Casting defect detection method and casting defect detection device

Abstract

High-precision detection of casting defects from X-ray images of castings. [Solution] A method for detecting casting defects using X-ray images of a casting is configured to include an acquisition step of acquiring an X-ray image of the casting, a generation step of generating a casting defect detection image by adding a periodic image of a striped pattern in which the pixel values ​​change periodically in a predetermined direction to the X-ray image, and a detection step of detecting casting defects based on the detection image.

Description

[Technical Field]

[0001] The present disclosure relates to a porosity detection method and a porosity detection apparatus for detecting porosity in a cast product using an X-ray image of the cast product. [Background Art]

[0002] Conventionally, techniques for detecting porosity generated during casting using X-ray images of cast products are known. When a cast product manufactured by pouring metal or resin into a mold has porosity caused by cavities or air bubbles, the porosity appears brighter than the surrounding area on an X-ray image captured by transmitting X-rays. Using this contrast, regions with high luminance on the X-ray image of a cast product can be detected as porosity. For example, Patent Document 1 discloses a technique for detecting porosity based on a luminance difference by comparing an X-ray image obtained by imaging a cast product with a master image of a component free of porosity.

[0003] If a threshold value is set based on the luminance difference between the porosity appearing in the X-ray image of the cast product and the surrounding area thereof, a region having a luminance higher than the threshold value can be detected as porosity. Even when an X-ray image includes a region having a different luminance from other regions due to the shape of the cast product, if adaptive threshold processing for setting a local threshold corresponding to the region is performed, a region having a luminance higher than the set threshold can be detected as porosity. [Prior Art Literature] [Patent Literature]

[0004] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2018-96955 [Summary of Invention] [Problem to be Solved by Invention]

[0005] However, even when applying the above-mentioned conventional techniques, there are cases where casting defects cannot be detected. For example, if the brightness of a region containing a casting defect gradually increases or decreases in an X-ray image due to changes in its shape, it becomes difficult to detect the casting defect based on the brightness difference between the defect and the surrounding region.

[0006] This disclosure has been made in view of the problems of the prior art described above, and one of its purposes is to provide a method and apparatus for detecting casting defects that can detect casting defects with high precision using X-ray images of castings. [Means for solving the problem]

[0007] The method for detecting casting defects according to this disclosure is a method for detecting casting defects using an X-ray image of a casting, and includes an acquisition step of acquiring an X-ray image of a casting, a generation step of generating a casting defect detection image by adding a periodic image of a striped pattern in which the pixel values ​​change periodically in a predetermined one direction to the X-ray image, and a detection step of detecting casting defects based on the detection image.

[0008] In the above configuration, the change in the pixel value in one direction in the periodic image may exhibit a sine wave shape or a triangular wave shape.

[0009] In the above configuration, the generation step includes a step of generating a plurality of detection images by adding to the X-ray image a plurality of periodic images in which at least one of the following is different: the phase in one direction, the angle indicating the one direction, the period of the pixel value change in one direction, and the amplitude of the pixel value change in one direction, and the detection step may include a step of detecting casting defects on each of the plurality of detection images generated in the generation step.

[0010] In the above configuration, the detection step may include a step of detecting casting defects based on the pixel value of each pixel that forms the X-ray image.

[0011] In the above configuration, the output step may further include synthesizing an image showing the detection result of a casting defect obtained from each of the multiple detection images in the detection step using a predetermined method, and outputting it as the casting defect detection result.

[0012] In the above configuration, the method may further include a step of determining whether a region is a casting defect or not based on the area of ​​the region detected as a casting defect from the detection image, and excluding regions that are determined not to be casting defects from the casting defect detection results.

[0013] The casting defect detection device according to this disclosure is a casting defect detection device that uses an X-ray image of a casting to detect a casting defect, and comprises an acquisition unit that acquires an X-ray image of a casting, and a detection unit that generates a casting defect detection image by adding a periodic image of a striped pattern in which the pixel values ​​change periodically in a predetermined direction to the X-ray image, and detects a casting defect based on the detection image. [Effects of the Invention]

[0014] According to the casting defect detection method and casting defect detection apparatus described herein, casting defects occurring in a casting can be detected with high accuracy by utilizing an X-ray image obtained by transmitting X-rays through the casting. [Brief explanation of the drawing]

[0015] [Figure 1] Figure 1 is a schematic diagram illustrating the principle of the casting defect detection method according to this embodiment. [Figure 2] Figure 2 is a flowchart showing an example of the process for detecting casting defects. [Figure 3] Figure 3 is a diagram illustrating an example of a periodic image. [Figure 4] Figure 4 is a diagram illustrating an example of image generation for detection. [Figure 5] Figure 5 is a diagram illustrating examples of several types of periodic images. [Figure 6] Figure 6 illustrates the post-processing of casting defect detection results obtained using periodic images. [Modes for carrying out the invention]

[0016] Hereinafter, embodiments of the shrinkage cavity detection method and shrinkage cavity detection apparatus according to the present disclosure will be described with reference to the accompanying drawings. Fig. 1 is a schematic diagram for explaining the outline of the principle of the shrinkage cavity detection method according to the present embodiment. As shown in Fig. 1, an X-ray irradiating apparatus 10 irradiates X-rays to a cast product 100 which is a shrinkage cavity detection target for checking whether there is a shrinkage cavity. An imaging device 20 receives X-rays transmitted through the cast product 100 and captures an X-ray image (X-ray transmission image) of the cast product 100.

[0017] The shrinkage cavity detection apparatus 1 acquires (A) an X-ray image 200 of the cast product 100 captured by the imaging device 20. The shrinkage cavity detection apparatus 1 may acquire the X-ray image 200 output from the imaging device 20 in real time, or may acquire the X-ray image 200 stored in a storage device after imaging by the imaging device 20. The configuration of the shrinkage cavity detection apparatus is not particularly limited as long as the shrinkage cavity detection method described in the present embodiment can be executed. For example, a computer apparatus including a storage unit, a control unit and a communication unit may be used as the shrinkage cavity detection apparatus 1. The shrinkage cavity detection apparatus 1 may be in an aspect including the X-ray irradiation apparatus 10 and the imaging apparatus 20.

[0018] The shrinkage cavity detection apparatus 1 detects shrinkage cavities using the grayscale X-ray image 200. Before the shrinkage cavity detection apparatus 1 executes the process of detecting shrinkage cavities, it may execute preprocessing image processing such as a process of removing noise from the X-ray image 200 and a process of adjusting contrast.

[0019] The shrinkage cavity detection apparatus 1 can detect a shrinkage cavity (B) based on the luminance difference between the shrinkage cavity and the surrounding area that appears in the X-ray image 200 of the cast product 100. That is, the shrinkage cavity detection apparatus 1 can detect the shrinkage cavity based on the contrast appearing in the X-ray image 200.

[0020] For example, as shown in FIG. 1, when a shrinkage cavity 301 appears brighter in an image 201 of a partial region 200a on an X-ray image 200 than in the surrounding area, the shrinkage cavity detection apparatus 1 detects the position of the shrinkage cavity 301 in the cast product 100 based on the luminance difference between the shrinkage cavity 301 and the surrounding area, that is, the contrast.

[0021] In FIG. 1, below the image 201 of the partial region 200a cut out from the X-ray image 200, a luminance curve 211 showing the relationship between the position of each pixel on a straight line passing through the approximate center of the shrinkage cavity 301 in the horizontal direction (left-right direction in the drawing) and the pixel value (luminance value) is shown. Similarly, for other images 202, 402, and 502, luminance curves 212, 412, and 512 showing the relationship between the position of each pixel and the pixel value are shown below each image. Note that since the image processed by the shrinkage cavity detection apparatus 1 is a grayscale image, the luminance value of each pixel forming the image corresponds to the pixel value.

[0022] When the cast product 100 has the shrinkage cavity 301, the shrinkage cavity 301 appears at a higher luminance than the surrounding area on the X-ray image 200. Therefore, as shown in the luminance curve 211, a pixel value difference (luminance difference) occurs between the shrinkage cavity 301 and the peripheral region of the shrinkage cavity 301. The shrinkage cavity detection apparatus 1 detects the shrinkage cavity 301 from the X-ray image 200 using this luminance difference. For example, the shrinkage cavity detection apparatus 1 executes adaptive threshold processing to set a pixel value threshold for detecting the shrinkage cavity 301, and detects the shrinkage cavity 301 based on a comparison result between the pixel value of each pixel and the threshold.

[0023] In addition to detecting shrinkage cavities based on the luminance difference appearing in the X-ray image 200, the shrinkage cavity detection apparatus 1 can change the luminance value of the X-ray image 200 and execute shrinkage cavity detection processing based on the changed luminance value (C).

[0024] For example, due to changes in the shape or thickness of the casting 100, the image 202 of the partial region 200b in which the casting defect 302 is visible on the X-ray image 200 may be an image in which the brightness gradually changes, as shown in the lower left of Figure 1. As can be seen from the brightness curve 212 shown below the image 202, in such an image 202, it is difficult to detect the casting defect 302 by simply setting a threshold pixel value, as in the case of image 201 where the pixel value in the surrounding region of the casting defect 301 is approximately constant.

[0025] Therefore, the casting defect detection device 1 generates a detection image 502 by adding a superimposed image 402, which shows a change in phase opposite to the change in pixel values ​​in the X-ray image 202, to the X-ray image 202 as an image for detecting the casting defects 302. In other words, the superimposed image 402, which shows a brightness gradient opposite to the brightness gradient that appeared in the X-ray image 202, is added to the X-ray image 202 to generate a detection image 502 of the casting defects 302.

[0026] For example, as shown in Figure 1, if the X-ray image 202 is an image in which the pixel values ​​gradually increase as you move to the right, the casting defect detection device 1 adds a superimposed image 402 in which the pixel values ​​gradually decrease as you move to the right to the X-ray image 202. As a result, it is possible to obtain a detection image 502 that cancels out the changes in pixel values ​​caused by the shape of the casting 100, while retaining the changes in pixel values ​​caused by the casting defects 302.

[0027] Specifically, as shown in Figure 1, when the pixel values ​​on the X-ray image 202 change as shown by the brightness curve 212, the pixel values ​​of each pixel forming the brightness curve 412, which has the opposite phase and changes in the opposite direction to the brightness curve 212, are added to the pixel values ​​of each pixel forming the brightness curve 212. As a result, the pixel values ​​of each pixel forming the brightness curve 212 change to the pixel values ​​shown by the brightness curve 512.

[0028] The detection image 502, like image 201, is an image in which the casting defect 302 is shown to be brighter than the surrounding area. The casting defect detection device 1 uses the detection image 502 to set a threshold for detecting the casting defect 302, as in the case of image 201, and detects the casting defect 302 based on the comparison result between the pixel value of each pixel and the threshold.

[0029] The casting defect detection device 1 outputs the casting defect detection result (D). The casting defect detection device 1 outputs the detection result which includes the casting defect 301 detected based on the brightness difference that appeared in the X-ray image 201 and the casting defect 302 detected based on the brightness difference that appeared in the detection image 502 obtained by changing the pixel values ​​of the X-ray image 202.

[0030] For example, as shown in Figure 1, detection results indicating the location and size of each casting defect 301 and 302 are output on an image 600 corresponding to the X-ray image 200. The method of outputting the detection results is not particularly limited, but for example, on a black and white binary image 600, the casting defects 301 and 302 may be shown in white or black, and the other areas may be shown in the opposite color, black or white. The detection results may be displayed on a display unit if the casting defect detection device 1 is equipped with a display unit, or the results may be output from the casting defect detection device 1 to an external device via the communication unit and displayed on a display unit of the external device.

[0031] Note that in Figure 1, for the sake of simplicity, the explanation focuses on partial regions 200a and 200b. However, the casting defect detection device 1 performs two processes on the entire X-ray image 200: one that detects casting defects based on the brightness difference of the X-ray image 200, as in the example of image 201, and another that detects casting defects based on the brightness difference of a detection image generated by changing the pixel values ​​of the X-ray image 200, as in the example of image 202. Multiple casting defects may be detected in each process, and the image 600 showing the casting defect detection results may contain a large number of casting defects.

[0032] In Figure 1, a superimposed image 402 in which the pixel value decreases in one direction (to the right in the drawing) was used as an example. However, the casting defect detection device 1 detects casting defects by adding each of the multiple types of superimposed images to the X-ray image 200. By utilizing multiple types of superimposed images, the casting defect detection device 1 generates a detection image that suppresses the influence of various types of brightness gradients that appear in the X-ray image 200 due to the shape of the casting 100, and detects casting defects based on each detection image. Details will be described later.

[0033] Figure 1 illustrates an example of a process for detecting casting defects using a threshold set for detection. However, the process is not limited to threshold-based processing, as long as casting defects can be detected based on the brightness difference or contrast between the defect and the surrounding area. For example, a template image of the casting defect could be used to search for the defect by template matching, or machine learning or AI technology could be used to detect the casting defect.

[0034] In the example shown in Figure 1, a method for detecting casting defects was explained using an image in which the casting defects appear brighter than the surrounding area. However, it goes without saying that an image in which the brightness and darkness of the image are inverted, so that the casting defects appear darker than the surrounding area, can also be used. In this case as well, the casting defects can be detected based on the brightness difference between the casting defects and the surrounding area, as described above. Furthermore, although a grayscale image is used in this embodiment, a monochrome grayscale image, such as an image with only red tones, may be used instead of a grayscale image.

[0035] Next, a specific example of the casting defect detection process performed by the casting defect detection device 1 will be explained. Figure 2 is a flowchart showing an example of the casting defect detection process flow. First, the casting defect detection device 1 acquires an X-ray image of the casting (step S1).

[0036] The casting defect detection device 1 performs preprocessing of the acquired X-ray image as needed (step S2). For example, preprocessing such as removing noise from the X-ray image, adjusting the contrast, or extracting a partial region image to be targeted for casting defect detection from the X-ray image is performed as needed. If preprocessing is performed, the following processes are performed on the processed X-ray image.

[0037] The casting defect detection device 1 detects casting defects based on the brightness difference between the casting defect appearing in the X-ray image and the surrounding area (step S3). This process corresponds to the process of detecting casting defects based on the pixel values ​​of the X-ray image, as explained in image 201 of Figure 1. The casting defect detection device 1 performs the casting defect detection process on the entire X-ray image and stores the casting defect detection results in the storage unit.

[0038] Next, the casting defect detection device 1 generates a detection image from the X-ray image and the periodic image (step S4), and detects the casting defect based on the brightness difference between the casting defect that appears in the detection image and the surrounding area (step S5). These processes correspond to the process of generating a detection image 502 by adding the superimposed image 402 to the X-ray image 202 to change the brightness value, as explained in image 202 of Figure 1, and detecting the casting defect based on the brightness difference that appears in the detection image 502. The periodic image added to the X-ray image in step S4 corresponds to the superimposed image 402 explained in Figure 1, and the periodic image is an image with a striped pattern in which the pixel values ​​change periodically multiple times. For example, an image made by arranging the superimposed image 402 vertically and horizontally to the same size as the X-ray image 200 shown in Figure 1 is used as the periodic image.

[0039] Figure 3 is a diagram illustrating an example of periodic image 450. Figure 3(a) shows an example of X-ray image 250, which is the target of casting defect detection. X-ray image 250 corresponds to image 200 explained in Figure 1, and is actually a grayscale transmission image taken by irradiating a casting with X-rays, but the casting is not shown in Figure 3(a).

[0040] Figure 3(b) shows an example of a periodic image 450. A periodic image 450 is the same size as an X-ray image 250 (horizontal Wa pixels, vertical Wb pixels), but the pixel values ​​change periodically in one direction on the image. The pixel values ​​in the horizontal direction (left-right direction in the drawing) are shown at the bottom of the periodic image 450 in Figure 3(b).

[0041] In the example shown in Figure 3(b), the periodic image 450 has the same pixel value for each row of pixels arranged vertically (up and down in the drawing), while the pixel values ​​for each row of pixels arranged horizontally change periodically in a sine wave 460 shape. The change in pixel value represents a sine wave 460 shape with a wavelength (period) of Wc pixels and an amplitude of 2V (+V to -V).

[0042] In Figure 3, only a portion of region 450a of the periodic image 450 is shown. However, in reality, the periodic image 450 is an image with vertical striped regions 450a arranged horizontally, resulting in an overall image of light and dark vertical stripes. Similarly, in Figures 4 and 5, the periodic image, in which the entire image is a light and dark striped pattern, is illustrated by only a portion of the stripes.

[0043] In step S4 shown in Figure 2, the casting defect detection device 1 adds a periodic image 450, in which the pixel values ​​change periodically in one direction, to the X-ray image 250 to generate a detection image capable of detecting casting defects.

[0044] Figure 4 is a diagram illustrating an example of generating a detection image. For example, as shown in Figure 4, suppose that the X-ray image 250 has a region 250a where the pixel value gradually increases as shown by the brightness curve 261, and a region 250b where the pixel value gradually decreases as shown by the brightness curve 262, and that casting defects 311 and 312 are visible in both regions. When the image of region 450b of the periodic image 450, where the pixel value changes in approximately the opposite phase to the brightness curve 261, is added to the X-ray image 251 of region 250a, a detection image is obtained in which the casting defects 311 can be detected by setting a threshold based on the brightness difference, as explained in image 202 of Figure 1. Similarly, a detection image obtained by adding the image of region 450c of the periodic image 450, where the pixel value changes in approximately the opposite phase to the brightness curve 262, to the X-ray image 252 of region 250b, also becomes an image in which the casting defects 312 can be detected based on the brightness difference. Since the periodic image 450 includes multiple regions 450b, 450c, etc., where the pixel values ​​change, the detection image generated by adding the periodic image 450 to the X-ray image 250 becomes an image capable of simultaneously detecting multiple casting defects 311, 312, etc.

[0045] In this way, the casting defect detection device 1 generates a detection image by adding a periodic image 450, in which the pixel values ​​change periodically multiple times in one direction, to the X-ray image, thereby enabling the detection of multiple casting defects from a single detection image. In other words, the casting defect detection device 1 can simultaneously detect multiple casting defects by performing the detection process of step S5 shown in Figure 2 just once.

[0046] However, in order to obtain a detection image capable of detecting casting defects, as explained in Figure 4, the periodic image 450 must contain a region showing a brightness gradient that is approximately in opposite phase to the region on the X-ray image 250 where a brightness gradient caused by the shape of the casting appears. Also, although Figure 4 shows an example in which a brightness gradient appears in the horizontal direction in the X-ray image 250, depending on the shape of the casting 100, a brightness gradient may also appear in the vertical or diagonal direction, for example. Furthermore, the amplitude of the pixel values ​​in the periodic image 450 must be large enough to cancel out the brightness gradient that appears in the X-ray image 250 due to the shape of the casting.

[0047] Therefore, the casting defect detection device 1 prepares multiple types of periodic images in which the phase, angle, wavelength (period), and amplitude of the sine wave 460 shown in Figure 3(b) differ from those of the periodic image 450, adds each of these to the X-ray image 250 to generate an image for casting defect detection, and performs casting defect detection processing on each detection image.

[0048] Figure 5 is a diagram illustrating examples of multiple types of periodic images 450, 451, 1450, 4501, etc. As shown in Figure 5(a), the casting defect detection device 1 shifts the vertical stripe pattern of the periodic image 450 by a predetermined number of pixels (Pa pixels) in one direction perpendicular to the amplitude direction of the sine wave 460, i.e., the periodic direction in which the pixel values ​​change periodically (left-right direction in the drawing), thereby generating phase-shifted periodic images 451, etc. Both the periodic image 450 and the phase-shifted periodic image 451 are images with a vertical stripe pattern throughout, but the position of the vertical stripe pattern is shifted horizontally by Pa pixels.

[0049] If the phase of periodic image 450 is set to 0 degrees, the phase of periodic image 451 will be phase P1 (= 360 degrees / Wc pixels × Pa pixels). For example, the casting defect detection device 1 generates multiple types of periodic images 451... with a phase shift of a predetermined number of pixels (Pa pixels) within the range of one wavelength (Wc pixels) of the sine wave 460 shown in Figure 3, that is, within the phase range of 0 to 360 degrees.

[0050] As shown in Figure 5(a), the casting defect detection device 1 rotates the periodic direction of the periodic image 450, that is, the horizontal axis direction of the sine wave 460 in which the pixel values ​​change periodically, by a predetermined angle (R degrees) to generate an angle-changed periodic image 1450. The angle-changed periodic image 1450 becomes an image of linear stripes with varying brightness in the diagonal direction. For example, the casting defect detection device 1 generates multiple types of periodic images 1450… by changing the angle by a predetermined angle within an angle range of 0 to 180 degrees.

[0051] The casting defect detection device 1 generates periodic images 1451… by shifting the phase of the angle-changed periodic image 1450 by a predetermined number of pixels (Pb pixels) in the periodic direction, similar to how it generates a phase-changed periodic image 451 from a periodic image 450. The phase of the periodic image 1451 shown in Figure 5(a) is phase P2 (=360 degrees / Wc pixels × Pb pixels). Both the periodic image 1450 and the phase-changed periodic image 1451 are images with diagonal linear stripes, but the position of the stripes is shifted by Pb pixels in a diagonal direction (periodic direction) perpendicular to the stripes.

[0052] For example, the casting defect detection device 1 generates multiple types of periodic images 1451… with a phase shift of a predetermined number of pixels (Pb pixels) within a range of one wavelength (Wc pixel), that is, within a phase range of 0 to 360 degrees. For example, the casting defect detection device 1 generates multiple types of periodic images 1451… with a phase shift for each of the multiple types of periodic images 1450… with a changed angle.

[0053] As shown in Figure 5(b), the casting defect detection device 1 generates a periodic image 4501 by changing the period of pixel value changes in the periodic image 450. Both the periodic image 450 and the periodic image 4501 with the changed period are images with vertical stripes throughout, but the periodic image 4501 is an image in which the periodic Wc pixels of the periodic image 450 are changed to periodic Wd pixels. In other words, the periodic image 4501 is an image in which the width of one wavelength of the sine wave 460 shown in Figure 3 is changed from Wc pixels to Wd pixels.

[0054] For example, the casting defect detection device 1 generates multiple types of periodic images 4501… within a predetermined pixel range, for each pixel, by changing the period, i.e., the width of one wavelength of the sine wave 460. Furthermore, the casting defect detection device 1 generates multiple types of periodic images with changed phases and angles, as explained in Figure 5(a), for each periodic image with a changed period.

[0055] Although not shown in Figure 5, the casting defect detection device 1 generates periodic images with altered amplitudes for each periodic image generated by changing the phase, angle, and period, as explained in Figures 5(a) and (b).

[0056] Specifically, for example, a 16-bit grayscale X-ray image 250 with 2000 pixels vertically and horizontally (Wa=2000 pixels, Wb=2000 pixels) is used to generate a periodic image 450, also in 16-bit grayscale.

[0057] For example, the periodic image 450 shown in Figure 3 is an image of light and dark vertical stripes representing a sine wave 460 with an amplitude of 20,000 pixels, where the pixel value changes between +10,000 and -10,000 (V=10,000) around a predetermined pixel value, with wavelength 200 pixels (Wc=200 pixels).

[0058] The casting defect detection device 1 generates 10 periodic images 451… by shifting the phase of a periodic image 450, which shows a light and dark stripe pattern with a period of 200 pixels, by 20 pixels (Pa=20 pixels) at a time. The casting defect detection device 1 also generates three types of periodic images 1450… by changing the angle in the periodic direction of the periodic image 450 to 45 degrees, 90 degrees, and 135 degrees (R=45 degrees, 90 degrees, 135 degrees). From each of the angle-changed periodic images 1450…, the casting defect detection device 1 generates 10 periodic images 1451… by shifting the phase of each image by 20 pixels (Pb=20 pixels) in the periodic direction.

[0059] The casting defect detection device 1 generates one or more periodic images 4501… by changing the period of the light and dark stripe pattern that forms the periodic image 450, i.e., the width of one wavelength of the sine wave 460 (Wc pixels). For each periodic image 4501…, three periodic images are generated by changing the angle by 45 degrees, 90 degrees, and 135 degrees. In addition, ten periodic images are generated by shifting each periodic image with a different angle by 20 pixels in the periodic direction to change the phase.

[0060] The casting defect detection device 1 generates each of the above-described periodic images as images with an amplitude of pixel values ​​from +10000 to -10000, centered on the average pixel value of all pixels forming the X-ray image 250. The casting defect detection device 1 also generates periodic images with modified amplitudes from each periodic image. For example, multiple types of amplitudes, such as +10000 to -10000 pixel values ​​and +5000 to -5000 pixel values, can be adopted to generate periodic images for each amplitude. Alternatively, for example, the minimum value may be set to 5000, the maximum value to 10000, and the step size to 1000, and multiple types of amplitudes, such as +5000 to -5000 and +6000 to -6000, can be set to generate periodic images for each amplitude. The order of generation of periodic images is not limited. For example, periodic images with modified amplitudes may be generated from periodic images with modified phase, angle, and period, or periodic images with modified phase, angle, and period may be generated from periodic images with modified amplitudes first.

[0061] In this way, the casting defect detection device 1 uses a periodic image 450 with a light and dark vertical stripe pattern and multiple types of periodic images in which at least one of the phase, angle, period, or amplitude differs from the periodic image 450, to detect casting defects on a detection image generated by adding each periodic image to the X-ray image 250.

[0062] For example, multiple types of periodic images can be generated in advance and stored in the memory unit of the casting defect detection device 1, and the casting defect detection device 1 can then perform the respective processing for each periodic image. The multiple types of periodic images may be generated by the casting defect detection device 1 or by another device.

[0063] In terms of the specific processing flow, as shown in Figure 2, the casting defect detection device 1 uses periodic images 450, 451… with different phases at an angle of 0 degrees (R=0 degrees in Figure 5), and while changing the phase (step S7), that is, changing to periodic images 451… with different phases, it adds them to the X-ray image 250 to generate a detection image (step S4), and then repeatedly performs the process of detecting casting defects from the detection image (step S5).

[0064] When the defect detection process is completed using all periodic images 450, 451... with different phases at an angle of 0 degrees (Step S6; Yes), the defect detection device 1 determines whether or not the defect detection process has been completed using all periodic images 1450... with changed angles (Step S8).

[0065] While there are unused periodic images 1450... (step S8; No), the casting defect detection device 1 changes the angle (step S9), that is, uses the angle-changed periodic images 1450... and continues processing. For example, the casting defect detection device 1 targets each of the multiple periodic images 1450... with angles changed to 45 degrees, 90 degrees, and 135 degrees, and further changes the phase, that is, changes the periodic images 1451... (step S7), adds them to the X-ray image 250 to generate a detection image (step S4), and repeatedly performs the process of detecting casting defects from the detection image (step S5).

[0066] After completing the process using the periodic images 1450, 1451… whose angle and phase have been changed (Step S8; Yes), the casting defect detection device 1 determines whether or not the casting defect detection process has been completed using all the periodic images 4501… whose periodicity of the light and dark stripe pattern, i.e., the wavelength of the sine wave 460, has been changed (Step S10).

[0067] While there are unused periodic images 4501... (step S10; No), the casting defect detection device 1 changes the period (step S11), that is, continues processing using the periodic images 4501... whose periods have been changed. The casting defect detection device 1 adds the periodic images 4501... whose periods have been changed and the periodic images 4501... whose phases and angles have been changed as described above, to the X-ray image 250 to generate a detection image (step S4), and repeatedly performs the process of detecting casting defects from the detection image (step S5).

[0068] After completing the process using the periodic images 4501... whose angle, phase, and period have been changed (step S10; Yes), the casting defect detection device 1 determines whether or not the casting defect detection process has been completed using all the periodic images in which the amplitude of the light and dark stripe pattern, i.e., the amplitude of the sine wave 460, has been changed (step S12).

[0069] While there are unused periodic images (step S12; No), the casting defect detection device 1 changes the amplitude (step S13), that is, continues processing using the periodic images with changed amplitude. The casting defect detection device 1 adds the periodic images with changed amplitude and the periodic images with changed phase, angle, and period of each periodic image as described above to the X-ray image 250 to generate a detection image (step S4), and repeatedly performs the process of detecting casting defects from the detection image (step S5).

[0070] Thus, for example, the detection results for casting defects can be obtained by adding multiple periodic images with different phases at angles (R) of 0 degrees, multiple periodic images with different phases at angles of 45 degrees, multiple periodic images with different phases at angles of 90 degrees, and multiple periodic images with different phases at angles of 135 degrees, as shown in Figure 5, to the X-ray image 250. Similarly, for periodic image 4501 with a changed period (wavelength), the detection results for casting defects can be obtained by adding multiple periodic images with different phases at angles of 0 degrees, multiple periodic images with different phases at angles of 45 degrees, multiple periodic images with different phases at angles of 90 degrees, and multiple periodic images with different phases at angles of 135 degrees, to the X-ray image 250. Furthermore, for periodic images with changed amplitudes, the detection results for casting defects can be obtained by adding each periodic image to the X-ray image 250.

[0071] The casting defect detection device 1 detects casting defects using each periodic image and stores the obtained casting defect detection results in the storage unit. The casting defect detection device 1 performs post-processing of multiple detection results (step S14), then outputs the final casting defect detection result (step S15), and terminates the process. By performing post-processing, the output of a detection result includes the casting defects detected in the process of step S3 and the casting defects detected in each detection image by repeatedly executing the processes of steps S4 and S5.

[0072] Figure 6 is a diagram illustrating the post-processing of the detection results of casting defects obtained using each periodic image. As shown in Figure 6(a), in the black and white binary image 700, where the detection results of casting defects are shown as white pixels against a black pixel background, areas other than the casting defects may appear as white pixels. For example, when detecting casting defects on a detection image generated by adding the periodic image 450 of light and dark vertical stripes shown in Figure 3 to the X-ray image 250, in the resulting image 700, areas where the edge portion of the casting 100 is in the same vertical direction as the stripes of the periodic image 450 tend to appear as white pixels.

[0073] In step S14 of the post-processing shown in Figure 2, the casting defect detection device 1 performs a process to exclude areas that do not meet the pre-set conditions for casting defects from the areas formed by white pixels included in the image 700. For example, if the area of ​​casting defects appearing in the casting defect detection image is set to the range of Sa to Sb, the casting defect detection device 1 determines that an area is not a casting defect if the area of ​​an area formed by multiple white pixels is less than the threshold Sa or exceeds Sb, and changes the white pixels in that area to black pixels. As a result, for example, the image 700 shown in Figure 6(a) changes to the detection image 710 shown in Figure 6(b).

[0074] The casting defect detection device 1 performs post-processing to remove areas other than casting defects from the image 700 obtained by performing casting defect detection processing on a detection image generated using a periodic image, and adopts the resulting detection image 710 as the casting defect detection result obtained from the periodic image.

[0075] The casting defect detection device 1 uses multiple periodic images to detect casting defects and performs post-processing on the image in which defects have been detected, changing white pixels other than those containing defects into black pixels. In this way, multiple detection images are obtained, such as the detection image 710 shown in Figure 6(b) and the detection image 720 shown in Figure 6(c).

[0076] In step S14 of the post-processing shown in Figure 2, the casting defect detection device 1 performs a process of combining multiple detection images 710 and 720 to generate an image showing the final casting defect detection result using periodic images.

[0077] For example, when the comparative light blending process is performed on the detection image 710 shown in Figure 6(b) and the detection image 720 shown in Figure 6(c), the image 610 shown in Figure 6(d) is obtained. The casting defect detection device 1 adopts the image 610 obtained by performing the comparative light blending process as the result of casting defect detection using the X-ray image 250. Although only the two detection images 710 and 720 are shown in Figure 6, in reality, the comparative light blending process is performed on all detection images obtained using each of the multiple types of periodic images, and an image showing the final detection result of casting defects obtained using periodic images is obtained. The image blending method is not limited to comparative light blending; other blending methods such as additive light blending and multiplicative blending may be used as long as the white pixels representing casting defects in each detection image can be displayed together in a single image.

[0078] By using the periodic image 450 with an angle of 0 degrees shown in Figure 5, it is possible to obtain a detection image in which casting defects are detected while suppressing the effect of pixel value fluctuations in areas where the pixel value gradually increases and decreases laterally on the X-ray image due to the shape of the casting 100, etc. By using the periodic image 451 with a changed phase, it is possible to obtain a detection image in which casting defects are detected while suppressing the effect of pixel value fluctuations in areas where the position does not match that of the periodic image 450 and the effect of pixel value fluctuations could not be suppressed. By using the periodic image 1450 with a changed angle, it is possible to obtain a detection image in which casting defects are detected while suppressing the effect of pixel value fluctuations in areas where the pixel value gradually increases and decreases in different directions on the X-ray image. By using the image 4501 with a changed period, it is possible to obtain a detection image in which casting defects are detected while suppressing the effect of pixel value fluctuations in areas where the width of the area on the X-ray image where the pixel value fluctuates does not match that of the periodic image 450 and the effect of pixel value fluctuations could not be suppressed. By using periodic images with altered amplitude, it is possible to obtain a detection image in which casting defects are detected. This image suppresses the effects of pixel value fluctuations in areas of the X-ray image where pixel values ​​fluctuate, in areas where the periodic image could not cancel out the fluctuations. Image 610, which combines the casting defect detection results from these multiple detection images, is considered the final casting defect detection result using periodic images.

[0079] Thus, the casting defect detection device 1 can output a casting defect detection result that includes both casting defects detected based on the brightness difference of the X-ray image and casting defects detected by generating a casting defect detection image from the X-ray image using a periodic image. However, the method of outputting the casting defect detection result is not particularly limited, and for example, the casting defect detection result using the X-ray image and the casting defect detection result using the periodic image may be output separately. Alternatively, the casting defect detection result using periodic images may be output separately for each type of periodic image, for example, casting defect detection results for periodic images with the same angle and casting defect detection results for periodic images with the same period.

[0080] In this embodiment, an example of using multiple types of periodic images with modified phase, angle, period, and amplitude has been described. However, it is also possible to use multiple types of periodic images with modified phase, angle, period, and amplitude. The numerical values ​​for phase, angle, period, and amplitude described above are examples only and do not limit the periodic images to be used. The periodic images to be used should be set according to the shape of the casting 100 to be detected for casting defects.

[0081] In this embodiment, an example was described in which multiple types of periodic images are prepared in advance, and each periodic image is added to an X-ray image obtained by imaging a casting to generate an image for detecting casting defects, and casting defects are detected based on the brightness difference that appears in the image. However, the periodic images may be generated at the time the process of adding them to the X-ray image is executed. The process of adding periodic images to the X-ray image may be executed as image processing, or it may be executed as a calculation process in which a periodically changed pixel value is added to the pixel value of each pixel forming the X-ray image.

[0082] The images shown in this embodiment are schematic images provided to illustrate the processes performed by the casting defect detection device 1, and do not limit the content of the images. Similarly, the pixel value curves and waveforms shown to illustrate the pixel values ​​(luminance values) and changes in pixel values ​​(luminance gradients) are illustrative examples, and do not limit the pixel values ​​or changes in pixel values ​​to these. For example, the waveform showing the change in pixel values ​​in the periodic images shown in Figures 3 to 5 is not limited to a sine wave 460, but may be a triangular wave or a waveform in which Gaussian waveforms are periodically repeated, as long as the pixel values ​​change periodically.

[0083] The configuration of the casting defect detection device 1 shown in this embodiment is functionally schematic, and the configuration of the casting defect detection device 1 is not physically limited to this configuration. The form of distribution and integration of the device is not limited to the example described above, and all or part of it can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions. For example, the casting defect detection device 1 may include an acquisition unit that acquires an X-ray image of the casting and a detection unit that detects casting defects based on a detection image generated from the X-ray image using periodic images as described above. The detection unit may be divided into a periodic image generation unit that generates periodic images, a detection image generation unit that generates a detection image by adding periodic images to the X-ray image, a detection processing unit that performs casting defect detection processing using the detection image, a processing unit that performs pre-processing and post-processing, etc. The casting defect detection device 1 may also be configured as a plurality of devices, such as a terminal and a server device.

[0084] While embodiments of the casting defect detection method and casting defect detection apparatus according to this disclosure have been described above with reference to the drawings, the configuration and operation of the casting defect detection apparatus 1 are not limited to the above embodiments, and may be implemented in various forms with improvements, changes, and modifications based on the knowledge of those skilled in the art, without departing from the spirit of the invention. [Industrial applicability]

[0085] As described above, the casting defect detection method and casting defect detection apparatus according to this disclosure are useful for detecting casting defects with high accuracy using X-ray images of castings. [Explanation of Symbols]

[0086] 1. Casting defect detection device 10 X-ray irradiation device 20 Imaging device

Claims

1. A method for detecting casting defects using X-ray images of a casting, The acquisition process involves obtaining X-ray images of the castings, A generation step of generating an image for detecting casting defects by adding a periodic image of a striped pattern in which the pixel values ​​change periodically in a predetermined direction to the X-ray image, A detection step for detecting casting defects based on the aforementioned detection image, A method for detecting casting defects, characterized by including the following:

2. The casting defect detection method according to claim 1, characterized in that the change in the pixel value in one direction in the periodic image exhibits a sine wave shape or a triangular wave shape.

3. The generation step includes a step of generating a plurality of detection images by adding to the X-ray image a plurality of periodic images, each of which differs in at least one of the following: phase in one direction, angle indicating one direction, period of pixel value change in one direction, and amplitude of pixel value change in one direction. The detection step includes a step of detecting casting defects in each of the multiple detection images generated in the generation step. The casting defect detection method according to feature 1.

4. The casting defect detection method according to claim 1, characterized in that the detection step includes a step of detecting casting defects based on the pixel value of each pixel forming the X-ray image.

5. Output step: In the above detection step, images showing the detection results of casting defects obtained from each of the multiple detection images are combined in a predetermined manner and output as the casting defect detection result. The casting defect detection method according to claim 3 or 4, further comprising:

6. The process involves determining whether a region is a casting defect or not based on the area of ​​the region detected as a casting defect from the aforementioned detection image, and excluding regions that are determined not to be casting defects from the casting defect detection results. The casting defect detection method according to claim 1, further comprising:

7. A casting defect detection device that detects casting defects using X-ray images of castings, An acquisition unit for acquiring X-ray images of castings, A detection unit generates a casting defect detection image by adding a periodic image of a striped pattern in which the pixel values ​​change periodically in a predetermined direction to the X-ray image, and detects casting defects based on the detection image. A casting defect detection device characterized by comprising the following features.

Images