam place.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201010521299.7A CN102455171B (en) | 2010-10-27 | 2010-10-27 | Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201010521299.7A CN102455171B (en) | 2010-10-27 | 2010-10-27 | Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102455171A true CN102455171A (en) | 2012-05-16 |
CN102455171B CN102455171B (en) | 2014-01-15 |
Family
ID=46038607
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201010521299.7A Expired - Fee Related CN102455171B (en) | 2010-10-27 | 2010-10-27 | Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102455171B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102854193A (en) * | 2012-08-30 | 2013-01-02 | 苏州天准精密技术有限公司 | Detection method and detection system used for image defect detection |
WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
CN105160656A (en) * | 2015-08-05 | 2015-12-16 | 哈尔滨工业大学 | Pipe fitting internal thread vision identification system and method based on gray scale co-occurrence matrix |
CN105300296A (en) * | 2014-07-15 | 2016-02-03 | 中国科学院沈阳自动化研究所 | Laser tailored welding curved surface abutted seam measuring method and realization device thereof |
CN105783712A (en) * | 2014-12-26 | 2016-07-20 | 北京中电科电子装备有限公司 | Method and device used for detecting knife mark |
CN106574665A (en) * | 2014-08-12 | 2017-04-19 | Ntn株式会社 | Device for inspecting junction-type outer joint member of constant velocity universal joint |
CN108620714A (en) * | 2018-07-06 | 2018-10-09 | 太原科技大学 | Welding deviation detecting system based on the molten baths GMAW contour feature and its detection method |
CN110930368A (en) * | 2019-10-31 | 2020-03-27 | 中船重工鹏力(南京)智能装备系统有限公司 | Method for extracting characteristics of real-time welding image of sheet lap weld |
CN110961821A (en) * | 2018-09-30 | 2020-04-07 | 宝山钢铁股份有限公司 | Method for realizing automatic judgment and release of strip steel laser splicing weld inspection |
CN113369641A (en) * | 2021-05-18 | 2021-09-10 | 宁波博视达焊接机器人有限公司 | Visual sensor for tracking welding seam |
CN113770585A (en) * | 2021-11-11 | 2021-12-10 | 中国科学院自动化研究所 | Method and device for controlling underwater welding quality |
CN115063361A (en) * | 2022-06-10 | 2022-09-16 | 东南大学 | Modularized welding seam recognition device and method |
CN117990025A (en) * | 2024-04-03 | 2024-05-07 | 河北省特种设备监督检验研究院 | Geometric dimension detection device for pipeline weld joint detection |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1782659A (en) * | 2004-12-02 | 2006-06-07 | 中国科学院自动化研究所 | Welding seam tracking sight sensor based on laser structure light |
WO2008028580A1 (en) * | 2006-09-06 | 2008-03-13 | Precitec Vision Gmbh & Co. Kg | Method and device for the optical assessment of welding quality during welding |
KR20090055278A (en) * | 2007-11-28 | 2009-06-02 | 에스티엑스조선주식회사 | Welding part geometry measurement and welding condition control system for welding automation |
CN101750416A (en) * | 2008-12-19 | 2010-06-23 | 中国科学院沈阳自动化研究所 | Visual welding seam surface quality detection sensor based on line structure light |
-
2010
- 2010-10-27 CN CN201010521299.7A patent/CN102455171B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1782659A (en) * | 2004-12-02 | 2006-06-07 | 中国科学院自动化研究所 | Welding seam tracking sight sensor based on laser structure light |
WO2008028580A1 (en) * | 2006-09-06 | 2008-03-13 | Precitec Vision Gmbh & Co. Kg | Method and device for the optical assessment of welding quality during welding |
KR20090055278A (en) * | 2007-11-28 | 2009-06-02 | 에스티엑스조선주식회사 | Welding part geometry measurement and welding condition control system for welding automation |
CN101750416A (en) * | 2008-12-19 | 2010-06-23 | 中国科学院沈阳自动化研究所 | Visual welding seam surface quality detection sensor based on line structure light |
Non-Patent Citations (2)
Title |
---|
YUAN LI,ET AL: "Measurement and Defect Detection of the Weld Bead Based on Online Vision Inspection", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》 * |
张万江 等: "基于结构光视觉的激光拼焊焊缝背面质量检测方法研究", 《仪表技术与传感器》 * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102854193A (en) * | 2012-08-30 | 2013-01-02 | 苏州天准精密技术有限公司 | Detection method and detection system used for image defect detection |
WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
CN105300296B (en) * | 2014-07-15 | 2018-02-02 | 中国科学院沈阳自动化研究所 | A kind of laser assembly solder curved surface seam measuring method and its realization device |
CN105300296A (en) * | 2014-07-15 | 2016-02-03 | 中国科学院沈阳自动化研究所 | Laser tailored welding curved surface abutted seam measuring method and realization device thereof |
CN106574665B (en) * | 2014-08-12 | 2020-05-12 | Ntn株式会社 | Device for inspecting joint type outer joint member of constant velocity universal joint |
CN106574665A (en) * | 2014-08-12 | 2017-04-19 | Ntn株式会社 | Device for inspecting junction-type outer joint member of constant velocity universal joint |
CN105783712A (en) * | 2014-12-26 | 2016-07-20 | 北京中电科电子装备有限公司 | Method and device used for detecting knife mark |
CN105783712B (en) * | 2014-12-26 | 2018-11-27 | 北京中电科电子装备有限公司 | A kind of method and device detecting tool marks |
CN105160656B (en) * | 2015-08-05 | 2017-11-03 | 哈尔滨工业大学 | A kind of pipe fitting internal thread visual identifying system and method based on gray level co-occurrence matrixes |
CN105160656A (en) * | 2015-08-05 | 2015-12-16 | 哈尔滨工业大学 | Pipe fitting internal thread vision identification system and method based on gray scale co-occurrence matrix |
CN108620714A (en) * | 2018-07-06 | 2018-10-09 | 太原科技大学 | Welding deviation detecting system based on the molten baths GMAW contour feature and its detection method |
CN110961821A (en) * | 2018-09-30 | 2020-04-07 | 宝山钢铁股份有限公司 | Method for realizing automatic judgment and release of strip steel laser splicing weld inspection |
CN110930368A (en) * | 2019-10-31 | 2020-03-27 | 中船重工鹏力(南京)智能装备系统有限公司 | Method for extracting characteristics of real-time welding image of sheet lap weld |
CN110930368B (en) * | 2019-10-31 | 2022-02-18 | 中船重工鹏力(南京)智能装备系统有限公司 | Method for extracting characteristics of real-time welding image of sheet lap weld |
CN113369641A (en) * | 2021-05-18 | 2021-09-10 | 宁波博视达焊接机器人有限公司 | Visual sensor for tracking welding seam |
CN113770585A (en) * | 2021-11-11 | 2021-12-10 | 中国科学院自动化研究所 | Method and device for controlling underwater welding quality |
CN115063361A (en) * | 2022-06-10 | 2022-09-16 | 东南大学 | Modularized welding seam recognition device and method |
CN117990025A (en) * | 2024-04-03 | 2024-05-07 | 河北省特种设备监督检验研究院 | Geometric dimension detection device for pipeline weld joint detection |
CN117990025B (en) * | 2024-04-03 | 2024-06-07 | 河北省特种设备监督检验研究院 | Geometric dimension detection device for pipeline weld joint detection |
Also Published As
Publication number | Publication date |
---|---|
CN102455171B (en) | 2014-01-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102455171B (en) | Method for detecting geometric shape of back of tailor-welding weld and implementing device thereof | |
CN106181162B (en) | A kind of real-time weld joint tracking detection method based on machine vision | |
CN101526484B (en) | Bearing defect detecting technique based on embedded-type machine vision | |
CN106226325B (en) | A kind of seat surface defect detecting system and its method based on machine vision | |
CN102175700B (en) | Method for detecting welding seam segmentation and defects of digital X-ray images | |
CN108760747A (en) | A kind of 3D printing model surface defect visible detection method | |
CN103480991B (en) | Thin steel plate narrow welding joint online visual inspection and control device | |
CN101750416A (en) | Visual welding seam surface quality detection sensor based on line structure light | |
CN105571502A (en) | Measuring method of weld gap in friction-stir welding | |
CN103983203B (en) | A kind of laser melting coating molten bath defocus measuring device and its measuring method | |
CN104112269A (en) | Solar cell laser-marking parameter detection method based on machine vision and system thereof | |
US7599050B2 (en) | Surface defect inspecting method and device | |
CN206074465U (en) | A kind of battery based on machine vision and connection sheet solder joint detection means | |
CN110567963B (en) | Alloy analysis visual positioning method and device and alloy analysis system | |
CN108489394A (en) | A kind of large-scale sheet metal works almost T-stable automatic detection device and method | |
CN111879264A (en) | Flatness measurement and evaluation system based on line structured light | |
CN111307812A (en) | Welding spot appearance detection method based on machine vision | |
CN104976959B (en) | A kind of spring sizes on-line measurement system and its method based on machine vision | |
CN115546125A (en) | Method for error detection and track deviation correction of additive manufacturing cladding layer based on point cloud information | |
CN104148838A (en) | Small feature point visual tracking and real-time quality test device for sheet steel weld joints | |
CN203992877U (en) | The little characteristic point vision of sheet metal weld seam is followed the tracks of and real-time quality detection device | |
CN116664508A (en) | Weld surface quality detection method and computer readable storage medium | |
CN106370673A (en) | Automatic lens flaw detection method | |
CN107797517A (en) | The method and system detected using realizing of Robot Vision steel band punching processing | |
CN109741311B (en) | Aluminum alloy fusion welding back face fusion width detection method with false edge |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140115 Termination date: 20201027 |