JP2009276250A - Device and method for inspecting quality of weld bead - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately measure an appearance profile of a weld bead, improve reliability of inspection and inspection accuracy of quality of the weld bead. <P>SOLUTION: A device for inspecting a weld bead includes a color-imaging section 3 for color-imaging the appearance profile of the weld bead 22, by receiving irradiated beams irradiated towards the surface of the weld beam 22; an image formation section 40a for forming an inspection image of each color component by decomposing color imaging data to a plurality of color components; a mask region formation section 40b for forming a mask region, by extracting a predetermined image region from an inspection image formed by one-color component of the inspection image, an image processing section 40c for conducting, by masking a position corresponding to the inspection image formed by other color component, based on the mask region and extracting an analyzed image in an non-mask region, predetermined image processing to the analyzed image, and an inspection section 41 for inspecting the quality of the weld bead, based on the analyzed image that has conducted image processing. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  More particularly, the present invention relates to a weld bead quality inspection apparatus and method for inspecting weld defects of a weld bead based on the appearance shape of the weld bead.

  2. Description of the Related Art In recent years, in a manufacturing process of an automobile, a joined structure in which a pair of base materials such as a thin steel plate is butt welded (arc welding) is often used. Such a bonded structure has a quality in the width direction such as undercuts, blowholes, burn-outs, perforations, and misalignment and constriction on the surface of the welded portion (weld bead) of thin steel plate during welding. Various welding defects such as defects may occur. If such a welding defect still occurs, there is a problem that a crack is generated in the welded part due to a load at the time of molding in press molding, or the joint strength of the welded part is significantly reduced.

  Thus, many techniques related to a quality inspection device for a weld bead for inspecting a weld defect of the weld bead have been proposed. The weld defect of the weld bead was determined by visually observing the appearance shape (surface shape, two-dimensional shape, etc.) of the weld bead. Welding defects can now be determined automatically.

  As a conventional weld bead quality inspection device, for example, the appearance shape of a weld bead is measured from a light cut image obtained by receiving a slit-shaped laser beam irradiated toward the surface of the weld bead, and such appearance A configuration for inspecting a weld defect of a weld bead based on a shape is known. In such a quality inspection apparatus, specifically, laser light reflected on the surface of the weld bead is imaged by an imaging means such as a CCD camera, an analysis image is extracted based on the imaged image data, and the extracted analysis image is further extracted. When the predetermined image processing is performed, the welding defect of the weld bead is inspected based on the analysis image subjected to the image processing.

As another weld bead quality inspection device, Patent Document 1 discloses a laser welded portion (weld bead) imaged by an infrared camera, and the difference in luminance in the obtained infrared image is determined as an infrared emissivity on the surface of the laser welded portion. A configuration is disclosed in which a welded region is detected as an index representing the difference between the two, and the quality of welding of the laser welded portion is determined based on the shape of the welded region, the distribution pattern of luminance values, and the like.
Further, Patent Document 2 recognizes a three-dimensional welded part outline from first image data in which a regular light pattern such as a stripe pattern is projected on a welded part (welded bead), and the three-dimensional welded part outline is obtained. For the unrecognizable region, the quality of the welded part is judged based on the completed three-dimensional welded part outline by complementing it with the two-dimensional welded outline from the second image data on which the light pattern is not projected. A configuration is disclosed.
JP 2003-65985 A JP 2006-275536 A

  However, in the configuration of the above-described conventional weld bead quality inspection device, the irradiated light or reflected light is diffused / scattered depending on the shape and configuration of the welded portion (where the weld bead is formed), and the appearance shape of the weld bead. Could not be measured accurately. In other words, the laser light emitted toward the surface of the weld bead is diffused and scattered by the uneven shape of the weld, so adjacent laser lights interfere with each other, and the appearance shape of the weld bead is determined in the analysis image. Could not be accurately identified. And if the accuracy of specifying the appearance shape of the weld bead is poor in this way, there is a problem that the inspection accuracy of the weld defect is reduced, such as determination of the presence or absence of a hole defect in the weld bead and specification of the type of hole defect. It was.

  Moreover, the quality inspection apparatus of the weld bead disclosed in Patent Document 1 and Patent Document 2 described above has an appearance shape of the weld bead based on an infrared image of the weld bead on which a regular light pattern such as a stripe pattern is projected. Since it is configured to measure, similarly to the above-described conventional quality inspection apparatus, it is inferior in the accuracy of specifying the analysis image due to the influence of the diffusion and scattering of the reflected light. Therefore, the measurement accuracy of the appearance shape of the weld bead is poor, and the inspection accuracy of the weld defect of the weld bead is poor.

  Therefore, the present invention relates to a weld bead quality inspection apparatus and method for solving the above-described conventional problems, and accurately identifies the appearance shape of the weld bead to increase the reliability of the inspection, and thus the weld bead. It is an object of the present invention to provide a weld bead quality inspection apparatus and method for improving the quality accuracy of the inspection.

  The problems to be solved by the present invention are as described above. Next, means for solving the problems will be described.

  That is, in claim 1, in the weld bead quality inspection device that inspects a weld defect of the weld bead based on the appearance shape of the weld bead, the reflected light of the irradiation light irradiated toward the surface of the weld bead is received. Color imaging means for color imaging of the appearance shape of the weld bead, image generation means for decomposing the color imaging data acquired by the color imaging means into a plurality of color components, and generating an inspection image of each color component, and Of the inspection images generated by the image generation means, a mask area generation means for extracting a predetermined image area from the inspection image generated with one color component and generating a mask area; and the mask area generation means Based on the mask area, the analysis image in the non-mask area is extracted by masking the corresponding position of the inspection image generated with other color components, and the analysis image Those having image processing means for performing predetermined image processing, and quality inspection means for inspecting the quality of the weld bead based on the analysis image which the image processing has been performed by the image processing means is.

  According to a second aspect of the present invention, the color image pickup means reflects reflected light of the infrared slit light irradiated along the length direction of the weld bead and the range light irradiated toward a predetermined measurement range of the weld bead. Receiving light respectively, color imaging of the appearance shape of the weld bead, the image generating means generates an inspection image of each color component of RGB, and the mask area generating means is the B component generated by the image generating means A mask region is generated by extracting an image region including a width direction region of the weld bead from the inspection image.

  The image processing means masks the corresponding position of the R component inspection image generated by the image generation means based on the mask area generated by the mask area generation means, and the quality The inspection means determines the presence or absence of a hole defect on the surface of the weld bead from a plurality of stripe lines appearing in the analysis image subjected to the image processing by the image processing means.

  According to a fourth aspect of the present invention, the quality inspection unit includes an inner region formed on the inner side by a predetermined amount with respect to the width direction of the weld bead and an analysis region that has been subjected to image processing by the image processing unit. And the position and type of the hole defect are specified.

  According to a fifth aspect of the present invention, the quality inspection means calculates a difference between the image area of the analysis image image-processed by the image processing means and a preset reference image area, and thereby determines the quality in the width direction of the weld bead. It is for judging a defect.

  According to a sixth aspect of the present invention, the color imaging means receives the reflected light of the range light irradiated toward the predetermined measurement range of the weld bead, performs color imaging of the appearance shape of the weld bead, and the image generation means , RGB inspection images are generated, and the mask area generation means extracts an image area of a predetermined area including the shading portion of the weld bead surface from the R component inspection image generated by the image generation means. To generate a mask region.

  The image processing means masks the corresponding position of the B component inspection image generated by the image generation means based on the mask area generated by the mask area generation means, and the quality The inspection means calculates the area of the shaded portion appearing in the analysis image subjected to the image processing by the image processing means, and determines the quality defect of the hole defect formed on the surface of the weld bead.

  According to claim 8, in the weld bead quality inspection method for inspecting weld defects of the weld bead based on the appearance shape of the weld bead, the reflected light of the irradiation light irradiated toward the surface of the weld bead is received, and welding is performed. A color imaging process for color imaging the appearance shape of the bead, an image generation process for decomposing the color imaging data acquired by the color imaging process into a plurality of color components and generating an inspection image of each color component, and the image generation A mask region generating step for generating a mask region by extracting a predetermined image region from an inspection image generated with one color component among the inspection images generated by the steps, and the mask generated by the mask region generating step Based on the region, the analysis image in the non-mask region is extracted by masking the corresponding position of the inspection image generated with other color components, and a predetermined value is applied to the analysis image. An image processing step of performing image processing, and has a quality inspection process for inspecting the quality of the weld bead based on the analysis image which the image processing has been performed by the image processing step.

  According to a ninth aspect of the present invention, the color imaging step includes the reflected light of the infrared slit light irradiated along the length direction of the weld bead and the range light irradiated toward the predetermined measurement range of the weld bead. Receiving light respectively, color imaging of the appearance shape of the weld bead, the image generation step generates an inspection image of each color component of RGB, and the mask region generation step includes the B component generated by the image generation step A mask region is generated by extracting an image region including a width direction region of the weld bead from the inspection image.

  The image processing step masks the corresponding position of the R component inspection image generated by the image generation step based on the mask region generated by the mask region generation unit, and the quality In the inspection step, the presence or absence of a hole defect on the surface of the weld bead is determined from a plurality of striped lines appearing in the analysis image subjected to the image processing in the image processing step.

  In the eleventh aspect, the color imaging step receives reflected light of the range light irradiated toward a predetermined measurement range of the weld bead to perform color imaging of the appearance shape of the weld bead, and the image generation step includes , RGB inspection images are generated, and the mask region generation step extracts an image region of a predetermined region including a shading portion of the weld bead surface from the R component inspection image generated by the image generation step. To generate a mask region.

  The image processing step masks the corresponding position of the B component inspection image generated by the image generation step based on the mask region generated by the mask region generation step, and the quality In the inspection step, the area of the shaded portion appearing in the analysis image that has been subjected to the image processing in the image processing step is calculated, and the quality defect of the hole defect formed on the surface of the weld bead is determined.

  By configuring as in the present invention, the analysis image is extracted based on the mask region, so that the inspection range is narrowed to improve the accuracy of the image processing, and the inspection image of a predetermined color component is used, so that the quality inspection is performed. According to the purpose, it is possible to obtain a highly accurate analysis image in which the influence of diffusion and scattering of laser light is reduced. Therefore, it is possible to accurately measure the appearance shape of the weld bead to increase the reliability of the inspection, and to improve the inspection accuracy of the quality of the weld bead.

Next, the best mode for carrying out the invention will be described.
1 is a diagram showing an overall configuration of a quality inspection apparatus according to a first embodiment of the present invention, FIG. 2 is a functional block diagram showing a configuration of a processing inspection unit, and FIG. 3 is generated by an image generation unit FIG. 4 is a diagram showing an inspection image of R component and B component, FIG. 4 is a diagram showing a mask region created by the mask region generation unit, FIG. 5 is a diagram showing an analysis image extracted by the image processing unit, and FIG. Is a diagram showing an actual measurement image region and a reference image region extracted by the inspection unit, FIG. 7 is a flowchart showing the quality inspection method of the present embodiment, and FIG. 8 is an overall view of the quality inspection apparatus according to the second embodiment. FIG. 9 is an enlarged plan view showing the welded portion of the joint structure in FIG. 8, FIG. 10 is a functional block diagram showing the configuration of the processing inspection unit, and FIG. 11 is generated by the image generating unit. FIG. 12 shows a mask region created by the mask region generation unit. The figure 13 is a flow chart showing the inspection method of the second embodiment.
In the following embodiments, the width direction of the weld bead 22 is the X direction, and the length direction of the weld bead 22 is the Y direction.

  As shown in FIG. 1, the quality inspection apparatus 1 of the present embodiment is applied in an inspection process of a joint structure 2 used for an automobile body in an automobile manufacturing process. In the joint structure 2, a pair of thin steel plates 20, 20 are joined as a base material by butt welding (laser welding). Specifically, the pair of thin steel plates 20, 20 are substantially T-shaped at the weld portion 21. And are welded together. A weld bead 22 is formed on the welded portion 21 by irradiation with laser light from a laser welding apparatus (not shown).

  The quality inspection apparatus 1 according to the present embodiment is an apparatus that inspects a welding defect of the weld bead 22 based on the appearance shape of the weld bead 22, and specifically, irradiation light irradiated toward the surface of the weld bead 22. The color imaging unit 3 as a color imaging unit that receives the reflected light and color-images the appearance of the weld bead 22, and generates an inspection image from the color imaging data acquired by the color imaging unit 3. A processing inspection unit 4 that performs predetermined image processing on the analysis image extracted from the image and inspects the quality of the weld bead based on the analysis image subjected to the image processing.

First, the color imaging unit 3 will be described in detail below.
As shown in FIG. 1, the color imaging unit 3 includes a light source unit 30 that irradiates irradiation light toward the surface of the welded part 21 (weld bead 22), and welding light among the irradiation light irradiated by the light source unit 30. The camera unit 31 is configured to receive the reflected light reflected by the unit 21 (weld bead 22) and to color-image the appearance shape (two-dimensional shape) of the weld bead 22. The light source unit 30 irradiates the first light source unit 32 that irradiates infrared slit light along the length direction (Y direction) of the weld bead 22, and the first light that irradiates range light toward a predetermined measurement range of the weld bead 22. The two light source sections 33 and 33 are configured.

  The first light source unit 32 is configured to be able to irradiate infrared laser light as linear slit light, and specifically, infrared laser light orthogonal to the length direction (Y direction) of the weld bead 22. Are irradiated so as to form slits at substantially equal intervals along the length direction of the weld bead 22. Moreover, the 2nd light source part 33 * 33 is comprised so that LED light can be irradiated toward the surface of the welding part 21 (weld bead 22), and, specifically, the width direction (X direction) of the weld bead 22 And LED light is irradiated toward the measurement range containing the predetermined | prescribed surface over a length direction (Y direction). The measurement range is also irradiated with the laser light from the first light source unit 32 described above.

  As the camera unit 31, a 3CCD camera having spectral sensitivity characteristics that can be imaged for each of R (red), G (green), and B (blue) color components is used, and the light source unit 30 ( Of the irradiation light irradiated by the first light source unit 32 and the second light source unit 33), the external shape of the weld bead 22 is received by receiving the reflected light reflected by the surface of the welded portion 21 (weld bead 22). (Two-dimensional shape) is color-imaged. And the camera part 31 is connected to the process test | inspection part 4 mentioned later, and the external appearance shape of the weld bead 22 imaged by the camera part 31 is sent to the process test | inspection part 4 as color imaging data.

  Thus, in the color imaging unit 3 of the present embodiment, the first light source unit 32 irradiates the slit light (laser light) along the length direction (Y direction) of the weld bead 22 and the second light source unit. The range light is irradiated toward the predetermined measurement range of the weld bead 22 by 33 and 33, and the irradiation light is reflected on the surface of the weld bead 22 and then received by the camera unit 31. 22 appearance shapes are color-imaged.

Next, the processing inspection unit 4 will be described in detail below.
As shown in FIGS. 2 to 6, the processing inspection unit 4 includes a CPU (not shown) that executes various processes, a memory (not shown) as a storage unit that stores various processing programs, and the like. An inspection image is generated from the color imaging data acquired by the color imaging unit 3, and a predetermined image processing is performed on the analysis image extracted from the inspection image, and image processing is performed by the processing unit 40. And an inspection unit 41 for inspecting the quality of the weld bead 22 based on the analyzed image. In particular, in the quality inspection apparatus 1 according to the present embodiment, the presence or absence of the hole defect 45 of the weld bead 22 is determined based on the plurality of stripe lines 44, 44... Based on the infrared slit light that appears on the surface of the weld bead 22. It is comprised so that it may determine (refer FIG. 3 etc.).

  In addition to the above-described configuration, the processing inspection unit 4 is connected to an output device 42 such as a monitor capable of outputting an image, an input device 43 such as a keyboard, a numeric keypad, and a touch panel. In particular, the output device 42 includes various images (such as inspection images, mask regions, and analysis images) extracted by the processing inspection unit 4 as described later, and the quality of the weld bead 22 determined by the inspection unit 41. The result of inspection is displayed.

  The processing unit 40 decomposes the color imaging data acquired by the camera unit 31 into a plurality of color components, and generates an image generation unit 40a as an image generation unit that generates an inspection image of each color component, and the image generation unit 40a generates A mask area generation unit 40b serving as a mask area generation unit that generates a mask area by extracting a predetermined image area from an inspection image generated with one color component, and a mask area generation unit 40b. Based on the generated mask area, the corresponding position of the inspection image generated with other color components is masked to extract the analysis image in the non-mask area, and the analysis image is subjected to predetermined image processing The image processing unit 40c and the like as processing means are configured.

  In the image generation unit 40a, the color imaging data acquired by the camera unit 31 of the color imaging unit 3 described above is received, the color imaging data is decomposed into RGB color components, and an inspection image of each RGB color component is obtained. Generated. In the image generation unit 40a of this embodiment, an R component inspection image, a G component inspection image, and a B component inspection image are generated. In addition, since the inspection image of each component is an image corresponding to the same predetermined range in the weld bead 22, for example, a predetermined position of the inspection image of one component and a corresponding position of the inspection image of the other component Are configured to match.

  Specifically, as shown in FIG. 3, the inspection image of the R component in which the stripe lines 44, 44... Based on the infrared slit light are clarified, and the influence of the infrared slit light is eliminated and the weld bead is removed. Each of the B component inspection images in which the contour shape of 22 is clarified is generated. Stripe lines 44, 44... Based on the infrared slit light clearly appear in the R component inspection image (see FIG. 3A), while the B component inspection image includes the infrared slit light. Although the striped lines 44, 44... Are not displayed, the edges 22a and 22a clearly appear as the contour in the width direction (X direction) of the weld bead 22 (see FIG. 3B).

  In the mask region generation unit 40b, first, an image region including the width direction region of the weld bead 22 is extracted from the B component inspection image generated by the image generation unit 40a (see FIG. 4A). The “image region including the width direction region of the weld bead 22” refers to a region including at least the region inside the edges 22a and 22a in the width direction (X direction) of the weld bead 22, and in this embodiment. In the width direction (X direction) of the weld bead 22, an image region is extracted by adding a region inside the edge portions 22a and 22a and a region formed larger than the region by a predetermined amount outside (see FIG. 3 (b)). The “predetermined amount” is set so that the undercut 23 is included in the above-described image region on the assumption that the undercut 23 is generated at the edges 22a and 22a of the weld bead 22 during welding.

  Next, the mask area generation unit 40b generates a mask area from the image area extracted as described above (see FIG. 4B). The “mask area” corresponds to an area obtained by inverting the extracted image area in the B component inspection image. In other words, the mask area in this embodiment is an area other than an image area including the width direction area of the weld bead 22 in the B component inspection image (hereinafter, this area is referred to as a “non-mask area”).

  In the image processing unit 40c, first, based on the mask region generated by the mask region generation unit 40b, the corresponding position of the R component inspection image generated by the image generation unit 40a is masked, and the above-described non-mask region The analysis image is extracted (see FIG. 5A). The analysis image extracted in this way corresponds to an image appearing in an image region including the width direction region of the weld bead 22 in the R component inspection image.

  Then, in the image processing unit 40c, predetermined image processing is performed on the extracted analysis image so that a striped line image based on the infrared slit light that appears in the analysis image becomes clearer. In the analysis image thus image-processed, the plurality of stripe lines 44, 44... Based on the infrared slit light are spaced apart at substantially equal intervals along the length direction of the weld bead 22. Appears (see FIG. 5B).

  In the inspection unit 41, the quality inspection of the weld bead 22 is performed based on the analysis image image-processed by the image processing unit 40c described above. In the inspection unit 41 of this embodiment, the presence / absence of the hole defect 45 on the surface of the weld bead 22 and the type of the hole defect 45 are determined from the plurality of stripe lines 44. The determination is referred to as “hole defect determination”), and the difference between the image area of the analysis image and a preset reference image area is calculated to determine the quality defect in the width direction of the weld bead 22 (hereinafter, referred to as “defect determination”). This determination is referred to as “width direction defect determination”).

  In the hole defect determination, the presence / absence of the defective portion 44a is detected by the striped lines 44, 44... Appearing in the analysis image, and the presence / absence of the hole defect 45 on the surface of the weld bead 22 is determined based on the presence / absence of the defective portion 44a. The In the present embodiment, when the defective portion 44 a is detected by the stripe line 44, it is determined that the hole defect 45 is generated on the surface of the weld bead 22, and the defective portion 44 a is not detected by the stripe line 44. It is determined that a hole defect 45 has occurred on the surface of the weld bead 22 (see FIG. 5B).

  In the analysis image subjected to the image processing, the normal striped line 44 appears as a single continuous line along the width direction of the weld bead 22, but the continuous line is cut in the middle, and a defective part 44 a is expressed. There is something. The cause of the appearance of the defect portion 44a in the stripe line 44 is not only blow holes, perforations and melted holes formed on the surface of the weld bead 22, but also welding such as an undercut 23 generated at the edge 22a of the weld bead 22. This is because the hole defect 45 is generated on the surface of the bead 22. That is, when a hole defect 45 occurs on the surface of the weld bead 22, the infrared laser beam irradiated toward the surface of the weld bead 22 is irregularly reflected by the hole defect 45, so that the luminance level of the portion in the analysis image is obtained. As a result, the striped line 44 is lost and a defective part 44a appears.

  When it is determined that the hole defect 45 has occurred, the position where the defect portion 44 a is detected is further identified by the stripe line 44, and the position and hole defect where the hole defect 45 has occurred on the surface of the weld bead 22. 45 types are identified. In particular, in the inspection unit 41 of the present embodiment, the analysis image is divided into an inner region formed inside a predetermined amount with respect to the width direction of the weld bead 22 and an outer region other than the inner region. The position and type of the hole defect 45 are specified according to the region.

  Specifically, first, the analysis image is formed larger in the width direction (X direction) of the weld bead 22 than the inner region (inner region) of the edges 22a and 22a and a predetermined amount outside the inner region. It is divided into an outer region (outer region) (see FIG. 5B). As an example of area division, a boundary line 46 is drawn inward by a predetermined amount from the side in the width direction of the analysis image, an area inside the boundary line 46 is an inner area, and an area outside the boundary line 46 is an outer area. As divided. The “predetermined amount” is a reference value set in advance so that the outer region includes the undercut 23.

  When the position of the defective portion 44a of the stripe line 44 is specified and the defective portion 44a is detected in the inner region, it is a hole defect 45 generated on the surface side of the weld bead 22, which is a blowhole or hole. , It is identified as a hole defect 45 caused by melting. On the other hand, when the defect part 44a is detected in the outer region, the hole defect 45 generated at the edge 22a of the weld bead 22 and identified as the hole defect 45 caused by the undercut 23 is specified.

  In the width direction defect determination, an actual measurement image region is extracted from an actual analysis image of the weld bead 22 to be inspected, and a difference between the actual measurement image region and a preset reference image region is calculated. A quality defect in the width direction of the weld bead is determined. The “image area” here is an image area including the width direction area of the weld bead 22 and corresponds to the above-described non-mask area. The “preliminarily set reference image region” is a reference image region that is extracted from an analysis image of the weld bead 22 that has no quality defect in the width direction in advance.

  The difference between the measured image area and the preset reference image area is compared with the position in the width direction of the measured image area and the reference image area, and the relative error is calculated. Specifically, the actual image area and the reference image area are matched (superposed), and as a result, the actual image area is shifted in the width direction (X direction) and does not overlap with the reference image area. It is determined that the weld bead 22 is defective in quality in the width direction such as misalignment and constriction. In particular, in this embodiment, when the measured image areas are shifted in the width direction over the entire length direction of the weld bead 22 and do not overlap (see FIG. 6A), the quality defect due to misalignment. If the measured image area is shifted in the width direction and does not overlap in a part of the length direction of the weld bead 22 (see FIG. 6B), it is a quality defect due to the constriction. Determined.

Next, the quality inspection method of the weld bead 22 using the quality inspection apparatus 1 of the present embodiment will be described below.
As shown in FIG. 7, the quality inspection method of the weld bead 22 using the quality inspection device 1 of the present embodiment receives the irradiation light irradiated toward the surface of the weld bead 22, and the appearance shape of the weld bead 22. A color imaging step (S100) for color imaging, an image generation step (S110) for decomposing the color imaging data acquired in the color imaging step into a plurality of color components and generating an inspection image for each color component, and image generation Of the inspection images generated in the process, a mask region generation step (S120) for generating a mask region by extracting a predetermined image region from the inspection image generated with one color component, and a mask region generation step. Based on the mask area, the corresponding position of the inspection image generated with other color components is masked to extract an analysis image in the non-mask area, and a predetermined image processing is performed on the analysis image. An image processing step of performing (S130), those having a quality inspection step (S140) or the like to check the quality of the weld bead based on the analysis image in the image processing by the image processing step is performed.

  First, in the color imaging step (S100), the surface of the weld bead 22 formed on the welded portion 21 is irradiated to the joint structure 2 in which the pair of thin steel plates 20 and 20 are butt welded in a substantially T shape. Light is irradiated, and the appearance shape of the weld bead 22 is color-imaged. Specifically, the length of the weld bead 22 from the light source unit 30 (the first light source unit 32 and the second light source units 33 and 33) toward the surface of the weld bead 22 by the color imaging unit 3 constituting the quality inspection apparatus 1. The infrared slit light irradiated along the vertical direction and the range light irradiated toward the predetermined measurement range of the weld bead 22 are irradiated (S101) and reflected by the surface of the weld bead 22 by the camera unit 31. By receiving the reflected light, the appearance shape (two-dimensional shape) of the weld bead 22 is color-imaged (S102).

  In the image generation step (S110), based on the color imaging data acquired by the camera unit 31 of the color imaging unit 3, the color imaging data is decomposed into RGB color components (S111), and an image of each RGB color component is obtained. By performing color binarization processing on the data, an R component inspection image and a B component inspection image are generated (S112) (see FIG. 3).

  In the mask region generation step (S120), the regions inside the edges 22a and 22a of the weld bead 22 are specified by the painting process based on the difference in luminance level appearing in the B component inspection image, and then the edge 22a. -By adding a region that is formed larger outside by a predetermined amount to the inner region of 22a, an image region including the width direction region of the weld bead 22 is extracted (S121). Then, the extracted image area including the width direction area of the weld bead 22 is inverted to create a target mask area (S122) (see FIG. 4).

  In the image processing step (S130), the image corresponding to the mask area is deleted from the R component inspection image, and only the image corresponding to the non-mask area is extracted as an analysis image (S131) (FIG. 5A). reference). Next, image processing is performed on the extracted analysis image (S132). Specifically, various processing steps such as thinning processing, area removal processing, and noise check processing are performed (FIG. 5B). )reference).

  In the quality inspection process (S140), hole defect determination and width direction defect determination of the weld bead 22 are performed based on the analysis image image-processed in the image processing process.

  In the hole defect determination, first, the analysis image is divided into an inner region with respect to the boundary line 46 in the width direction (X direction) of the weld bead 22 and an outer region formed larger by a predetermined amount than the inner region. (S141). Then, based on each area of the analysis image, the position of the missing portion 44a of the stripe line 44 is specified (S142), and the type of the hole defect 45 is determined (S143). When the defect part 44a is detected in the inner region, the hole defect 45 generated on the surface side of the weld bead 22 is identified as the hole defect 45 caused by blowhole, perforation, or burn-through. . On the other hand, when the defect part 44a is detected in the outer region, it is specified as the hole defect 45 generated in the edge 22a of the weld bead 22 and caused by the undercut 23 (FIG. 5 (b)).

  In the width direction defect determination, a measured image region is extracted from a measured analysis image of the weld bead 22 to be inspected (S144), and the measured image region and the reference image region are compared with each other in the width direction. And the difference between the reference image area and the reference image area are calculated (S145). As a result, when the actually measured image area is shifted in the width direction (X direction) and does not overlap with the reference image area, the weld bead 22 to be inspected has a quality defect in the width direction such as misalignment or constriction. If it is determined that the weld bead 22 is in the width direction, the quality defect in the width direction is determined (S146) (see FIG. 6).

  As described above, in the quality inspection apparatus 1 according to the present embodiment, in the quality inspection apparatus 1 for the weld bead 22 that inspects the welding defect of the weld bead 22 based on the appearance shape of the weld bead 22, the surface is directed toward the surface of the weld bead 22. The color imaging unit 3 that receives the irradiated irradiation light and performs color imaging of the appearance shape of the weld bead 22, and the color imaging data acquired from the color imaging unit 3 are separated into a plurality of color components, and each color component An image generation unit 40a that generates an inspection image, and a mask region that generates a mask region by extracting a predetermined image region from an inspection image generated with one color component among the inspection images generated by the image generation unit 40a Based on the mask region generated by the generation unit 40b and the mask region generation unit 40b, the corresponding position of the inspection image generated with other color components is masked to be in the non-mask region. An image processing unit 40c that extracts an analysis image and performs predetermined image processing on the analysis image, and an inspection unit 41 that inspects the quality of the weld bead based on the analysis image subjected to image processing by the image processing unit 40c. Therefore, it is possible to accurately measure the appearance shape of the weld bead 22 to increase the reliability of the inspection, and to improve the inspection accuracy of the quality of the weld bead 22.

That is, the quality inspection apparatus 1 according to the present embodiment selects an inspection image based on one color component based on an inspection image decomposed into predetermined color components, generates a mask region, and sets other color components as the other color components. An analysis image is extracted from the non-mask region of the based inspection image. As described above, by configuring the analysis image to be extracted based on the mask region, the inspection range is narrowed down, the accuracy of the image processing is improved, and the calculation speed of the image processing can be increased.
Further, in the present embodiment, since it is configured to separate the original color component inspection image for creating the mask region and the original color component inspection image for extracting the analysis image, It is possible to select an inspection image suitable for specifying the appearance shape of the weld bead 22 and perform image processing, and to obtain a highly accurate analysis image in which the influence of diffusion and scattering of laser light is reduced according to the purpose of quality inspection. Obtainable.
Therefore, the appearance shape of the weld bead 22 can be specified with high accuracy based on the analysis image, and the reliability of the inspection can be improved. And the inspection precision of the quality quality of the weld bead 22 can be improved from the external appearance shape of the weld bead 22 specified in this way.

  In particular, in the quality inspection apparatus 1 according to the present embodiment, the color imaging unit 3 has a predetermined measurement range of the infrared slit light irradiated along the length direction (Y direction) of the weld bead 22 and the weld bead 22. The range light irradiated toward each is received, and the appearance shape of the weld bead 22 is color-imaged. The image generation unit 40a generates an inspection image of each color component of RGB, and the mask region generation unit 40b Since the image area including the width direction area of the weld bead is extracted from the inspection image of the B component generated by the generation unit 40a and the mask area is generated, the diffusion / scattering of the infrared laser beam is suppressed. By using the inspection image of the B component with reduced influence, the region in the width direction of the weld bead 22 can be clearly specified, and the mask region can be generated with high accuracy.

  In the present embodiment, the image processing unit 40c masks the corresponding position of the R component inspection image generated by the image generation unit 40a based on the mask region generated by the mask region generation unit 40b, and performs the inspection. Since the part 41 determines the presence or absence of the hole defect 45 on the surface of the weld bead 22 from the plurality of stripe lines 44, 44... Appearing in the analysis image subjected to the image processing by the image processing unit 40c, By extracting the analysis image using the R component inspection image in which the infrared laser beam clearly appears, the hole defect 45 on the surface of the weld bead 22 is accurately identified from the striped lines 44, 44. It is possible to improve the determination accuracy of the hole defect 45.

  Further, in the present embodiment, the inspection unit 41 includes an inner region formed on the inner side by a predetermined amount with respect to the width direction of the weld bead 22, and the inner region of the analysis image that has been subjected to image processing by the image processing unit 40c. In order to identify the position and type of the hole defect 45, the hole defect 45 is formed on the surface of the weld bead 22, the hole defect 45 is based on the perforation or melted down, and the undercut 23. Since the hole defect 45 based on it can be specified separately, the inspection accuracy of the quality of the weld bead 22 can be further improved.

  Further, in the present embodiment, the inspection unit 41 calculates the difference between the image region of the analysis image subjected to the image processing by the image processing unit 40c and a preset reference image region, so that the welding bead 22 in the width direction is calculated. Since the quality defect is determined, the difference between the image regions can be easily calculated, and the inspection speed for determining the quality unnecessary in the width direction can be improved.

  In addition, about the quality inspection apparatus 1 and its method of a present Example, it is not limited to the Example mentioned above, A various change is possible unless it deviates from the objective of this invention.

  That is, in the configuration of the quality inspection apparatus 1 of the above-described embodiment, the surface of the weld bead 22 is irradiated with infrared slit light, and the weld bead 22 is obtained from the plurality of stripe lines 44. The presence / absence of a hole defect 45 on the surface is determined, but the inspection content of the weld defect of the weld bead 22 inspected by the quality inspection apparatus 1 is not limited to this, and is formed on the surface of the weld bead 22, for example. It may be configured to determine the quality defect of the hole defect 145 by specifying the shaded portion 144 and calculating the area value of the shaded portion 144.

  In the quality inspection apparatus 101 of the second embodiment shown in FIGS. 8 to 12, the processing inspection unit 104 generates an inspection image from the color imaging data acquired by the color imaging unit 103 and analyzes extracted from the inspection image. A processing unit 140 that performs predetermined image processing on the image, an inspection unit 141 that inspects the quality of the weld bead 122 based on the analysis image that has been subjected to image processing by the processing unit 140, and the like. In particular, the quality inspection apparatus 101 according to the present embodiment is configured such that the quality defect of the hole defect 145 of the weld bead 122 is determined based on the plurality of shade portions 144, 144... Appearing on the surface of the weld bead 122. Has been.

  In the configuration of the quality inspection apparatus 101 of the present embodiment, unless otherwise specified, the members having the same names as those of the quality inspection apparatus 1 of the above-described embodiment are configured in the same manner, and the detailed description thereof Is omitted.

  In the joint structure 102 of this embodiment, a pair of thin steel plates 120 and 120 are welded as base materials in a state where both ends are overlapped at a welded portion 121. A weld bead 122 is formed on the welded portion 121 by irradiation with laser light from a laser welding apparatus (not shown). In addition, on the surface of the weld bead 122, a hole defect 145 formed in a concave shape and a slug 146 raised in a convex shape are formed (see FIG. 9 and the like). The hole defect 145 and the slag 146 appear as a shaded portion 144 in an inspection image or the like to be described later (see FIG. 11 and the like).

  In the quality inspection apparatus 101 of the present embodiment, the color imaging unit 103 includes a light source unit 130 that irradiates irradiation light toward the surface of the welded part 121 (weld bead 122), and among the irradiation light that is irradiated by the light source unit 130. The camera unit 131 is configured to receive the reflected light reflected by the welded part 121 (weld bead 122) and perform color imaging of the appearance shape (two-dimensional shape) of the weld bead 122. In particular, unlike the first embodiment described above (see FIG. 1), the light source unit 130 includes only second light source units 133 and 133 that emit range light toward a predetermined measurement range of the weld bead 122. Yes.

  In the image generation unit 140a, the color imaging data acquired by the camera unit 31 of the color imaging unit 3 described above is received, the color imaging data is decomposed into RGB color components, and an inspection image of each RGB color component is obtained. Is generated (see FIG. 11). In the inspection image of each color component generated by the image generation unit 140a of the present embodiment, a plurality of shade portions 144, 144... Appear on the surface of the weld bead 122, respectively. In the present embodiment, due to the difference in the brightness value of the light and shade portion 144, both the hole defect 145 and the light and shade portion 144 based on the slag 146 appear clearly in the R component inspection image, and the hole defect 145 appears in the B component inspection image. The outline of the shaded portion 144 based on the above appears clearly (not shown).

  The “dark portion 144” refers to a substantially circular pattern in plan view that appears on the surface of the weld bead 122 in the inspection image. The shading portion 144 includes a portion based on a slag 146 formed in a convex shape in addition to a portion based on a hole defect 145 formed in a concave shape on the surface of the weld bead 122. In the quality inspection apparatus 101 of the present embodiment, as will be described later, only the shading portion 144 based on the hole defect 145 is accurately identified through a predetermined processing process in the processing unit 140.

  In the mask region generation unit 140b, an image region of a predetermined region including the shading portion 144 on the surface of the weld bead 122 is extracted from the R component inspection image generated by the image generation unit 140a, and a mask region is generated from the image region. (See FIG. 12). The “image region of the predetermined region including the light and dark portion 144” refers to a region that includes at least the light and dark portion 144 and is larger than the light and dark portion 144 by a predetermined amount.

  Specifically, in the mask area generation unit 140b according to the present embodiment, first, the light and shade portions 144, 144,... Having luminance values equal to or higher than a predetermined threshold are specified for the R component inspection image. The threshold value is appropriately set according to a predetermined condition for specifying the shaded portions 144, 144... Based on the hole defect 145. Then, the center of gravity of each of the shaded areas 144, 144... Is calculated, and an image area of a predetermined area including the substantially circular shaded part 144 is extracted around the calculated center of gravity. From this, a mask area is generated.

  In this step, the shaded portion 144 based on the hole defect 145 is selected to some extent, but the shaded portion 144 based on the slag 146 may be mixed without being completely eliminated. In the present embodiment, the image processing unit 140c described later performs predetermined image processing on the analysis image, so that only the shading portion 144 based on the hole defect 145 is specified.

  The image processing unit 140c masks the corresponding position of the B component inspection image generated by the image generation unit 140a based on the mask region generated by the mask region generation unit 140b. An analysis image is extracted. Then, the image processing unit 140c performs predetermined image processing (binarization processing, contrast enhancement processing, area removal processing, etc.) on the extracted analysis image. Specifically, image processing for selecting only the shading portion 144 based on the hole defect 145 from the shading portion 144 appearing in the analysis image, and image processing for clarifying the peripheral portion of the selected shading portion 144 Each is done.

  In the inspection unit 141, the area of the shading portion 144 that appears in the analysis image that has been subjected to image processing by the image processing unit 140c is calculated, and the quality defect of the hole defect 145 formed on the surface of the weld bead 122 is determined. (Hereinafter, this determination is referred to as “hole defect determination”). Specifically, in the hole defect determination, the total value of the areas of the shading portions 144, 144,... Appearing in the analysis image is calculated, and it is determined whether or not the total value is larger than a preset reference value. The As a result, when the total value of the area description of the shading portion 144 is larger than a predetermined reference value, it is determined that the weld bead 122 has a defective quality of the hole defect 145.

  Further, in the processing inspection unit 104 of the present embodiment, the finally specified analysis image (the shading portion 144 thereof) is filled and displayed on the output device 142 in a state where it is superimposed on the inspection image of a predetermined component.

Next, the quality inspection method of the weld bead 122 using the quality inspection apparatus 101 of the present embodiment will be described below.
As shown in FIG. 13, in the color imaging step (S200), a weld bead formed on a welded portion 121 of a bonded structure 102 in which a pair of thin steel plates 120 and 120 are butt welded in a substantially T shape. Irradiation light is irradiated on the surface 122, and the appearance shape of the weld bead 22 is color-imaged. Specifically, the color imaging unit 103 constituting the quality inspection apparatus 101 is directed from the light source unit 130 (second light source unit 133/133) toward the surface of the weld bead 122 toward a predetermined measurement range of the weld bead 22. The irradiated range light is irradiated (S201), and the reflected light reflected by the surface of the weld bead 122 is received by the camera unit 131, whereby the appearance shape (two-dimensional shape) of the weld bead 122 is color-captured. (S202).

  In the image generation step (S210), based on the color imaging data acquired by the camera unit 131 of the color imaging unit 103, the color imaging data is decomposed into RGB color components (S211), and each RGB color component image is displayed. Color binarization processing is performed on the data (S212), and an R component inspection image and a B component inspection image are generated.

  In the mask region generating step (S220), after the binarization process and the area removal process are performed on the shading portions 144, 144... Appearing in the R component inspection image based on the luminance value equal to or higher than a predetermined threshold, The center of gravity of each of the shaded areas 144, 144... Is calculated, and an image area of a predetermined area including the substantially circular shaded part 144 is extracted around the calculated center of gravity (S221). Then, the image area of the predetermined area including the shading portion 144 is inverted, and a target mask area is created (S222).

  In the mask area generation step (S220) of the present embodiment, when the area of the non-mask area is filled and the filled areas overlap each other during the transition to the image processing process (S230) described later. In each of the shaded areas 144 in the overlapped area, the area removal process is performed. Then, a predetermined image area including the shading portion 144 is extracted again, and a mask area is generated.

  In the image processing step (S230), only an image corresponding to a non-mask area is extracted as an analysis image from the B component inspection image (S231), and then predetermined image processing is performed on the extracted analysis image. (S232). Specifically, various processing steps such as dark portion emphasis processing, binarization processing, and area removal processing are performed. In particular, in the image processing step of the present embodiment, preferably, binarization processing or the like is performed in a plurality of times. For example, by adjusting the binarization threshold level and performing a plurality of binarization processes and the like, the accuracy of the selection of the light / dark part 144 (the light / dark part 144 based on the hole defect 145 and the slag 146) is improved and the light / dark The resolution of the peripheral part of the part 144 can be improved.

  In the quality inspection process (S240), the hole defect determination of the weld bead 122 is performed based on the analysis image subjected to the image processing in the image processing process. That is, the total value of the areas of the shaded portions 144, 144,... Appearing in the analysis image is calculated (S241), and it is determined whether or not the total value is greater than a preset reference value (S242). As a result, when the total value of the area description of the shading portion 144 is larger than a predetermined reference value, it is determined that the weld bead 122 has a defective quality of the hole defect 145 (S243).

  As described above, in the quality inspection apparatus 101 according to the present embodiment, unlike the above-described embodiment (see FIG. 1 and the like), as the inspection content of the welding defect, from the light and dark portion 144 formed on the surface of the weld bead 22, Although it is configured to determine the quality defect of the defect 145, even with such a configuration, the external shape of the weld bead 22 is accurately measured as in the above-described embodiment (see FIG. 1 and the like). As a result, the reliability of the inspection can be improved, and as a result, the inspection accuracy of the quality of the weld bead 22 can be improved.

  In particular, the color imaging unit 103 receives the range light irradiated toward the predetermined measurement range of the weld bead 122, performs color imaging of the appearance shape of the weld bead 122, and the image generation unit 140a performs RGB components. The mask region generation unit 140b extracts an image region of a predetermined region including the shading portion 144 on the surface of the weld bead 122 from the R component inspection image generated by the image generation unit 140a to generate a mask. Since it is configured to generate a region, it is configured to use an inspection image of the R component in which the contrast of the light and shade part 144 clearly appears, so that the R component in which the influence of diffusion and scattering of range light is reduced By using this inspection image, it is possible to reliably specify the shaded portion 144 that appears on the surface of the weld bead 122, and to generate the mask region with high accuracy.

  Further, in the present embodiment, the image processing unit 140c masks the corresponding position of the B component inspection image generated by the image generation unit 140a based on the mask region generated by the mask region generation unit 140b. The unit 141 calculates the area of the shaded portion 144 that appears in the analysis image that has been subjected to image processing by the image processing unit 140c, and determines the quality defect of the hole defect 145 formed on the surface of the weld bead 122. Therefore, the analysis image is extracted using the B component inspection image in which the shading portion 144 based on the hole defect 145 clearly appears, so that the opening area of the shading portion 144 that appears in the analysis image is specified with high accuracy. It is possible to improve the accuracy of determining the quality defect of the hole defect 145.

  For example, in the above-described embodiment (see FIG. 1 and the like), for example, the processing inspection unit 4 in which the processing unit 40 and the inspection unit 41 are integrally formed is provided. The processing unit 40 and the inspection unit 41 may be configured as separate bodies.

  Further, the processing unit 40 in the above-described embodiment (see FIGS. 1 and 8, etc.) can extract an inspection image of a predetermined color component according to the inspection purpose and inspection conditions. Further, the image processing step is not limited to the above-described binarization processing, thin line processing, and the like, and a known processing step can be used as appropriate according to the purpose of the image processing.

  Furthermore, in the above-described embodiment (see FIG. 1 and the like), the inspection unit 41 is configured to perform both hole defect determination and width direction defect determination, but at least one of the determinations is performed. What is necessary is just to be comprised. In particular, regarding the width direction defect determination, when the actual measurement image region is shifted to any one of the width direction with respect to the reference image region and does not overlap, the area value of the non-overlap region of the actual measurement image region is a predetermined threshold value. In the case where it is larger than that, it may be determined that a quality defect in the width direction has occurred.

The figure which showed the whole structure of the quality inspection apparatus which concerns on 1st Example of this invention. The functional block diagram which showed the structure of the process test | inspection part. The figure which showed the test | inspection image of R component and B component which are produced | generated by an image generation part. The figure which showed the mask area | region produced with a mask area | region production | generation part. The figure which showed the analysis image extracted by an image process part. The figure which showed the measurement image area | region and reference | standard image area | region extracted by the test | inspection part. The flowchart which showed the quality inspection method of a present Example. The figure which showed the whole structure of the quality inspection apparatus which concerns on a 2nd Example. The top view which expanded and showed the welding part of the joint structure in FIG. The functional block diagram which showed the structure of the process test | inspection part. The figure which showed the test | inspection image produced | generated by an image generation part. The figure which showed the mask area | region produced with a mask area | region production | generation part. The flowchart which showed the quality inspection method of 2nd Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Quality inspection apparatus 2 Connection structure 3 Color imaging part (color imaging means)
4 processing inspection part 22 welding bead 23 undercut 30 light source part 31 camera part 32 first light source part 33 second light source part 40 processing part 40a image generation part (image generation means)
40b Mask area generator (mask area generator)
40c Image processing unit (image processing means)
41 Inspection Department (quality inspection means)
44 Striped line 44a Defect site 45 Hole defect

Claims (12)

In a weld bead quality inspection device that inspects weld defects of the weld bead based on the appearance shape of the weld bead,
A color imaging unit that receives reflected light of the irradiation light irradiated toward the surface of the weld bead and performs color imaging of the appearance shape of the weld bead;
Image generating means for decomposing color imaging data acquired by the color imaging means into a plurality of color components and generating an inspection image of each color component;
A mask area generating means for extracting a predetermined image area from an inspection image generated with one color component out of the inspection images generated by the image generating means, and generating a mask area;
Based on the mask area generated by the mask area generation means, the analysis image in the non-mask area is extracted by masking the corresponding position of the inspection image generated with other color components, and the analysis image Image processing means for performing predetermined image processing;
And a quality inspection means for inspecting the quality of the weld bead based on the analysis image subjected to the image processing by the image processing means. The color imaging means receives the reflected light of the infrared slit light irradiated along the length direction of the weld bead and the range light irradiated toward the predetermined measurement range of the weld bead, respectively. Take a color image of the appearance of the bead,
The image generation means generates an inspection image of each color component of RGB,
2. The mask area is generated by extracting an image area including a width direction area of a weld bead from the B component inspection image generated by the image generation means. The weld bead quality inspection device described. The image processing means masks the corresponding position of the R component inspection image generated by the image generation means based on the mask area generated by the mask area generation means;
The said quality inspection means determines the presence or absence of the hole defect of a weld bead surface from the several striped line which appeared in the analysis image by which the image processing was performed by the said image processing means. Weld bead quality inspection device.   The quality inspection unit divides the analysis image that has been subjected to image processing by the image processing unit into an inner region formed inside a predetermined amount with respect to the width direction of the weld bead and an outer region other than the inner region. 4. The weld bead quality inspection device according to claim 3, wherein the position and type of the hole defect are specified.   The quality inspection means calculates a difference between an image area of the analysis image image-processed by the image processing means and a preset reference image area, and determines a quality defect in the width direction of the weld bead. The quality inspection device for a weld bead according to any one of claims 2 to 4, wherein the quality inspection device is a weld bead. The color imaging means receives reflected light of range light irradiated toward a predetermined measurement range of the weld bead, and performs color imaging of the appearance shape of the weld bead,
The image generation means generates an inspection image of each component of RGB,
The mask area generation unit generates a mask area by extracting an image area of a predetermined area including a shading portion of a weld bead surface from an R component inspection image generated by the image generation unit. Item 3. A weld bead quality inspection device according to Item 1. The image processing means masks the corresponding position of the B component inspection image generated by the image generation means based on the mask area generated by the mask area generation means;
The quality inspection means calculates the area of the shaded part that appears in the analysis image that has been subjected to image processing by the image processing means, and determines the quality defect of the hole defect formed on the surface of the weld bead. The weld bead quality inspection device according to claim 6. In the weld bead quality inspection method for inspecting weld defects of the weld bead based on the appearance shape of the weld bead,
A color imaging step of receiving reflected light of the irradiation light irradiated toward the surface of the weld bead and performing color imaging of the appearance shape of the weld bead;
An image generation step of decomposing the color imaging data acquired by the color imaging step into a plurality of color components and generating an inspection image of each color component;
A mask region generation step of generating a mask region by extracting a predetermined image region from an inspection image generated with one color component among the inspection images generated by the image generation step;
Based on the mask region generated by the mask region generation step, the analysis image in the non-mask region is extracted by masking the corresponding position of the inspection image generated with other color components, and the analysis image An image processing step for performing predetermined image processing;
A quality inspection step for inspecting the quality of the weld bead based on the analysis image subjected to the image processing in the image processing step. The color imaging step receives the reflected light of the infrared slit light irradiated along the length direction of the weld bead and the range light irradiated toward the predetermined measurement range of the weld bead, respectively. Take a color image of the appearance of the bead,
The image generation step generates an inspection image of each color component of RGB,
9. The mask area generation step generates a mask area by extracting an image area including a width direction area of a weld bead from an inspection image of a B component generated by the image generation process. The quality inspection method of the described weld bead. The image processing step masks the corresponding position of the R component inspection image generated by the image generation step based on the mask region generated by the mask region generation unit;
The said quality inspection process determines the presence or absence of the hole defect of a weld bead surface from the several striped line which appeared in the analysis image by which the image process was performed by the said image processing process. Weld bead quality inspection method. The color imaging step receives reflected light of a range light irradiated toward a predetermined measurement range of the weld bead, and color-images the appearance shape of the weld bead,
The image generation step generates an inspection image of each component of RGB,
The mask area generation step extracts a predetermined image area including a shading portion on the surface of the weld bead from the R component inspection image generated by the image generation process to generate a mask area. Item 9. A method for quality inspection of a weld bead according to Item 8. The image processing step masks the corresponding position of the B component inspection image generated by the image generation step based on the mask region generated by the mask region generation step;
The quality inspection step is characterized by calculating an area of a shaded portion appearing in an analysis image subjected to image processing by the image processing step, and determining a quality defect of a hole defect formed on a weld bead surface. The weld bead quality inspection method according to claim 11.

Images