CN104084669A - Visual detecting method and device based on decoupling detection for circular seams - Google Patents

Visual detecting method and device based on decoupling detection for circular seams Download PDF

Info

Publication number
CN104084669A
CN104084669A CN201410279698.5A CN201410279698A CN104084669A CN 104084669 A CN104084669 A CN 104084669A CN 201410279698 A CN201410279698 A CN 201410279698A CN 104084669 A CN104084669 A CN 104084669A
Authority
CN
China
Prior art keywords
weld
welding gun
welding
point
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201410279698.5A
Other languages
Chinese (zh)
Other versions
CN104084669B (en
Inventor
陈海永
王萌
马子涵
孙鹤旭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hebei University of Technology
Original Assignee
Hebei University of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hebei University of Technology filed Critical Hebei University of Technology
Priority to CN201410279698.5A priority Critical patent/CN104084669B/en
Publication of CN104084669A publication Critical patent/CN104084669A/en
Application granted granted Critical
Publication of CN104084669B publication Critical patent/CN104084669B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B23MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23KSOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00Arc welding or cutting
    • B23K9/12Automatic feeding or moving of electrodes or work for spot or seam welding or cutting
    • B23K9/127Means for tracking lines during arc welding or cutting
    • B23K9/1272Geometry oriented, e.g. beam optical trading
    • B23K9/1274Using non-contact, optical means, e.g. laser means
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B23MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23KSOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00Arc welding or cutting
    • B23K9/02Seam welding; Backing means; Inserts
    • B23K9/028Seam welding; Backing means; Inserts for curved planar seams

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Plasma & Fusion (AREA)
  • Mechanical Engineering (AREA)
  • Optics & Photonics (AREA)
  • Geometry (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention relates to a visual detecting method based on decoupling detection for circular seams. The method is used for detecting, recognizing and welding the welding seams of circular welding pieces with V-shaped grooves. The method is characterized by comprising the following steps: first, processing pictures and extracting features; and second, decoupling image information. For the method, according to the three-dimensional welding seam feature information extracted by decoupling single frame images, the deviation value of a welding gun in a two-dimensional plane relative to the welding seam center can be calculated, the accurate positioning of welding spots in three-dimensional space can be realized, the decoupling algorithm is performed by monocular vision when the two-dimensional welding seam information is obtained, the tedious matrix transformation when the welding seam information is demarcated is avoided, the demarcating and calculating time is greatly saved, and the real-time performance of a system is obviously improved. On one hand, building a complex imaging model of a camera is avoided, and on the other hand, the demarcating precision, the robustness and the applicability of the camera are improved.

Description

A kind of circumferential weld visible detection method and device detecting based on decoupling zero
Technical field
The present invention relates to determine detection method and the location of circumferential weld bond pad locations, specifically a kind of circumferential weld visible detection method and device detecting based on decoupling zero.
Background technology
Along with modernization welding production is more and more higher to the requirement of product quality and production efficiency, how to guarantee welding quality, improve Automation of Welding level and become the key subjects of pendulum in face of people.The working environment of welding process is very severe for welding robot human visual system, as interference such as welding job inevitably exists a large amount of arc lights, splashes, flue dust; Through the smooth workpiece in polishing rear surface, under LASER Light Source irradiates, because surperficial multiple reflections presents irregular reflective interference; Surface of the work is irregularity sometimes, often has the vestiges such as iron filings, greasy dirt, chalk mark; The weld shape of workpiece is different, and the material of weldment is also incomplete same.All of these factors taken together has formed the challenge to the processing of vision system image and acquisition of information.In the situation that strong jamming is severe, obtain and the image of visual information processed the technical bottleneck that Study on Problems becomes the current welding robot vision-based detection application of restriction, realizes Weld Seam Tracking Control.In general camera calibration process, often first determine the inside and outside parameter of camera, camera inner parameter only need be demarcated once, but external parameter need to change constantly along with the rotation of weldment, be difficult for demarcating, and operand is very large.The information that binocular vision need to be obtained two video cameras when obtaining weld seam three-dimensional information merges, and amount of calculation is large, affects system real time.
Chinese patent CN101954557 discloses the two-dimensional position information extracting method that a kind of structure light welding seam image is followed the tracks of, the method is obtained weld image by a structured light formula image weld seam tracking sensor, sensor is processed this image, thereby obtain the center of weld seam, for weld joint tracking directly provides the left-right deviation information of weld seam; Simultaneously, utilize the geometry site of sensor and welding work pieces, the former frame image of Real-time Obtaining and a rear frame are compared, draw the variable quantity of welding gun short transverse, thereby provide welding gun height tolerance information, its weak point is: in welding process, because light, arc light, impact reflective, the factor such as splash can cause the center of the current frame image that video camera obtains inaccurate, if front and back two two field pictures compare, will make deviation more and more inaccurate.
Summary of the invention
For the deficiencies in the prior art, the technical problem that quasi-solution of the present invention is determined is that a kind of circumferential weld visible detection method and device detecting based on decoupling zero is provided.This detection method is carried out decoupling zero according to the three-dimensional characteristics of weld seam information that single-frame images is extracted, calculate the departure at welding gun phase butt welded seam center in two dimensional surface, do not relate to extensive matrixing, amount of calculation is less, make not have coupled relation between the variation of characteristics of image and welding gun motion, can realize the accurate location at three dimensions solder joint.This checkout gear vision detection system comprises monocular-camera, laser instrument and the support that connects both, and it is simple in structure, can identify accurately and rapidly bond pad locations.
The present invention solves the technical scheme that described method and technology problem adopts, and designs a kind of circumferential weld visible detection method detecting based on decoupling zero, for detection, the identification of the weld seam of V-butt annular weldment and weld, it is characterized in that the concrete steps of the method are:
The first step, image is processed and feature extraction
1-1 image acquisition: obtain the welded seam area image that comprises laser stripe by smart camera sensor;
The scope of the laser stripe of the image that 1-2 region of interesting extraction: determining step 1-1 obtains in image;
1-3 image threshold self adaptation is cut apart: on the basis of step 1-2, determine the available point of laser stripe;
1-4 laser structure light center line drawing: adopt the mode of extracting by column to extract V-butt circular weld center line array method;
The filtering of 1-5 center line and interpolation: the missing point in the laser structure light center line that interpolation completion is extracted, adopts gauss low frequency filter further to eliminate noise jamming afterwards;
1-6Hough straight line conversion: on the basis of step 1-5, the laser stripe after differential filtering is carried out to hough fitting a straight line;
1-7 feature point extraction: the ordinate value corresponding with the straight line obtaining after hough conversion according to the ordinate value of the center line group of the laser stripe extracting compares the flex point that obtains laser stripe, and the mid point of two flex points is exactly weld bead feature points.
Second step, image information decoupling zero
2-1 horizontal direction deviation identification: utilize the linear relationship of weld width and weld seam developed width on the plane of delineation to obtain the departure of welding-torch horizontal direction;
2-1 vertical direction deviation identification: obtain the departure of welding gun vertical direction in the space geometry relation of vertical direction formation according to the weld seam on the plane of delineation and the welding gun in space.
The technical scheme that the present invention solves described device technique problem employing is, design a kind of circumferential weld vision inspection apparatus detecting based on decoupling zero that adopts above-mentioned detection method, it is characterized in that this device comprises vision detection system, the first support, cross adjusting bracket, welding gun and fixture, described vision detection system comprises video camera, the second support and laser instrument, described video camera is connected with cross adjusting bracket by the first support, by the second support and laser instrument, is fixedly linked; Cross adjusting bracket bottom is connected with welding gun top, and cross adjusting bracket can drive welding gun to move together, and laser instrument is apart from welding gun the place ahead 5-10cm; Below welding gun, be provided with fixture, by the scroll chuck on fixture, can fix V-butt annular weldment.
Compared with prior art, the three-dimensional characteristics of weld seam information that the inventive method is extracted by decoupling zero single-frame images, can calculate the departure at welding gun phase butt welded seam center in two dimensional surface, and then can realize the accurate location at three dimensions solder joint.Apparatus of the present invention adopt monocular vision when obtaining weld seam two-dimensional signal, to carry out decoupling algorithm, avoided matrixing loaded down with trivial details when demarcating weld seam information, greatly saved and demarcated and computing time, significantly improved the real-time of system, avoid setting up complicated camera imaging model on the one hand, strengthened on the other hand precision, robustness and the adaptability of camera calibration.In the process of image detection, realize decoupling zero, simplified control algolithm, reduced the requirement to controller.
Accompanying drawing explanation
Fig. 1 is the hardware configuration schematic diagram that the present invention is based on the circumferential weld visible detection method of decoupling zero detection and install a kind of embodiment assembling test platform;
Fig. 2 the present invention is based on the circumferential weld visible detection method of decoupling zero detection and installs the vision detection system elements relative position relationship schematic diagram in a kind of embodiment decoupling zero process;
Fig. 3--6 for the present invention is based on the position relationship of video camera 1 and weld seam 9 in circumferential weld visible detection method that decoupling zero detects and a kind of embodiment decoupling zero process of device and laser stripe at the schematic diagram of the position relationship of the plane of delineation;
Fig. 7 the present invention is based on the circumferential weld visible detection method of decoupling zero detection and installs the geometric graph that in a kind of embodiment decoupling zero process, weld seam Y-direction deviation is asked for;
Fig. 8 is the geometric graph that the present invention is based on the circumferential weld visible detection method of decoupling zero detection and install the partial enlarged drawing of Fig. 7 in a kind of embodiment decoupling zero process;
Fig. 9 the present invention is based on the circumferential weld visible detection method of decoupling zero detection and the analogous diagram of installing the abscissa of characteristic point while not adding decoupling algorithm there is no the starting the arc;
Figure 10 the present invention is based on the circumferential weld visible detection method of decoupling zero detection and the analogous diagram of installing the ordinate of characteristic point while not adding decoupling algorithm there is no the starting the arc;
Figure 11 is the analogous diagram that the present invention is based on circumferential weld visible detection method that decoupling zero detects and the device abscissa of characteristic point while adding decoupling algorithm not have the starting the arc;
Figure 12 is the analogous diagram that the present invention is based on circumferential weld visible detection method that decoupling zero detects and the device ordinate of characteristic point while adding decoupling algorithm not have the starting the arc;
Figure 13 is the analogous diagram that the present invention is based on the lateral deviation of welding gun while adding the decoupling algorithm starting the arc of circumferential weld visible detection method that decoupling zero detects and device;
Figure 14 is the analogous diagram that the present invention is based on circumferential weld visible detection method that decoupling zero detects and the device along track bias of welding gun while adding the decoupling algorithm starting the arc;
In the drawings, video camera 1, the second support 2, laser instrument 3, cross adjusting bracket 4, welding gun 5, V-butt annular weldment 6, fixture 7, the first supports 8, weld seam 9, the plane of delineation 10, straight plate weldment 11, laser stripe 12.
The specific embodiment
Below in conjunction with embodiment and accompanying drawing, the invention will be further described.
Circumferential weld visible detection method (the abbreviation method detecting based on decoupling zero of the present invention's design, referring to Fig. 1-14) be mainly used in industry spot, the detection of V-butt circular weld, identification and welding are carried out simultaneously, detection algorithm fully adapts to field condition, utilize two-dimentional weld image to obtain the positional information of three-dimensional weld seam, make not have coupled relation between the variation of characteristics of image and welding gun motion, the concrete steps of the method are:
The first step, image is processed and feature extraction
1-8 image acquisition: obtain the welded seam area image that comprises laser stripe by smart camera sensor;
The scope of the laser stripe of the image that 1-9 region of interesting extraction: determining step 1-1 obtains in image;
1-10 image threshold self adaptation is cut apart: on the basis of step 1-2, determine the available point of laser stripe;
1-11 laser structure light center line drawing: adopt the mode of extracting by column to extract V-butt circular weld center line array method;
The filtering of 1-12 center line and interpolation: the missing point in the laser structure light center line that interpolation completion is extracted, adopts gauss low frequency filter further to eliminate noise jamming afterwards;
1-13Hough straight line conversion: on the basis of step 1-5, the laser stripe after differential filtering is carried out to hough fitting a straight line;
1-14 feature point extraction: the ordinate value corresponding with the straight line obtaining after hough conversion according to the ordinate value of the center line group of the laser stripe extracting compares the flex point that obtains laser stripe, and the mid point of two flex points is exactly weld bead feature points.
Second step, image information decoupling zero
2-1 horizontal direction deviation identification: utilize the linear relationship of weld width and weld seam developed width W on the plane of delineation to obtain the departure of welding-torch horizontal direction;
2-1 vertical direction deviation identification: obtain the departure of welding gun vertical direction in the space geometry relation of vertical direction formation according to the weld seam on the plane of delineation and the welding gun in space.
The inventive method is further characterized in that the overall diameter of described V-butt annular weldment 6 is greater than 180mm, and groove width is 10-20mm, and groove depth is greater than 4mm.
The inventive method is further characterized in that described weld bead feature points is the mid point of two flex point lines of weld image V-butt.
The inventive method is further characterized in that the mid point of the two flex point lines that described weld seam initial characteristics point is weld image V-butt; In the situation that the not starting the arc and weldment do not rotate, limited N (5<N<10) two field picture in vision system acquisition time sequence, records the image coordinate at every two field picture plane V-butt center i=1,2 ... N, and the arithmetic mean of instantaneous value that calculates N frame center point value obtains weld seam initial characteristics point (x int, y int), computing formula is (1) formula
x Int = 1 N &Sigma; i = 1 N x mid i y Int = 1 N &Sigma; i = 1 N y mid i - - - ( 1 ) .
The inventive method is further characterized in that described horizontal departure knows method for distinguishing and be: adopt the method for real-time calibration to constantly update the ratio of weld width on weld seam developed width and the plane of delineation, every collection one two field picture recalculates horizontal direction proportionality coefficient k xdvalue, establish the position (x of weld seam initial characteristics point int, y int) and the position coordinates of weld seam n frame characteristic point the characteristic point coordinate that wherein n two field picture horizontal departure calculates for this frame with the welding gun initial characteristics point coordinates (x having calculated int, y int) compare, obtain plane of delineation weld seam deviation amount, by line style, convert, welding gun is in the horizontal direction with respect to the departure of actual welds, computing formula is (4) formula
ΔX=k xd (n)×δx (4)。
The welding gun that the further feature of the inventive method is also the identification of described vertical direction deviation is (8) formula with respect to the computing formula of the departure of actual welds in the vertical direction
AD=AA 1/cosθ=AB/(cos(90°-α)×cosθ) (8)。
The present invention adopts the circumferential weld vision inspection apparatus detecting based on decoupling zero of above-mentioned detection method (to be called for short device, referring to Fig. 1-2) comprise vision detection system, the first support 8, cross adjusting bracket 4, welding gun 5 and fixture 7, described vision detection system comprises video camera 1, the second support 2 and laser instrument 3, video camera 1 is connected with cross adjusting bracket 4 by the first support 8, by the second support 2, is fixedly linked with laser instrument 3; Cross adjusting bracket 4 bottoms are connected with welding gun 5 tops, and cross adjusting bracket 4 can drive welding gun 5 to move together, and laser instrument 3 is apart from welding gun 5 the place ahead 5-10cm; Below welding gun 5, be provided with fixture 7, by the scroll chuck on fixture 7, can fix V-butt annular weldment 6.
Apparatus of the present invention are further characterized in that described laser instrument 3 and video camera 1 optical axis included angle α=45 °, laser instrument 3 is installed 20 ° of angle theta ≈ with vertical direction, video camera 1 is to the distance L on V-butt annular weldment 6 between 180mm~200mm, and the horizontal range of the laser stripe on pad and V-butt annular weldment 6 is apart between 45mm~50mm.
The further feature of apparatus of the present invention is also that described laser instrument 3 is laser line generator.
The present invention should reflect weld seam horizontal departure when choosing weld bead feature points, reflects again weld seam vertical missing, and the mid point that considers the two flex point lines of selecting weld image V-butt is the most suitable as weld bead feature points.
Vision detection system and welding gun 5 are rigidly connected, and before welding starts, welding gun are adjusted to suitable welding initial position, on V-butt center line; In the situation that the not starting the arc and weldment do not rotate, limited N (5<N<10) two field picture (N selected value 6 in this example) in vision system acquisition time sequence, records the image coordinate at every two field picture plane V-butt center i=1,2 ... N, and the arithmetic mean of instantaneous value that calculates N frame center point value obtains weld seam initial characteristics point (x int, y int), computing formula is (1) formula, so, and acquiescence weld seam initial characteristics point (x int, y int) be that welding gun is in the position of the plane of delineation 10;
x Int = 1 N &Sigma; i = 1 N x mid i y Int = 1 N &Sigma; i = 1 N y mid i - - - ( 1 ) .
Apparatus of the present invention are separated (referring to Fig. 2) by vision detection system, welding gun, V-ring seam weldment from test platform, embodiment illustrated in fig. 2 showing, laser instrument 3 and video camera 1 optical axis included angle α=45 °, laser instrument 3 is installed 20 ° of angle theta ≈ with vertical direction, video camera 1 is to the distance L on V-butt annular weldment 6 between 180mm~200mm, and the horizontal range of the laser stripe 12 on pad and the annular weldment 6 of V-butt is between 45mm~50mm.
The change in location of video camera 1 of the present invention and laser stripe 12 variation relation (referring to Fig. 3-6) on the plane of delineation 10, adopt this device, straight plate weldment 11 with weld seam 9 is placed on fixture 7, open laser instrument 3, in the vertical direction of weld seam 9, occur a laser stripe 12, video camera 1 is above laser stripe 12.
1) when the left side of video camera 1 at weld seam 9, on the plane of delineation 10, the groove of laser stripe 12 is positioned at the right side (referring to Fig. 3) of the plane of delineation 10, when video camera 1 level moves to right a segment distance, video camera 1 is when the right side of weld seam 9, and the groove of the laser stripe 12 on the plane of delineation 10 is understood horizontal left (referring to Fig. 4);
2) when video camera 1 is above weld seam 9, on the plane of delineation 10, laser stripe 12, in the position (referring to Fig. 5) of side on the lower side of the plane of delineation 10, when video camera 1 moves a certain distance straight down, moves (referring to Fig. 6) in laser stripe 12 meetings vertically on image 10.
From the analysis of weld image (referring to 3-6), the point that can reflect weld width information is the intersection point that deformation part and undeformed part occur laser stripe 12, i.e. two intersection points of both sides straight line portion and intermediate curve part.Extraction for this point can adopt the method that starts within the scope of certain length, center line array to be asked first derivative from image on one side, but this method is inapplicable to irregular situation in welding bead.
So, the feature of combining image, and the straight line of matching after hough conversion above, determine flex point, and then definite weld bead feature points, and concrete steps are:
A. according to system accuracy, setting threshold T c;
B. from image one side, start search, according to y seam(t)-y i(t) formula judges whether viewpoint to > T (i=1,2), if difference is greater than threshold value, records this point for x 1acontinue search, when difference is less than threshold value, recording this point is x 1b, these two points are one group, record successively x 1a, x 1b, x 2a, x 2b
Wherein, y seam(t) be the ordinate of the centerline points that lists of t, y 1(t) be that t is listed as the ordinate that the left-half of laser stripe is carried out to the Points on Straight Line that hough conversion obtains, y 2(t) be that t is listed as the ordinate that the right half part of laser stripe is carried out to the Points on Straight Line that hough conversion obtains, if t is in the left-half of laser stripe, with y 1(t) value compares, if at the right half part of laser stripe, should with y 2(t) value compares.
The position of the trisection point of the line segment c. every group of array being become is detected, and determines whether groove (as the groove of the laser stripe 12 on the plane of delineation 10), and no, and one group of point is rejected;
It is little that the deformation quantity that can cause than weld seam in situation is removed in the deformation one of the structured light center line d. causing due to interfere information, therefore finds two numbers can be judged as striped flex point apart from a group of maximum, and image coordinate corresponding to two flex points of V-butt is
E. the mid point of flex point is exactly weld bead feature points.
Can occur to change according to the horizontal level of the groove of laser stripe 12 on the plane of delineation 10 accordingly showing that welding gun 5 exists deviation in the horizontal direction for weld seam 9, the vertical position of the groove of the laser stripe 12 on the plane of delineation 10 has occurred to change and has shown that welding gun 5 is for have deviation on weld seam 9 vertical directions.
(1) horizontal direction deviation identification
In the horizontal direction, can regard the demarcation on two dimensional surface as, on the plane of delineation, weld width wpixel and weld seam developed width W (known) are linear, and wherein weld seam developed width W is measured value, and be fixed value, in welding process, can be considered invariant.Adopt the method for real-time calibration to constantly update the ratio of weld width wpixel on weld seam developed width W and the plane of delineation, every collection one two field picture recalculates horizontal direction proportionality coefficient k xdvalue.If the position (x of weld seam initial characteristics point int, y int) and weld seam before the position coordinates of frame characteristic point the characteristic point coordinate that wherein n two field picture horizontal departure calculates for this frame with the welding gun initial characteristics point coordinates (x having calculated int, y int) compare, obtain plane of delineation weld seam deviation amount, by line style, convert, obtain welding gun in the horizontal direction with respect to the departure of actual welds, computing formula is (2), (3), (4) formula, concrete steps are:
A. calculate proportionality coefficient
k xd ( n ) = W / | x 2 ( n ) - x 1 ( n ) | - - - ( 2 )
B. obtain plane of delineation weld seam deviation amount
&delta;x = x mid ( n ) - x Int - - - ( 3 )
C. calculate output controlled quentity controlled variable, welding gun is in the horizontal direction with respect to the departure of actual welds
ΔX=k xd (n)×δx (4)
(2) vertical direction deviation identification
In actual welding, due to the position deviation of V-butt annular weldment fixture clamping, or due to the cause of backing welding dislocation, there will be the annular weldment geometrical axis of rotation and V-butt angled, in rotary course, weld seam phase Butt welding gun there will be larger departure.The algorithm of a kind of effective practicality of this employing is obtained to the regulated quantity of welding gun vertical direction, reach the welding effect of expection.
The algorithm (referring to Fig. 7-8) of vertical direction deviation of the present invention identification, the position of weld seam 9 when curve S 1 is initial acquisition, curve S 2 is the image of weld seam 9 spatial position change that collect of n frame.Former laser structure light and V-butt annular weldment meet at A point, and rear because V-butt annular weldment 6 is when rotating, the position of weld seam 9 can change, and intersects at A with curve S 2 1point.Straight line L1 is that the light, the straight line L2 that while taking n two field picture, reflex to camera review plane reflex to the light of camera review plane 10 while taking n+1 two field picture, because video camera 1 is 180mm left and right with the distance of V-butt annular weldment 6, therefore deviation maximum is in 10mm, and two light straight line L1 and straight line L2 can be similar to and regard parallel lines as.Straight line L3 is the straight line at LASER Light Source place.α is the angle between laser instrument 3 and video camera 1 optical axis, and during actual Design of Vision System, in order to guarantee imaging effect, α can not be zero, and θ is that laser instrument 3 vertical direction are installed angle, is measurable magnitude, after vision system fixed installation, is given value.
If PQ is the plane vertical with video camera 1 optical axis, by experiment, record and knownly on the vertical two dimensional surface of camera optical axis, follow linear transformation, and k xd≈ k yd.AB line segment the value of the plane of delineation by laser stripe straight line longitudinally offset deviation embody, the physical length of line segment AB is AA 1length, can try to achieve according to formula (5), (6), (7) formula,
A. obtain AB line chart as plane welded seam deviate
&delta;y = y mid ( n ) - y Int - - - ( 5 )
B. calculate AB
AB=k xd (n)×δy (6)
C. pass through geometric operation, AA 1by formula (7), tried to achieve
AA 1=AB/cos(90°-α) (7)
D.A 1d is the straight line vertical with LASER Light Source projection plane, hands over AC in D point (referring to Fig. 8), calculates AD length, is welding gun in the vertical direction with respect to the departure of actual welds:
AD=AA 1/cosθ=AB/(cos(90°-α)×cosθ) (8)
In short transverse, offset distance should be AC, therefore need to analyze the value of CD.In the situation that other welding surroundings are constant, the caliber of weldment is less, and CD value is larger; At weldment caliber one regularly, weldment side-play amount in the vertical direction of x-y plane is larger, and the value of CD is also larger.Therefore can obtain, the caliber of weldment and weldment side-play amount in the vertical direction of x-y plane is the factor of major effect CD value, and wherein weldment pipeline radius is the leading factor that affects CD.
If A 1d and A 1the angle of C direction tangent line is β, due to arc A 1c is very short, can be approximated to be straight line, therefore has following relational expression:
&angle; ADA 1 = &pi; 2 - &theta; - - - ( 9 )
&angle; ADA 1 = &pi; 2 - &theta; - &beta; - - - ( 10 )
A 1D=ADsinθ (11)
According to sine, obtain:
sin &angle; ACA 1 A 1 D = sin &beta; CD - - - ( 13 )
Bringing formula (10), (11) into formula (12) obtains:
CD = AD &times; sin &theta; &times; sin &beta; sin ( &pi; 2 - &theta; - &beta; ) - - - ( 13 )
Further arrange:
CD = sin &theta; sin &beta; cos ( &theta; + &beta; ) AD - - - ( 14 )
Because θ value is known, CD → 0, β → 0.β ≠ 0 o'clock, formula (14) can further arrange again and be:
CD = 1 cot &theta; cot &beta; - 1 AD - - - ( 15 )
Now, CD is along with the angle of β increases and increases, and β value is affected by weldment pipeline radius and eccentric distance, the minimum pipe that native system is measured is 180mm, the welding system maximum deviation that requires to rotate a circle is no more than 10mm, now β maximum is 33 °, CD ≈ 0.3AD now, known by experiment, within the scope of the maximum deviation allowing at welding system, the AD value that every frame draws is below 1mm, so CD length is less than 0.3mm, and this value does not add up, next frame obtains new less value while calculating.Therefore, can not consider the impact that CD regulates system.
Embodiment 1
The present embodiment connects device on request, and selecting the overall diameter of V-butt annular weldment is 195mm, and groove width is 14mm, and groove depth is 7mm.
First open camera power supply, mobile welding gun is to initial weld position, and simultaneous camera 1 gathers image and processes, and obtains weld image coordinate and welding seam position information, the reference picture coordinate that records weld seam initial characteristics point, can start soldering joint automatic tracking function after completing.In soldering joint automatic tracking process, weld image coordinate of each sampling period extraction, as the value of feedback of current position while welding, by the decoupling zero detection algorithm based on the plane of delineation, is directly calculated the actual deviation amount on welding gun 5 both directions.
Vision detection system and welding gun 5 are rigidly connected, and before welding starts, welding gun are adjusted to suitable welding position, on V-butt center line.In the situation that the not starting the arc and weldment do not rotate, limited 6 two field pictures in vision detection system acquisition time sequence, the image coordinate at every two field picture plane V-butt center is respectively (296,455), (295,455), (296,455), (296,455), (296,455), (296,455).And by formula (1), the arithmetic mean of instantaneous value that calculates 6 frame center's point values obtains (296,455).So, this point is defaulted as welding gun 5 in the position of the plane of delineation 10, i.e. the reference picture coordinate of weld seam initial characteristics point.
x Int = 1 N &Sigma; i = 1 N x mid i y Int = 1 N &Sigma; i = 1 N y mid i - - - ( 1 ) .
Wherein, for the coordinate of every two field picture, i=1,2 ... N, (x int, y int) weld seam initial characteristics dot image coordinate.
(1) horizontal departure identification
In the horizontal direction, can regard the demarcation on two dimensional surface as, on the plane of delineation, weld width wpixel and weld seam developed width W (known) are linear, wherein W is measured value, and be fixed value, in welding process, can be considered invariant, in the present embodiment W=14.Adopt the method continuous renewal weld seam developed width of real-time calibration and the ratio of plane of delineation weld width, every collection one two field picture recalculates horizontal direction proportionality coefficient k xdvalue.The position (296 of weld seam initial characteristics point, 455), V-butt annular weldment 6 continuous capturing pictures of video camera 1 in the process of rotation, chooses 5 frames wherein in this example, its coordinate is respectively (295,424), (291,426), (294,425), (295,426), (293,425), the coordinate of this 5 two field picture is made comparisons with weld seam initial characteristics point coordinates (296,455) respectively, obtain plane of delineation departure, by line style, convert, obtain welding gun 5 in the horizontal direction with respect to the departure of actual welds.Can try to achieve by formula (2), (3) and (4), wherein the coordinate of two flex points of present frame weld seam, position coordinates for weld seam present frame characteristic point:
A. calculate proportionality coefficient
k xd ( n ) = W / | x 2 ( n ) - x 1 ( n ) | - - - ( 2 )
B. obtain pixel planes weld seam deviation value.
&delta;x = x mid ( n ) - x Int - - - ( 3 )
C. calculate output controlled quentity controlled variable, welding gun 5 is in the horizontal direction with respect to the departure of actual welds
ΔX (n)=k xd (n)×δx (4)
By above step, obtain Δ X (1)=-6, Δ X (2)=-2, Δ X (3)=-1, Δ X (4)=-4, Δ X (5)=-4.
(2) vertical direction deviation identification (referring to Fig. 7-8)
In actual welding, due to the position deviation of V-butt annular weldment 6 fixtures clampings, or due to the cause of backing welding dislocation, there will be annular weldment 6 geometrical axis of rotation and V-butt angled, in rotary course, weld seam phase Butt welding gun there will be larger departure.The algorithm of a kind of effective practicality of this employing is obtained to the regulated quantity of welding gun 5 vertical direction, reach the welding effect of expection.
If PQ is the plane vertical with video camera 1 optical axis, owing to following linear transformation on the vertical two dimensional surface of video camera 1 optical axis, and k xd≈ k yd.AB line segment the value of the plane of delineation by laser stripe straight line longitudinally offset deviation embody, i.e. AD length.Therefore the physical length of line segment AB can be tried to achieve according to (5)-(7),
A. obtain AB line chart as plane welded seam deviate
&delta;y = y mid ( n ) - y Int - - - ( 5 )
B. calculate AB
AB=k xd (n)×δy (6)
C. pass through geometric operation, AA 1by formula (7), tried to achieve
AA 1=AB/cos(90°-α) (7)
D.A 1d is the straight line vertical with LASER Light Source projection plane, hands over AC in D point, calculates AD length, is welding gun in the vertical direction with respect to the departure of actual welds:
AD=AA 1/cosθ=AB/(cos(90°-α)×cosθ) (8)
By checking above, CD can ignore on the impact of system, and like this, the vertical regulated quantity of welding gun that obtains 5 two field pictures by above step is respectively-2,2,0,2,0.
So far, obtained the welding gun 5 of 5 two field pictures in the side-play amount of level and vertical direction, again the validity of departure is judged, if coordinate does not surpass limit value, can be considered virtual value, PLC (programmable logic controller (PLC)) calculates output pulsed quantity according to both differences, by cross adjusting bracket 4, drives welding gun 5 to do correction adjustment.
In order to obtain controlling stably effect, need to introduce PID controller, this experiment need be controlled with increment type PID, because the output signal of controller is the pulse signal of stepper motor, therefore need to before PID controller, increase a pulse equivalency, namely eliminate the stepper motor umber of pulse that the image error of Yi Ge unit need to be exported.The increment type PID controller of tape pulse equivalent is as shown in formula (16):
p(k)=S{K p[e(k)-e(k-1)]+K ie(k)+K d[e(k)-2e(k-1)+e(k-2)]} (16)
Wherein, p (k) is k pulse output constantly, and S is pulse equivalency, K p, K i, and K drespectively ratio, integration, the differential gain of PID controller.In test, pid parameter is respectively: K p=0.5, K i=0.07 and K d=0.005.
In above-described embodiment, laser instrument used is laser line generator.
For the reliability of verification algorithm, while using rotation, serious eccentric test block is tested, and in the situation that not playing arc welding, while not adding decoupling algorithm, 100ms gathers a two field picture, records weld bead feature points at the coordinate figure of the plane of delineation 10.Obtain the abscissa value (referring to Fig. 9) of weld bead feature points and the curve map of ordinate value (referring to Figure 10), Figure 10 can find out that the ordinate of weld bead feature points has reached the position of high 480 pixels of image, to cause owing to regulating, the light-struck position of structure has exceeded the visual field of video camera 1, illustrates that the axle center of test block and axis of rotation angle are larger.The weld bead feature points being calculated by front 6 two field pictures is considered as welding gun 5 positions, while rotating so, can find out that welding gun 5 has obviously departed from weld seam 9 centers.
Use test part (the same), in the situation that not playing arc welding, adds decoupling algorithm, and 100ms gathers a two field picture, records weld bead feature points transverse and longitudinal coordinate figure, represents respectively with Figure 11, Figure 12; In the situation that playing arc welding, add decoupling algorithm, 100ms gathers a two field picture, records transversal displacement and vertical misalignment amount that welding gun need to move, represents respectively with Figure 13 and Figure 14.
Add after decoupling algorithm, constantly Butt welding gun 5 carries out horizontal and vertical adjustment, welding gun reaches 6 pixels (referring to Figure 11) in the deviation maximum of plane of delineation x direction, current Proportional coefficient K 1=0.075, and welding gun maximum deviation is 0.45mm (referring to Figure 13).In y direction, maximum deviation is 11 pixels (referring to Figure 12), and short transverse maximum deviation is 0.75mm (referring to Figure 14).And known by Figure 13 and Figure 14, the calculating of Butt welding gun side-play amount, meets rule, illustrate that decoupling zero detection algorithm realized the accurate positioning control of welding gun to three-dimensional weld seam.
Equipment related in above-described embodiment all can be by commercially available.
The inventive method is mainly used in industry spot, the detection of V-butt circular weld, identification and welding are carried out simultaneously, detection algorithm fully adapts to field condition, utilize two-dimentional weld image to obtain the positional information of three-dimensional weld seam, make not have coupled relation between the variation of characteristics of image and welding gun motion.Obtaining of the horizontal and vertical deviation of directivity for weld seam is that each two field picture obtaining according to video camera after the starting the arc 1 compares with the center of the V-butt circular weld that starting the arc does not obtain before, the deviate obtaining.Because welding surroundings is very complicated, in welding process may due to light, arc light, impact reflective, the factor such as splash cause the center of the current frame image that video camera 1 obtains may be inaccurate, if front and back two two field pictures compare, may make deviation more and more inaccurate, if but make comparisons with the center that starting the arc does not obtain before, even the incorrect deviation that also can not have influence on next frame image weld seam of the deviation that present frame obtains.Improved the accuracy rate that weld seam deviation is corrected.Apparatus of the present invention can provide corresponding warning message according to the anomaly occurring in welding process, also can observe the relevant information in welding process by display, the extraction and the location that only need video camera 1 (comprising camera lens, optical filter etc.) and laser instrument 3 just can complete weld seam information.
The present invention does not address part and is all applicable to prior art.

Claims (9)

1. the circumferential weld visible detection method detecting based on decoupling zero, for detection, the identification of the weld seam of V-butt annular weldment and weld, is characterized in that the concrete steps of the method are:
The first step, image is processed and feature extraction
1-1 image acquisition: obtain the welded seam area image that comprises laser stripe by smart camera sensor;
The scope of the laser stripe of the image that 1-2 region of interesting extraction: determining step 1-1 obtains in image;
1-3 image threshold self adaptation is cut apart: on the basis of step 1-2, determine the available point of laser stripe;
1-4 laser structure light center line drawing: adopt the mode of extracting by column to extract V-butt circular weld center line array method;
The filtering of 1-5 center line and interpolation: the missing point in the laser structure light center line that interpolation completion is extracted, adopts gauss low frequency filter further to eliminate noise jamming afterwards;
1-6Hough straight line conversion: on the basis of step 1-5, the laser stripe after differential filtering is carried out to hough fitting a straight line;
1-7 feature point extraction: the ordinate value corresponding with the straight line obtaining after hough conversion according to the ordinate value of the center line group of the laser stripe extracting compares the flex point that obtains laser stripe, and the mid point of two flex points is exactly weld bead feature points;
Second step, image information decoupling zero
2-1 horizontal direction deviation identification: utilize the linear relationship of weld width and weld seam developed width on the plane of delineation to obtain the departure of welding-torch horizontal direction;
2-1 vertical direction deviation identification: obtain the departure of welding gun vertical direction in the space geometry relation of vertical direction formation according to the weld seam on the plane of delineation and the welding gun in space.
2. the circumferential weld visible detection method detecting based on decoupling zero according to claim 1, is characterized in that the overall diameter of described V-butt annular weldment is greater than 180mm, and groove width is 10-20mm, and groove depth is greater than 4mm.
3. the circumferential weld visible detection method detecting based on decoupling zero according to claim 1, is characterized in that described weld bead feature points is the mid point of two flex point lines of weld image V-butt.
4. the circumferential weld visible detection method detecting based on decoupling zero according to claim 3, it is characterized in that in the situation that the not starting the arc and weldment do not rotate, limited N two field picture in vision system acquisition time sequence, the span of N is 5<N<10, records the image coordinate at every two field picture plane V-butt center i=1,2 ... N, and the arithmetic mean of instantaneous value that calculates N frame center point value obtains weld seam initial characteristics point (x int, y int), computing formula is (1) formula:
x Int = 1 N &Sigma; i = 1 N x mid i y Int = 1 N &Sigma; i = 1 N y mid i - - - ( 1 ) .
5. the circumferential weld visible detection method detecting based on decoupling zero according to claim 1, it is characterized in that described horizontal departure knows method for distinguishing and be: adopt the method for real-time calibration to constantly update the ratio of weld width on weld seam developed width and the plane of delineation, every collection one two field picture recalculates horizontal direction proportionality coefficient k xdvalue, establish the position (x of weld seam initial characteristics point int, y int) and the position coordinates of weld seam n frame characteristic point the characteristic point coordinate that wherein n two field picture horizontal departure calculates for this frame with the welding gun initial characteristics point coordinates (x having calculated int, y int) compare, obtain plane of delineation weld seam deviation amount and convert by line style, obtain welding gun in the horizontal direction with respect to the departure of actual welds, computing formula is (4) formula:
ΔX=k xd (n)×δx (4)。
6. the circumferential weld visible detection method detecting based on decoupling zero according to claim 1, the welding gun that it is characterized in that described vertical direction deviation identification is (8) formula with respect to the computing formula of the departure of actual welds in the vertical direction:
AD=AA 1/cosθ=AB/(cos(90°-α)×cosθ) (8)。
7. the circumferential weld vision inspection apparatus detecting based on decoupling zero that adopts above-mentioned detection method, it is characterized in that this device comprises vision detection system, the first support, cross adjusting bracket, welding gun and fixture, described vision detection system comprises video camera, the second support and laser instrument, video camera is connected with cross adjusting bracket by the first support, by the second support and laser instrument, is fixedly linked; Cross adjusting bracket bottom is connected with welding gun top, and cross adjusting bracket can drive welding gun to move together, and laser instrument is apart from welding gun the place ahead 5-10cm; Below welding gun, be provided with fixture, by the scroll chuck on fixture, can fix V-butt annular weldment.
8. the circumferential weld vision inspection apparatus detecting based on decoupling zero according to claim 7, it is characterized in that described laser instrument and camera optical axis angle α=45 °, laser instrument and vertical direction are installed 20 ° of angle theta ≈, video camera is to the distance L on V-butt annular weldment between 180~200mm, and the horizontal range of the laser stripe on pad and the annular weldment of V-butt is between 45~50mm.
9. according to the circumferential weld vision inspection apparatus detecting based on decoupling zero described in claim 7 or 8, it is characterized in that described laser instrument is laser line generator.
CN201410279698.5A 2014-06-20 2014-06-20 A kind of circumferential weld visible detection method detected based on decoupling zero Active CN104084669B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410279698.5A CN104084669B (en) 2014-06-20 2014-06-20 A kind of circumferential weld visible detection method detected based on decoupling zero

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410279698.5A CN104084669B (en) 2014-06-20 2014-06-20 A kind of circumferential weld visible detection method detected based on decoupling zero

Publications (2)

Publication Number Publication Date
CN104084669A true CN104084669A (en) 2014-10-08
CN104084669B CN104084669B (en) 2016-02-03

Family

ID=51632486

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410279698.5A Active CN104084669B (en) 2014-06-20 2014-06-20 A kind of circumferential weld visible detection method detected based on decoupling zero

Country Status (1)

Country Link
CN (1) CN104084669B (en)

Cited By (29)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104722922A (en) * 2014-12-23 2015-06-24 张家港浦项不锈钢有限公司 Dynamic welding method of laser welding machine
CN104816072A (en) * 2015-05-15 2015-08-05 大连大学 Method and system for laser measurement welding tracking of tiny joint seams of complex curved surface thin wall board
CN104933242A (en) * 2015-06-11 2015-09-23 北京卫星制造厂 Pipeline welding line information extraction method based on three-dimensional design model
CN105157603A (en) * 2015-07-29 2015-12-16 华南理工大学 Line laser sensor and method for calculating three-dimensional coordinate data of line laser sensor
CN105458462A (en) * 2015-12-22 2016-04-06 河北工业大学 Variable gap trapezoidal welding line multi-parameter synchronous visual inspection tracking method
CN106124177A (en) * 2016-06-29 2016-11-16 明基材料有限公司 The monitoring method of roller rotation and monitoring system
CN106141426A (en) * 2015-04-28 2016-11-23 大族激光科技产业集团股份有限公司 A kind of optical imaging system followed the trail of for laser welded seam
CN106141382A (en) * 2016-07-01 2016-11-23 江苏科技大学 Bevel for welding marginal position visual sensing detection method
CN107957247A (en) * 2017-12-22 2018-04-24 易思维(天津)科技有限公司 A kind of hand-held gap, face difference vision measurement device and method
CN108568624A (en) * 2018-03-29 2018-09-25 东风贝洱热系统有限公司 A kind of mechanical arm welding system and welding method based on image procossing
CN108673005A (en) * 2018-07-20 2018-10-19 常州机电职业技术学院 Deviation-rectifying type automatic welding line tracking system and working method thereof
CN108765419A (en) * 2018-06-20 2018-11-06 天津大学 Structure light vision weld image information self-adapting extracting method
CN108994488A (en) * 2018-07-13 2018-12-14 上海拓璞数控科技股份有限公司 Welding seam tracking method, system, device and readable storage medium storing program for executing
CN109741386A (en) * 2018-12-26 2019-05-10 豪威科技(武汉)有限公司 The Enhancement Method and stereo visual system of stereo visual system
CN110814465A (en) * 2019-11-28 2020-02-21 华东交通大学理工学院 Universal method for automatically extracting welding seam contour
CN111055054A (en) * 2020-01-13 2020-04-24 北京博清科技有限公司 Welding seam identification method and device, welding robot and storage medium
CN111405957A (en) * 2017-11-30 2020-07-10 吴圣圭 Welding automation system using shape and three-dimensional coordinates of welding part and welding method using the same
CN111633358A (en) * 2020-05-25 2020-09-08 西咸新区大熊星座智能科技有限公司 Laser-based weld parameter measuring method and device
CN111968187A (en) * 2020-10-21 2020-11-20 三代光学科技(天津)有限公司 Annular structure optical parameter calibration method
CN112161929A (en) * 2020-10-27 2021-01-01 宁波积微速成智能科技有限公司 Circumferential weld detection device and method for frequency-doubling differential anti-interference specular reflection orthogonal imaging
CN112365439A (en) * 2020-09-24 2021-02-12 南昌大学 Method for synchronously detecting forming characteristics of GMAW (gas metal arc welding) weld of galvanized steel and welding gun direction in real time
CN112355478A (en) * 2020-10-30 2021-02-12 北京博清科技有限公司 Welding control method and device and electronic equipment
CN112621033A (en) * 2020-12-10 2021-04-09 北京博清科技有限公司 Welding gun positioning method, device, equipment and storage medium
CN112621034A (en) * 2020-12-10 2021-04-09 北京博清科技有限公司 Welding gun positioning method, device, equipment and storage medium
US20210154773A1 (en) * 2019-11-22 2021-05-27 GM Global Technology Operations LLC Real-Time Weld Quality Analysis Systems And Methods
CN114043081A (en) * 2021-11-24 2022-02-15 苏州全视智能光电有限公司 Laser welding multi-weld type feature point identification method and system
CN114473153A (en) * 2022-04-01 2022-05-13 电王精密电器(北京)有限公司 Oil-gas long-distance pipeline welding system and method
CN115138949A (en) * 2022-08-29 2022-10-04 南通泰胜蓝岛海洋工程有限公司 High-position circular seam submerged-arc welding system for large steel pipe pile and control method of high-position circular seam submerged-arc welding system
CN115781094A (en) * 2022-11-29 2023-03-14 天津工业大学 Welding seam deviation detection method and device

Citations (5)

* Cited by examiner, † Cited by third party
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
CN102049638A (en) * 2010-07-27 2011-05-11 天津大学 Circular seam welding robot device
CN103071902A (en) * 2013-01-11 2013-05-01 北京工业大学 Method and device for controlling welding process based on stability of perforated molten bath
CN103600155A (en) * 2013-10-25 2014-02-26 上海工程技术大学 Real-time monitoring device for welding robot welding process
CN203973025U (en) * 2014-06-20 2014-12-03 河北工业大学 A kind of circumferential weld vision inspection apparatus detecting based on decoupling zero

Patent Citations (5)

* Cited by examiner, † Cited by third party
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
CN102049638A (en) * 2010-07-27 2011-05-11 天津大学 Circular seam welding robot device
CN103071902A (en) * 2013-01-11 2013-05-01 北京工业大学 Method and device for controlling welding process based on stability of perforated molten bath
CN103600155A (en) * 2013-10-25 2014-02-26 上海工程技术大学 Real-time monitoring device for welding robot welding process
CN203973025U (en) * 2014-06-20 2014-12-03 河北工业大学 A kind of circumferential weld vision inspection apparatus detecting based on decoupling zero

Cited By (42)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104722922A (en) * 2014-12-23 2015-06-24 张家港浦项不锈钢有限公司 Dynamic welding method of laser welding machine
CN104722922B (en) * 2014-12-23 2017-12-15 张家港浦项不锈钢有限公司 Laser welder dynamic welding method
CN106141426A (en) * 2015-04-28 2016-11-23 大族激光科技产业集团股份有限公司 A kind of optical imaging system followed the trail of for laser welded seam
CN104816072A (en) * 2015-05-15 2015-08-05 大连大学 Method and system for laser measurement welding tracking of tiny joint seams of complex curved surface thin wall board
CN104933242B (en) * 2015-06-11 2017-09-22 北京卫星制造厂 A kind of pipeline weld information extracting method based on Three Dimensional Design Model
CN104933242A (en) * 2015-06-11 2015-09-23 北京卫星制造厂 Pipeline welding line information extraction method based on three-dimensional design model
CN105157603B (en) * 2015-07-29 2017-12-01 华南理工大学 A kind of line laser sensor
CN105157603A (en) * 2015-07-29 2015-12-16 华南理工大学 Line laser sensor and method for calculating three-dimensional coordinate data of line laser sensor
CN105458462A (en) * 2015-12-22 2016-04-06 河北工业大学 Variable gap trapezoidal welding line multi-parameter synchronous visual inspection tracking method
CN105458462B (en) * 2015-12-22 2017-08-29 河北工业大学 A kind of trapezoidal weld seam multi-parameter synchronizing visual detecting and tracking method of Varied clearance
CN106124177A (en) * 2016-06-29 2016-11-16 明基材料有限公司 The monitoring method of roller rotation and monitoring system
CN106124177B (en) * 2016-06-29 2018-08-31 明基材料有限公司 The monitoring method and monitoring system of idler wheel rotation
CN106141382A (en) * 2016-07-01 2016-11-23 江苏科技大学 Bevel for welding marginal position visual sensing detection method
CN111405957A (en) * 2017-11-30 2020-07-10 吴圣圭 Welding automation system using shape and three-dimensional coordinates of welding part and welding method using the same
CN107957247A (en) * 2017-12-22 2018-04-24 易思维(天津)科技有限公司 A kind of hand-held gap, face difference vision measurement device and method
CN107957247B (en) * 2017-12-22 2023-09-08 易思维(天津)科技有限公司 Handheld gap and surface difference vision measurement device and method
CN108568624A (en) * 2018-03-29 2018-09-25 东风贝洱热系统有限公司 A kind of mechanical arm welding system and welding method based on image procossing
CN108765419A (en) * 2018-06-20 2018-11-06 天津大学 Structure light vision weld image information self-adapting extracting method
CN108765419B (en) * 2018-06-20 2022-09-06 天津大学 Structured light vision welding seam image information self-adaptive extraction method
CN108994488A (en) * 2018-07-13 2018-12-14 上海拓璞数控科技股份有限公司 Welding seam tracking method, system, device and readable storage medium storing program for executing
CN108673005A (en) * 2018-07-20 2018-10-19 常州机电职业技术学院 Deviation-rectifying type automatic welding line tracking system and working method thereof
CN109741386A (en) * 2018-12-26 2019-05-10 豪威科技(武汉)有限公司 The Enhancement Method and stereo visual system of stereo visual system
US11633813B2 (en) * 2019-11-22 2023-04-25 GM Global Technology Operations LLC Real-time weld quality analysis systems and methods
US20210154773A1 (en) * 2019-11-22 2021-05-27 GM Global Technology Operations LLC Real-Time Weld Quality Analysis Systems And Methods
CN110814465B (en) * 2019-11-28 2021-09-03 华东交通大学理工学院 Universal method for automatically extracting welding seam contour
CN110814465A (en) * 2019-11-28 2020-02-21 华东交通大学理工学院 Universal method for automatically extracting welding seam contour
CN111055054A (en) * 2020-01-13 2020-04-24 北京博清科技有限公司 Welding seam identification method and device, welding robot and storage medium
CN111055054B (en) * 2020-01-13 2021-11-16 北京博清科技有限公司 Welding seam identification method and device, welding robot and storage medium
CN111633358B (en) * 2020-05-25 2021-11-09 西咸新区大熊星座智能科技有限公司 Laser-based weld parameter measuring method and device
CN111633358A (en) * 2020-05-25 2020-09-08 西咸新区大熊星座智能科技有限公司 Laser-based weld parameter measuring method and device
CN112365439A (en) * 2020-09-24 2021-02-12 南昌大学 Method for synchronously detecting forming characteristics of GMAW (gas metal arc welding) weld of galvanized steel and welding gun direction in real time
CN112365439B (en) * 2020-09-24 2023-10-03 南昌大学 Method for synchronously detecting forming characteristics of GMAW welding seam of galvanized steel and direction of welding gun in real time
CN111968187A (en) * 2020-10-21 2020-11-20 三代光学科技(天津)有限公司 Annular structure optical parameter calibration method
CN112161929A (en) * 2020-10-27 2021-01-01 宁波积微速成智能科技有限公司 Circumferential weld detection device and method for frequency-doubling differential anti-interference specular reflection orthogonal imaging
CN112355478A (en) * 2020-10-30 2021-02-12 北京博清科技有限公司 Welding control method and device and electronic equipment
CN112621033A (en) * 2020-12-10 2021-04-09 北京博清科技有限公司 Welding gun positioning method, device, equipment and storage medium
CN112621034A (en) * 2020-12-10 2021-04-09 北京博清科技有限公司 Welding gun positioning method, device, equipment and storage medium
CN114043081A (en) * 2021-11-24 2022-02-15 苏州全视智能光电有限公司 Laser welding multi-weld type feature point identification method and system
CN114043081B (en) * 2021-11-24 2023-12-22 苏州全视智能光电有限公司 Multi-weld-joint type feature point identification method and system for laser welding
CN114473153A (en) * 2022-04-01 2022-05-13 电王精密电器(北京)有限公司 Oil-gas long-distance pipeline welding system and method
CN115138949A (en) * 2022-08-29 2022-10-04 南通泰胜蓝岛海洋工程有限公司 High-position circular seam submerged-arc welding system for large steel pipe pile and control method of high-position circular seam submerged-arc welding system
CN115781094A (en) * 2022-11-29 2023-03-14 天津工业大学 Welding seam deviation detection method and device

Also Published As

Publication number Publication date
CN104084669B (en) 2016-02-03

Similar Documents

Publication Publication Date Title
CN104084669B (en) A kind of circumferential weld visible detection method detected based on decoupling zero
CN203973025U (en) A kind of circumferential weld vision inspection apparatus detecting based on decoupling zero
CN107186319B (en) A kind of online tracking of welding robot cosmetic welding based on laser sensor
CN110064819B (en) Cylindrical surface longitudinal weld characteristic region extraction and weld tracking method and system based on structured light
CN105458462B (en) A kind of trapezoidal weld seam multi-parameter synchronizing visual detecting and tracking method of Varied clearance
EP2906384B1 (en) Methods and systems for weld control
Nayak et al. Intelligent seam tracking for robotic welding
CN106624262A (en) Method and device for intelligent welding of intersection curve of round pipes
JP5142775B2 (en) Welding quality inspection method and apparatus
US20090179021A1 (en) Welding robot
CN103008881A (en) Seam tracking method based on template matching
CN104588838A (en) Automatic tracking control and online correction system with welding gun and control method thereof
CN103231162A (en) Device and method for visual detection of welding quality of robot
CN104677301A (en) Helical weld pipe line outer diameter measuring device and method based on view detection
CN102284769A (en) System and method for initial welding position identification of robot based on monocular vision sensing
Sung et al. Development of a multiline laser vision sensor for joint tracking in welding
CN109822194A (en) A kind of weld tracker and welding method
Wu et al. Research on robust laser vision feature extraction method for fillet welds with different reflective materials under uncertain interference
JP2008302428A (en) Arc welding quality inspection method
KR101245767B1 (en) Method and system for tracking pipe welding seam
CN109128540A (en) A kind of T connector laser welding focal track determines method
Li et al. Narrow weld joint recognition method based on laser profile sensor
Wang et al. Feature information extraction method for narrow gap U-type groove based on laser vision
Cai et al. A structured light-based visual sensing system for detecting multi-layer and multi-track welding
KR101284852B1 (en) Apparatus for inspecting weld toe grinding and methord thereof

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