Flaw detection method, inspection system, and program

Abstract

To provide a flaw detection method capable of improving the detection accuracy of a flaw on the surface of a slab.SOLUTION: In a flaw detection method, a surface of a steel piece is imaged by a camera to generate a pre-marking image, a flaw on the surface of the steel piece is marked with a coloring material, the surface of the steel piece on which the flaw is marked is imaged by the camera to generate a post-marking image, and the marked flaw is detected on the basis of a difference between the pre-marking image and the post-marking image.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a defect detection method, an inspection system, and a program. [Background technology]

[0002] Patent Document 1 discloses a technology in which a mark is made using a coloring agent at a location where a scratch has formed on the surface of a steel piece, and the surface of the steel piece, including the edge, is photographed using an image capturing device while the steel piece is being moved, and the position of the mark and edge in the image is detected based on the mark and edge. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6627389 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the above-mentioned techniques may not be able to detect defects on the surface of a steel billet with sufficient accuracy.

[0005] The present invention has been made in view of the above-mentioned problems, and a main object of the present invention is to provide a defect detection method, an inspection system, and a program that can improve the accuracy of detecting defects on the surface of a steel billet. [Means for solving the problem]

[0006] In order to solve the above problems, one aspect of the present invention provides a defect detection method comprising the steps of capturing an image of the surface of a billet with a camera to generate a pre-marking image, marking defects on the surface of the billet with a coloring agent, capturing an image of the surface of the billet with the camera on which the defects have been marked to generate a post-marking image, and detecting the marked defects based on the difference between the pre-marking image and the post-marking image. This method makes it possible to improve the accuracy of detecting defects on the surface of the billet.

[0007] The above aspect may further include a step of calculating an actual position of the defect in the billet based on an in-image position of the defect in the post-marking image. This makes it possible to calculate the actual position of the defect.

[0008] In the above aspect, the method may further include a step of extracting an edge of the billet in the post-marking image, and the step of calculating the actual position of the defect may calculate the actual distance from the edge of the billet to the defect based on the in-image distance from the edge to the defect in the post-marking image. This makes it possible to calculate the actual distance from the edge of the billet to the defect.

[0009] In the above aspect, the method may further include a step of detecting an actual distance from the rangefinder to an edge of the billet using the rangefinder, the camera capturing an image of a field of view including the surface of the billet and the reference marker, and the step of calculating the actual position of the defect may calculate the actual distance from the edge of the billet to the defect based on the actual distance from the rangefinder to the edge of the billet, the actual distance from the rangefinder to the camera, and the in-image distance from a reference position determined by the reference marker in the post-marking image to the defect. This makes it possible to calculate the actual distance from the edge of the billet to the defect.

[0010] In the above aspect, the method may further include a step of using the camera to capture an image of a field of view including the surface of the billet and the background while illuminating the background of the billet with light, and extracting an edge of the billet based on the contrast between the surface of the billet and the background in the pre-marking image or the post-marking image. This makes it possible to extract the edge of the billet.

[0011] In the above aspect, the step of marking the flaw may include detecting the flaw by a magnetic particle inspection test using fluorescent magnetic powder. This makes it possible to detect the flaw by the magnetic particle inspection test.

[0012] In the above aspect, the camera may be capable of detecting red, green, and blue, have a filter that cuts blue light, and capture an image of the surface of the billet under ultraviolet light irradiation, which makes it possible to capture an image of the surface of the billet even under ultraviolet light irradiation.

[0013] In the above aspect, the defect detection step may be performed based on a difference between a green component of the pre-marking image and a green component of the post-marking image, thereby further improving the defect detection accuracy.

[0014] In the above aspect, the fluorescent magnetic particles may emit green fluorescence, which can further improve the accuracy of flaw detection.

[0015] In the above aspect, the coloring material may be white or yellow, which can further improve the accuracy of flaw detection.

[0016] In the above aspect, the method may further include a step of: using the camera to capture an image of a field of view including the surface and background of the billet while illuminating the background of the billet with red light; and extracting the edge of the billet based on the contrast between the surface and background of the billet in the red component of the pre-marking image or the post-marking image. This makes it possible to improve 414 the accuracy of extracting the edge of the billet.

[0017] Another aspect of the present invention provides an inspection system that includes a camera that captures an image of a billet surface to generate a pre-marking image, and an image of the billet surface on which defects have been marked with a coloring agent to generate a post-marking image, and a processing unit that detects the marked defects based on the difference between the pre-marking image and the post-marking image, thereby enabling improved detection accuracy for defects on the billet surface.

[0018] In addition, a program according to another aspect of the present invention causes a computer to execute the following steps: acquiring a pre-marking image of the surface of a steel billet captured by a camera; acquiring a post-marking image of the surface of the steel billet on which defects have been marked with a coloring agent captured by the camera; and detecting the marked defects based on the difference between the pre-marking image and the post-marking image. This makes it possible to improve the accuracy of detecting defects on the surface of the steel billet. [Effects of the Invention]

[0019] According to the present invention, it is possible to improve the accuracy of detecting defects on the surface of a steel billet. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 10 is a diagram illustrating an example of a procedure for a flaw detection method. [Figure 2] FIG. 1 is a diagram illustrating an example of the configuration of an inspection system. [Figure 3] FIG. 1 is a diagram illustrating an example of the configuration of an inspection system. [Figure 4] FIG. 1 is a diagram illustrating an example of the configuration of an inspection system. [Figure 5] FIG. 10 is a diagram illustrating an example of a procedure for marking detection processing. [Figure 6] FIG. 10 is a diagram for explaining a marking detection process. [Figure 7] FIG. 10 is a diagram for explaining a marking detection process. [Figure 8] FIG. 10 is a diagram illustrating an example of a procedure for outputting a marking position. [Figure 9] FIG. 10 is a diagram for explaining a marking position output. [Figure 10] FIG. 10 is a diagram for explaining a marking position output. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In this specification and the drawings, elements similar to those described above with reference to the previous drawings will be designated by the same reference numerals, and detailed descriptions thereof may be omitted as appropriate.

[0022] Fig. 1 is a diagram showing an example of the procedure of the flaw detection method. Fig. 2 and Fig. 3 are diagrams showing an example of the configuration of an inspection system 100 for implementing the flaw detection method, mainly showing the configuration installed in a visual inspection room (VEM). Fig. 4 is a block diagram showing an example of the configuration of the inspection system 100.

[0023] The X and Y directions in the figure are horizontal directions, with the X direction being the longitudinal direction of the billet BL and the Y direction being the width direction of the billet BL. The Z direction is the upward direction. The billet BL has, for example, a rectangular pillar shape extending in the longitudinal direction X.

[0024] The defect detection method is a method for detecting the location of defects that have occurred on the surface of a steel billet (BL) and for removing the detected defects, and is used in steel mills, metal processing plants, etc. Black or similar colored defects occur on the surface of the steel billet (BL) during the manufacturing process.

[0025] As shown in Figure 2, the visual inspection room VEM is equipped with multiple ultraviolet light sources 7 that irradiate the steel billet BL with ultraviolet light. The visual inspection room VEM is a darkroom. The ultraviolet light sources 7 are installed above the steel billet BL and lined up in the longitudinal direction X.

[0026] As shown in Fig. 3, in the visual inspection room VEM, a defect inspection stand 9 is installed beside and along the steel billet BL. On the defect inspection stand 9, a plurality of reference markers 8 are installed at predetermined intervals P in the longitudinal direction X.

[0027] The inspection system 100 is equipped with multiple cameras 2 that capture images of the surface of the steel block BL. Similar to the ultraviolet light source 7, the cameras 2 are installed in a row in the longitudinal direction above the steel block BL, and form a field of view FV looking directly down on the steel block BL (see FIG. 2).

[0028] The camera 2 is a digital camera capable of detecting red, green, and blue. In this embodiment, the image of the steel billet BL by the camera 2 is captured in a visual inspection room VEM under ultraviolet irradiation from an ultraviolet light source 7, and the lens of the camera 2 is provided with a cut filter that blocks blue light to avoid the effects of ultraviolet light.

[0029] Each camera 2 captures an image of a field of view FV that includes the surface of the billet BL, the background of the billet BL, and the reference marker 8 (see Figure 3). The field of view FV is a collective term for fields of view FV1 to FV4. The backgrounds are located on both sides of the billet BL in the width direction Y when viewed from above. Specifically, the background is the ground of the visual inspection room VEM.

[0030] Two adjacent fields of view FV are set so that their ends in the longitudinal direction X overlap. Furthermore, each field of view FV is set so that a reference position RL determined by a reference marker 8 is included in the center in the longitudinal direction X. The reference position RL is a general term for the reference positions RL1 to RL4.

[0031] The inspection system 100 is equipped with a plurality of background illuminators 3 for illuminating the background of the steel block BL. Specifically, the background illuminators 3 are installed below the steel block BL and illuminate the light toward the ground in the visual inspection room VEM (see FIG. 3). In this embodiment, the light from the background illuminators 3 is red light.

[0032] The camera 2 captures an image of the surface and background of the billet BL while illuminating the background of the billet BL with light from the background illuminator 3. Illuminating the background with light makes it possible to accurately detect the edge of the billet BL in the image. Furthermore, by pointing the background illuminator 3 toward the ground, it is possible to reduce the adhesion of flaw detection fluid and dirt.

[0033] The distance meter 4 measures the actual position of the edge of the billet BL in the longitudinal direction X (see FIG. 3). The distance meter 4 is, for example, a laser distance meter. Specifically, the distance meter 4 measures the distance B from the distance meter 4 to the edge of the billet BL in the longitudinal direction X.

[0034] In the following description, the edge in the longitudinal direction X of the billet BL, whose actual position is measured by the rangefinder 4, is also referred to as the "tip of the billet BL." In addition, the camera 2 and the reference marker 8 may be referred to by ordinal numbers in order of proximity to the rangefinder 4.

[0035] A distance A from the range finder 4 to the reference position RL1 of the first reference marker 8 is measured in advance. The actual position of the leading edge of the billet BL is at a distance B from the range finder 4, and at a distance AB from the reference position RL1 of the first reference marker 8.

[0036] 4, the inspection system 100 includes a PC (personal computer) 10. The PC 10 is connected to a camera 2, a background light 3, a rangefinder 4, and a storage unit 19.

[0037] The PC 10 is a computer including a CPU, RAM, ROM, non-volatile memory, an input / output interface, etc., and executes information processing according to a program. The program may be supplied from a non-transitory storage medium or via a communication line.

[0038] The PC 10 includes an image capture processing unit 11 and a mark detection unit 12. These functional units are realized by a CPU executing information processing in accordance with a program. The operations of the image capture processing unit 11 and the mark detection unit 12 will be described later.

[0039] An example of the procedure of the flaw detection method shown in FIG. 1 will be described below.

[0040] The flaw detection methods S1 to S4 are performed in a visual inspection room VEM. S1 to S2 and S4 are realized by an image processing unit 11 of the PC 10, and S5 to S6 are realized by a mark detection unit 12 of the PC 10.

[0041] S1 is a step of turning on the background light 3. When the billet BL is carried to the inspection position by an elevator (not shown), the background light 3 is turned on to irradiate the background of the billet BL with light.

[0042] S2 is a step in which the surface of the steel billet BL is imaged by the camera 2 to generate a pre-marking image before the inspector marks defects.

[0043] S3 is a process in which defects on the surface of the steel billet BL are marked with chalk. In this process, inspectors visually search for defects on the surface of the steel billet BL and mark any defects they find with chalk.

[0044] Chalk is an example of a coloring material. However, other materials such as pens, paints, or colored tape may be used. The color of the chalk is preferably a color that is easily distinguishable from the surface of the steel piece BL and is also different from the background lighting 3, such as white or yellow.

[0045] Defects on the surface of the steel billet BL are detected by magnetic particle testing using fluorescent magnetic powder. In magnetic particle testing, the steel billet BL is magnetized, the fluorescent magnetic powder is scattered on the surface of the steel billet BL, ultraviolet light is irradiated from an ultraviolet light source 7, and the fluorescence emitted by the fluorescent magnetic powder that has gathered in the leakage magnetic field of the defects is visually confirmed by an inspector. In this embodiment, fluorescent magnetic powder that emits green fluorescence is used.

[0046] S4 is a step in which, after the inspector has marked the defects, the surface of the steel billet BL on which the defects have been marked is photographed by the camera 2 to generate a post-marking image.

[0047] S5 is a marking detection process. This process is a process for detecting marked defects in the post-marking image. Details of the marking detection process S5 will be described later.

[0048] S6 is a marking position output. This is a step of calculating and outputting the actual position of the defect on the steel billet BL based on the in-image position of the marked defect in the post-marking image. Details of the marking position output S6 will be described later.

[0049] S7 is a grinding process. In this process, the actual defect positions output in S6 are used to grind away the defects using a defect removal device such as a grinder. The defect grinding may be automated or performed manually.

[0050] A specific example of the procedure for the marking detection process S5 and the marking position output S6 will be described below.

[0051] Fig. 5 is a flow diagram showing an example of the procedure of the marking detection process S5. The PC 10 of the inspection system 100 executes the information processing shown in the diagram in accordance with a program. Figs. 6 and 7 are diagrams for explaining the marking detection process S5.

[0052] First, the PC 10 acquires a pre-marking image BF and a post-marking image AF captured by the camera 2 (S51). Next, the PC 10 performs predetermined pre-processing on the pre-marking image BF and the post-marking image AF (S52).

[0053] Next, the PC10 recognizes the edge ED of the steel block BL in the width direction Y (S53). Specifically, the PC10 extracts the edge ED of the steel block BL in the width direction Y based on the contrast between the surface of the steel block BL and the background HK in the pre-marking image BF or the post-marking image AF.

[0054] In this embodiment, the camera 2 captures an image with red light illuminating the background of the billet BL using the background illumination 3, and the PC 10 extracts the edge ED of the billet BL in the width direction Y based on the contrast between the surface of the billet BF and the background HK in the red component. The red light illuminating the background is a different color from the fluorescent magnetic powder and chalk, making it easier to extract the edge ED.

[0055] Next, the PC 10 recognizes the edge of the metal block BL in the longitudinal direction X (S54). The actual position of the leading edge of the metal block BL is measured by the range finder 4, and is located at a distance B from the range finder 4 (see FIG. 3). The actual position of the leading edge of the metal block BL is also located at a distance AB from the reference position RL1 of the first reference marker 8.

[0056] Next, the PC 10 detects the chalk marks MK affixed to the flaws (S55), i.e., detects the marked flaws. As shown in Fig. 6, the PC 10 calculates a difference image DF between the pre-marking image BF and the post-marking image AF, and applies binarization processing to the difference image DF to detect the chalk marks MK.

[0057] In this embodiment, the detection of the chalk mark MK is performed based on the difference between the green component of the pre-marking image BF and the green component of the post-marking image AF. Because the camera 2 includes many green sensors to match the human visual sensitivity, it is preferable to compare the green components of both images to detect white or yellow chalk.

[0058] In particular, the lens of camera 2 is equipped with a cut filter that cuts out blue light, and the chalk mark MK in the post-marking image AF appears green or yellow, and is represented primarily by the contrast of the green component, so it is preferable to compare the green component of both images.

[0059] Not limited to this, the chalk mark MK can be detected by calculating the difference between the red, green, and blue components of the pre-marking image BF and the post-marking image AF, and the chalk mark MK can be detected as a color image, or by converting it to grayscale.

[0060] Furthermore, after the binarization process, the PC 10 labels each of the chalk marks MK, treating each group of consecutive bright areas exceeding the threshold as a single chalk mark MK, and determines a detection frame SC circumscribing the chalk mark MK for each label (see Figure 7). In this way, the chalk mark MK is detected in the shape of the detection frame SC that surrounds it.

[0061] Before binarizing the difference image DF, PC10 may perform a sharpening process to emphasize the chalk marks MK or a bilateral filter to reduce rubbing of the chalk marks MK, or after binarizing the difference image DF, it may expand bright parts that exceed a threshold value using an expansion process to combine adjacent bright parts.

[0062] Fig. 8 is a flow diagram showing an example of the procedure for marking position output S6. The PC 10 of the inspection system 100 executes the information processing shown in the diagram in accordance with a program. Figs. 9 and 10 are diagrams for explaining the marking position output S6.

[0063] First, the PC 10 calculates the chalk mark coordinates that represent the actual positions of the chalk marks MK on the billet BL based on the intra-image position of the detection frame SC in the post-marking image AF (S61).

[0064] 9 and 10, the chalk mark coordinates are expressed by distances X1 and X2 along the longitudinal direction X and distances Y1 and Y2 along the width direction Y, based on an origin O defined on the billet BL. The origin O of the billet BL is defined, for example, at the leading end of the billet BL and one end in the width direction Y (the lower left end in FIG. 9).

[0065] Distance X1 is the distance from the origin O to the near end of the detection frame SC along the longitudinal direction X, and distance X2 is the distance to the far end. Distance Y1 is the distance from the origin O to the near end of the detection frame SC along the width direction Y, and distance Y2 is the distance to the far end.

[0066] The distances Y1 and Y2 in the width direction Y are calculated as follows: The PC 10 uses the edge ED in the width direction Y of the steel billet BL extracted in the post-marking image AF in S53 above to calculate the distances Y1 and Y2 in the width direction Y based on the in-image distance in the width direction Y from the edge ED to the detection frame SC. The correspondence relationship between the in-image distance and the actual distance is determined in advance.

[0067] The distances X1 and X2 in the longitudinal direction X are calculated as follows: The PC 10 calculates the distances X1 and X2 in the longitudinal direction X based on the actual position of the leading edge of the steel billet BL measured by the rangefinder 4 and the in-image distance in the longitudinal direction X from the reference position RL to the detection frame SC in the post-marking image AF.

[0068] For example, as shown in Figure 3, when the field of view FV2 of the second camera 2 includes the detection frame SC, the distance C from the reference position RL2 of the second reference marker 8 to the detection frame SC is calculated based on the in-image distance in the longitudinal direction X from the reference position RL2 to the detection frame SC in the post-marking image AF generated by the second camera 2.

[0069] The actual position of the tip of the steel piece BL is at a distance B from the range finder 4, and the actual position of the reference position RL2 of the second reference marker 8 is at a distance A+P from the range finder 4, so the distance in the longitudinal direction X from the tip of the steel piece BL to the detection frame SC is A+P+CB.

[0070] It is also possible to determine in advance where in the longitudinal direction X the center of the camera 2 is located, and then determine the distance in the longitudinal direction X from the leading edge of the billet BL to the detection frame SC based on this determination, without using the reference marker 8. However, even in this case, it is easier to adjust the field of view FV of the camera 2 if the reference marker 8 is present.

[0071] Next, the PC 10 outputs the calculated chalk mark coordinates to the database of the storage unit 19 or a higher-level system (S62). The chalk mark coordinates thus output are used when specifying a grinding position for a flaw removal device in the subsequent grinding step S7.

[0072] According to the embodiment described above, the detection accuracy of the chalk marks MK can be improved by detecting the chalk marks MK based on the difference between the pre-marking image BF and the post-marking image AF. In addition, by outputting the chalk mark coordinates that represent the actual positions of the chalk marks MK, the defect positions and occurrence frequency can be quantified, which can be utilized for quality analysis and automation of surface defect removal.

[0073] The method disclosed in Patent Document 1 involves transporting the billet to a position away from the visual inspection equipment to detect the marking position, and since it is assumed that images are taken while the billet is moving, there is a risk that the chalk marks on the billet surface will fade due to contact with the roller table during transport, making it difficult to detect the chalk marks. Also, while the image can be used to analyze quality by retaining the pattern of flaws, there is a risk that the image cannot be used directly for quality analysis because the magnetic powder liquid dries out when the billet is transported to a position away from the visual inspection equipment and the optical conditions are different from those of the visual inspection equipment.

[0074] In contrast to this, in this embodiment, the surface of the steel billet BL is imaged by the camera 2 in the visual inspection room VEM, so it is possible to solve the problems that are a concern with the method disclosed in Patent Document 1 above.

[0075] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and it goes without saying that various modifications can be made by those skilled in the art. [Explanation of symbols]

[0076] 2 camera, 3 background lighting, 4 rangefinder, 7 ultraviolet light source, 8 reference marker, 9 defect inspection stand, 10 PC, 11 photography processing unit, 12 mark detection unit, 19 memory unit, 100 inspection system, VEM visual inspection room, BL steel billet, FV field of view, RL reference position, BF image before marking, AF image after marking, DF difference image, ED edge, HK background, MK chalk mark, SC detection frame

Claims

1. a step of capturing an image of the surface of the steel billet with a camera to generate a pre-marking image; marking defects on the surface of the steel billet with a coloring material; a step of capturing an image of the surface of the steel billet on which the defect has been marked by the camera to generate a post-marking image; detecting the marked defect based on a difference between the pre-marking image and the post-marking image; A flaw detection method comprising:

2. The method further includes a step of calculating an actual position of the defect in the billet based on an in-image position of the defect in the post-marking image. The flaw detection method according to claim 1 .

3. The method further includes extracting an edge of the billet in the post-marking image, the step of calculating the actual position of the defect includes calculating the actual distance from the edge of the billet to the defect based on an in-image distance from the edge to the defect in the post-marking image; The flaw detection method according to claim 2.

4. further comprising a step of detecting an actual distance from the range finder to an edge of the billet by the range finder; the camera captures an image of a field of view including the surface of the steel piece and the fiducial marker; the step of calculating the actual position of the defect includes calculating the actual distance from the edge of the billet to the defect based on the actual distance of the edge from the range finder to the billet, the actual distance from the range finder to the camera, and the in-image distance from a reference position determined by the reference marker in the post-marking image to the defect; The flaw detection method according to claim 2.

5. the camera captures an image of a field of view including the surface of the billet and the background while illuminating the background of the billet with light; extracting an edge of the billet based on a contrast between a surface of the billet and a background in the pre-marking image or the post-marking image. The flaw detection method according to claim 1 .

6. The step of marking the defects includes detecting the defects by a magnetic particle inspection test using fluorescent magnetic particles. The flaw detection method according to claim 1 .

7. The camera is capable of detecting red, green, and blue colors; It has a filter that cuts blue light, imaging the surface of the billet under ultraviolet irradiation; The flaw detection method according to claim 6.

8. the defect detection step is performed based on a difference between a green component of the pre-marking image and a green component of the post-marking image. The flaw detection method according to claim 7.

9. The fluorescent magnetic particles emit green fluorescence. The flaw detection method according to claim 8.

10. The colorant is white or yellow. The flaw detection method according to claim 8.

11. the camera captures an image of a field of view including the surface of the billet and the background while irradiating the background of the billet with red light; extracting an edge of the billet based on a contrast between a surface of the billet and a background in a red component of the pre-marking image or the post-marking image. The flaw detection method according to claim 8.

12. a camera that captures an image of the surface of the steel billet to generate a pre-marking image, and captures an image of the surface of the steel billet on which defects have been marked with a coloring material to generate a post-marking image; a processing unit that detects the marked defect based on a difference between the pre-marking image and the post-marking image; An inspection system comprising:

13. Obtaining a pre-marking image of the surface of the steel piece by capturing the image with a camera; Obtaining a post-marking image by capturing an image of the surface of the steel billet on which the defects have been marked with a coloring material using the camera; and detecting the marked defect based on a difference between the pre-marking image and the post-marking image; A program that causes a computer to execute the following.

Images