A method for automatically generating detection points for X-ray detection of head welds

Abstract

The present invention discloses a method of extracting the weld contour coordinates through a size picture of a head weld by relying on image processing technology, converting the picture pixel information into a geometric model, and finally obtaining the coordinates of the X-ray detection points of the head weld through corresponding calculations. This solution establishes a detection coordinate system in the picture, calculates the optimal detection coordinates of the weld in the picture, and finally maps the coordinates in the picture to the real detection coordinate system through the magnification ratio relationship. The beneficial effect of the present invention is that all detection points are automatically generated, which solves the problem of manual editing of points under existing detection conditions and greatly improves efficiency. At the same time, it can calculate the angle that the detector needs to rotate to the tangent position of the head weld, as well as the coordinates that the detector needs to move at this angle. Compared with manual naked eye observation, this method is simple to operate and has extremely high accuracy. It solves the difficulties and accuracy problems of editing points and ensures the accuracy of subsequent X-ray flaw detection.

Description

Technical Field

[0001] The present invention relates to the field of ray detection methods, and in particular to a method for automatically generating detection points during X-ray detection of a head weld. Background Art

[0002] With the continuous advancement of technology, X-ray nondestructive testing (NDT) has become widely used in inspecting tank container head welds. X-ray inspection uses a light tube to transmit X-rays through an object. A detector receives the rays and electronically converts them into a digital image. Inspectors can determine the weld's conformity based on the resulting image. However, since the weld on the head is curved and the detector is flat, the two surfaces don't fit perfectly. Therefore, finding the point where the detector surface intersects the weld curve and then capturing the image is crucial to ensure that no defects are missed.

[0003] The existing detection method uses a universal point programming method. Before testing each detection point, the inspector needs to enter the lead room, manually control the rotation angle of the detector, and determine the tangent position of the detector and the detection point by visual observation. At the same time, it is also necessary to manually measure the distance between the detector and the weld, and then move the detector to an equidistant position. When the position is confirmed to be correct, click the software to save the current point information, and repeat the above steps in sequence to complete the programming of all points. The edited point program is then sent to the controller in sequence, and the controller controls the corresponding hardware to move to the corresponding position according to the programmed coordinates, thereby realizing automatic detection.

[0004] The traditional method of editing points has three drawbacks: First, the inspector must visually confirm the position during programming, which can result in significant errors and easily lead to missed defects. Second, it is inefficient, as each inspection position must be manually adjusted before programming, which is very time-consuming. This is especially true when inspecting customized heads, as each head may have different dimensions, requiring time-consuming reprogramming for each head. Third, implementation is difficult, requiring repeated adjustments to the detector's rotation angle and the balance point at which the detector is equidistant from the weld.

[0005] Therefore, a method is needed to automatically generate the detection points of the head weld to improve the efficiency and accuracy of generating the point coordinates. Summary of the Invention

[0006] In response to the above-mentioned deficiencies in the prior art, the present invention provides a method for automatically generating detection points for X-ray detection of head welds, which automatically calculates the detection points based on the image, greatly improving the efficiency of X-ray detection of head welds.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] 1. A method for automatically generating detection points for X-ray detection of head welds, characterized by comprising the following steps:

[0009] S1. Use drawing software to draw a proportionally reduced cross-sectional outline of the weld according to the actual curvature of the head;

[0010] S2. Preprocessing the drawn contour image, including image filtering, image binarization, and obtaining the coordinate points of the inner contour and the outer contour;

[0011] S3. Eliminate the inner contour coordinate points and take the coordinates of two points with the same y coordinate, where the coordinate P should meet the following conditions:

[0012]

[0013] Then the inner contour coordinates are P1(x1, y), the outer contour coordinates are P2(x2, y),

[0014] S4. Based on the known conditions of the distance d between the detector and the cross-section of the weld inspection point, the focal length f between the detector and the light tube, and the width w of the detector imaging area, calculate the detectable weld length r at each inspection point. r satisfies the following formula:

[0015] r=(fd)*w / f

[0016] S5, take each detection point corresponding to the detectable contour coordinate starting point s (x i*r ,y i*r ), end point e(x i*r+r ,y i*r+r ), (where i represents the number of welds currently being tested, and r represents the length of the weld), according to the slope formula

[0017] k=-(x i*r+r -x i*r ) / (y i*r+r -y i*r )

[0018] S6. Then the rotation angle α of the detector and the horizontal coordinate x and vertical coordinate y of the detector can be obtained.

[0019] α=arctan(k)

[0020] x=d*cos(α)

[0021] y=d*sin(α)

[0022] S7, according to the known conditions, the actual radius of the head weld is R, the radius of the contour map drawn in step S1 is R1, and according to the proportional conversion formula, the real coordinates (x m ,y m ),

[0023] x m =R / R1*x

[0024] y m =R / R1*y

[0025] Furthermore, the cross-sectional profile of the head weld in step S1 is drawn using CAD software.

[0026] The beneficial effects of the present invention include: automatically generating the detection points and deflection angles of the weld, improving the detection accuracy, replacing manual positioning, and greatly improving the detection efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the detection structure of the present invention;

[0028] Figure 2 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0029] The present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0030] like Figure 1 The head weld inspection model shown in the figure includes an X-ray tube positioned above the arc-shaped head weld, and a detector positioned below the weld. The tube's ray-emitting port is always aligned with the detector's receiving plate. The detector receives a range of images, so during inspection, multiple segments are stitched together to form an image of the entire weld to determine weld quality. Specifically, the detector's inspection trajectory is divided into multiple inspection points, requiring precise positioning of the detector's position and rotation angle at each inspection point, which can be fed back to the motion control unit to drive the detector's movement.

[0031] In the present invention, the Figure 1 The method of automatically generating detection points for X-ray detection of head welds shown

[0032] One approach uses image processing technology to extract the weld contour coordinates from a size image of the head weld, converting the image pixel information into a geometric model. Finally, through appropriate calculations, the coordinates of the X-ray inspection points on the head weld are obtained. This approach establishes an inspection coordinate system within the image, calculates the optimal inspection coordinates for the weld within the image, and finally maps the image coordinates to the actual inspection coordinate system using a magnification ratio. The specific approach is as follows:

[0033] 1. Draw the size picture of the head weld and use CAD or other drawing software to draw a proportional reduction outline of the weld. It must be consistent with the curvature of the actual weld to ensure the accuracy of the software calculation.

[0034] 2. Preprocess the size image of the weld head, including image filtering, image binarization, and obtain the coordinate points of the inner and outer contours.

[0035] 3. Eliminate useless inner contour coordinate points and take the coordinates of two points with the same y coordinate, the inner contour coordinate P1 (x1, y), the outer contour coordinate P2 (x2, y), and the obtained coordinate P should meet the following conditions:

[0036]

[0037] 4. The distance d between the detector and the weld section, the focal length f between the imaging plate and the light tube, and the width w of the detector imaging area are known conditions. Based on the properties of similar triangles, the weld length r that the imaging plate can detect is calculated as shown in Formula 2

[0038] r=(fd)*w / f (2)

[0039] 5. Take the starting point s(x i*r ,y i*r ), end point e(x i*r+r ,y i*r+r ),(in i Indicates the weld section currently being tested. r As shown in step 4), k can be obtained as shown in formula 3

[0040] k=-(x i*r+r -x i*r ) / (y i*r+r -y i*r ) (3)

[0041] 6. Then the detector's rotation angle α and the detector's horizontal coordinate x and vertical coordinate y can be obtained, as shown in formulas 4, 5, and 6

[0042] α=arctan(k) (4)

[0043] x=d*cos(α) (5)

[0044] y=d*sin(α) (6)

[0045] 7. The coordinates obtained in the previous step are relative coordinates in the image and need to be mapped to the physical coordinate system. The actual radius R of the weld is a known condition. The radius R1 of the weld in the image can be calculated. Finally, the physical relative coordinates of the detector can be obtained by magnification conversion, that is, the actual relative coordinates of the detection position, as shown in Formula 7.

[0046] x m =R / R1*x

[0047] y m=R / R1*y (7).

[0048] The specific implementation process of this method is

[0049] 1. Transport the head to the inspection room and use the lifting and clamping mechanism to lift the head to the inspection position. During this process, it must be ensured that the head is level with the ground without tilt. This is equivalent to the reference platform of all coordinate systems. Any tilt here will directly lead to inaccurate inspection position.

[0050] 2. Use a tape measure to measure the following four length information related to the head weld, including the actual length L of the head weld measured by the tape measure, the distance d between the detector and the light tube, the distance f between the detector and the weld section, the focal length between the imaging plate and the light tube, and the width w of the detector imaging area.

[0051] 3. Draw the size picture of the head weld and use CAD or other drawing software to draw a proportionally reduced outline of the weld. It must be consistent with the curvature of the actual weld to ensure the accuracy of the software's later calculations.

[0052] 4. Import the generated size image and pre-process it to improve the accuracy of the measurement. First, use the mean filter kernel to perform convolution operation on the image to eliminate pseudo contour points. The specific filtering formula is as follows:

[0053]

[0054] Where, the filter kernel h is:

[0055]

[0056] Secondly, the convolution image is binarized. The binarization formula is as follows:

[0057]

[0058] Thresh is the threshold, which is usually 128.

[0059] 5. Through binarization, the contour coordinates of the weld can be obtained. These coordinates include the inner contour coordinates, which are useless coordinates and need to be eliminated. The elimination method is to take the coordinates of two points with the same y coordinate, the inner contour coordinate P1 (x1, y), the outer contour coordinate P2 (x2, y), and the obtained coordinate P should meet the conditions of formula 1

[0060] 6. The image range that the imaging plate can capture is small, while the length of the weld to be captured is very long, generally more than 1 meter. Therefore, the weld length r that can be detected by the imaging plate needs to be calculated, which can be calculated according to formula 2

[0061] 7. Take the starting point s(x i*r ,y i*r), end point e(x i*r+r ,y i*r+r ), (where i represents the weld segment currently being inspected, and r is obtained from Formula 2). Formula 3 allows us to obtain k, and thus the detector's rotation angle α, as well as the detector's horizontal coordinate x and vertical coordinate y, which are calculated using Formulas 4, 5, and 6 in the scheme.

[0062] 8. The coordinates obtained in the previous step are relative coordinates in the image and need to be mapped to the physical coordinate system. The actual radius R of the weld is a known condition. The radius R1 of the weld in the image can be calculated. Finally, the physical relative coordinates of the detector can be obtained through magnification conversion, that is, the actual relative coordinates of the detection position. The obtained detector coordinates are calculated by Formula 7 in the solution and stored in the database. At the same time, they are sent to the hardware control unit of the X-ray flaw detection equipment. The hardware control unit controls the corresponding mechanism to move to the specified coordinates to complete the image acquisition.

[0063] The technical solutions provided by the embodiments of the present invention are introduced in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the embodiments of the present invention. The description of the above embodiments is only applicable to help understand the principles of the embodiments of the present invention. At the same time, for those skilled in the art, according to the embodiments of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A method for automatically generating detection points for X-ray inspection of head welds, characterized by: The following steps are included: S1. Use drawing software to draw a proportionally reduced cross-sectional outline of the weld according to the actual curvature of the head; S2. Preprocessing the drawn contour image, including image filtering, image binarization, and obtaining the coordinate points of the inner contour and the outer contour; S3. Eliminate the inner contour coordinate points and take the coordinates of two points with the same y coordinate, where the coordinate P should meet the following conditions: Then the inner contour coordinate P1(x1, y) should be discarded and the outer contour coordinate P2(x2, y) should be retained. S4. Based on the known conditions of the distance d between the detector and the cross-section of the weld inspection point, the focal length f between the detector and the light tube, and the width w of the detector imaging area, calculate the detectable weld length r at each inspection point. r satisfies the following formula: r=(fd)*w / f S5, take each detection point corresponding to the detectable contour coordinate starting point s (x i*r ,y i*r ), end point e(x i*r+r ,y i*r+r ), where i represents the number of welds currently being tested, and r is the length of the weld. According to the slope formula k=-(x i*r+r -x i*r ) / (and i*r+r -and i*r ) S6. Then the rotation angle α of the detector and the horizontal coordinate x and vertical coordinate y of the detector can be obtained. α=arctan(k) x=d*cos(α) y=d*sin(α) S7, according to the known conditions, the actual radius of the head weld is R, the radius of the contour map drawn in step S1 is R1, and according to the proportional conversion formula, the real coordinates (x m ,y m ), x m =R / R1*x y m =R / R1*y 2. The method for automatically generating detection points for X-ray detection of head welds according to claim 1, characterized in that: The cross-sectional profile of the head weld in step S1 is drawn using CAD software.

Images