JP2022092083A - Product defect detection method - Google Patents

Abstract

To provide a defect detection method which can detect a blow hole of a casting easily.SOLUTION: A defect detection method comprises the steps of: acquiring a cross section image obtained by cutting in round slices a prescribed number of same-shaped castings at the same position; calculating the average luminance obtained by averaging the luminance of each corresponding image area of the cross section images and generating a reference cross section image made of the respective image areas applied with the average luminance; acquiring a detection cross section image obtained by cutting in round slices a casting having the same shape as the casting becoming the defect detection object at the same position; calculating the difference luminance taking a difference in the luminance between each image area of the detection cross section image and each image area of the reference cross section image equivalent to each image area and generating a cross section image for defect detection made of the image areas applied with the difference luminance; and specifying a blow hole from the cross section image for defect detection.SELECTED DRAWING: Figure 8

Description

The present invention relates to a method for detecting defects in a product, and more particularly to a method suitable for detecting cavities, which are internal defects occurring in a cast product.

Attempts have been made to use CT (computed tomography), which can obtain sliced cross-sectional images to detect defects in products, for example, cavities that occur in castings (Patent Document 1), but CT images have a peculiarity of artifacts. Since noise is generated, the deterioration of defect detection accuracy due to this is a problem. Therefore, for example, Patent Document 2 shows a method for removing a metal artifact generated in a medical CT image. Here, a circular detection region centered on a pixel point of the metal region is set, and the inside of the circular detection region is set. When the ratio of the metal region to the non-metal region is equal to or higher than the preset value, the metal artifact is removed by executing the boundary smoothing treatment with the circular detection region as the boundary region.

JP-A-2005-351875 JP 2015-85198

However, since the above-mentioned conventional artifact removing method requires complicated calculations, there is a problem that defect detection of a product cannot be easily performed.

Therefore, the present invention solves such a problem, and an object of the present invention is to provide a defect detection method capable of easily detecting defects in a product by reducing the computational burden of removing artifacts from CT images. And.

In order to achieve the above object, in the first invention, a cross-sectional image obtained by cutting a predetermined number of products of the same shape into round slices at the same position is obtained, and the average brightness of these cross-sectional images obtained by averaging the brightness of each corresponding image region is obtained. A reference cross-sectional image consisting of the image regions to which the average brightness is applied is generated by calculation, while a detected cross-sectional image obtained by cutting a product having the same shape as the product to be a defect detection target at the same position is obtained. Defect detection consisting of an image region to which the differential brightness is imparted by calculating the differential brightness obtained by taking the difference between the brightness of each image area of the detected cross-sectional image and the brightness of each image area of the reference cross-sectional image corresponding to the image region. It is characterized in that a defect is generated and a defect is identified from the defect detection sectional image. The "image area" may be a one-pixel area or a predetermined plurality of pixel area.

In the first invention, in the reference cross-sectional image, the image region is provided with the average brightness obtained by averaging the brightness of each corresponding image region of the cross-sectional image obtained by cutting a predetermined number of products of the same shape at the same position. Even if there is a local defect inside each product and the brightness changes significantly in this part, the change in brightness due to the defect is diluted by averaging and becomes sufficiently small, and a reference cross-sectional image without a defect image part can be obtained. .. At this time, if the installation position and orientation of the same-shaped product are the same, the artifacts on each of the cross-sectional images always appear at the same position with the same brightness. Therefore, even if the brightness is averaged as described above, the artifacts are displayed on the reference cross-sectional image. It remains as it is. On the other hand, on the detected cross-sectional image, the defect image portion and the same artifact as on the reference cross-sectional image appear superimposed on the defect image portion. Therefore, a defect consisting of an image region to which the difference brightness is given is calculated by calculating the difference brightness obtained by taking the difference between the brightness of each image area of the detected cross-sectional image and the brightness of each image area of the reference cross-section image corresponding to the image area. When the cross-sectional image for detection is generated, the artifact is removed in the cross-sectional image for defect detection and only the image of the defect remains, so that the defect can be easily and surely identified. In this way, according to the first aspect of the present invention, it is possible to easily detect defects in a product by a simple operation of only a luminance averaging operation and a luminance difference calculation without requiring a complicated calculation.

In the second invention, a required number of defect detection cross-sectional images are generated at different round slice positions of the product, a defect detection three-dimensional image is generated from these defect detection cross-sectional images, and a defect detection three-dimensional image is generated from the defect detection three-dimensional image. Identify defects.

According to the second invention, since the three-dimensional image for defect detection is generated, the size (volume) and the existence position of the defect can be detected more accurately, and the defect can be identified more reliably.

In the third invention, obvious defects are removed in advance from each of the cross-sectional images acquired in the predetermined number.

According to the third aspect of the present invention, since obvious defects are removed in advance from each cross-sectional image by automatic or visual inspection or the like, the efficiency of identifying the remaining defects is improved.

In the fourth invention, the product is a casting and the defect is a cavity.

As described above, according to the defect detection method of the present invention, the calculation burden of removing artifacts from the CT image is reduced, and the defect detection of the product can be easily performed.

It is a schematic block diagram of the apparatus for carrying out the method of this invention. It is a flowchart which shows the procedure for generating a reference cross-sectional image. It is a figure which shows an example of the cross-sectional image at three positions in the Z direction of a casting. It is a figure which shows an example of the cross-sectional image at three positions in the Z direction of the CAD three-dimensional image of a cast product. It is a flowchart which shows the detailed procedure of position correction. It is a figure which shows the corner of a CAD cross-sectional image, and the corner center of gravity of a CT cross-sectional image. It is a flowchart which shows the detailed procedure of the average luminance calculation. It is a flowchart which shows the procedure for generating the section image for defect detection. It is a flowchart which shows the detailed procedure of the difference luminance calculation. It is a flowchart which shows the detailed procedure of edge noise removal. It is a flowchart which shows the detailed procedure of cavities identification.

The embodiments described below are merely examples, and various design improvements made by those skilled in the art within the scope of the present invention are also included in the scope of the present invention.

FIG. 1 shows an example of an apparatus for carrying out the method of the present invention. As a product, for example, a casting W is placed on a table (not shown) that rotates in an XY plane. The X-ray irradiator 1 and the line sensor 2 constituting the CT (computed tomography) are located in a plane parallel to the XY plane so as to face each other across the cast product W, from the X-ray irradiator 1. The X-ray R, in which the beam is narrowed flat by a collimator or the like and spreads in a fan shape, passes through the cast product W in a slice shape and is input to the line sensor 2.

The output from the line sensor 2 is input to the computer 3 in the control device as digital data. By rotating the casting W on a table, transmitted X-rays from all directions are input to the line sensor 2, and digital data from the line sensor 2 is input by a method known to the computer 3 at a predetermined position in the Z direction. A sliced cross-sectional image of the cast product W of the above is obtained.

In the present embodiment, the table on which the casting W is placed, the X-ray irradiator 1 and the line sensor 2 are relatively movable in the Z direction, and are obtained at a plurality of positions in the Z direction by calculation in the computer 3. A three-dimensional image of the cast product W can be obtained from the cross-sectional image. Rotation of the X-ray irradiator 1 and the table, relative movement in the Z direction, and the like are appropriately performed by the output signal from the computer 3. Each procedure described in the flowchart below is executed by the program of the computer 3.

A. Generation of reference cross-section image (acquisition of cross-section image)
FIG. 2 shows a procedure for generating a reference cross-sectional image. First, in step 101, a cross-sectional image of the cast product at a predetermined position in the Z direction is acquired. In the present embodiment, cross-sectional images of a plurality of positions in the Z direction are acquired for one cast product. This is shown in FIG. 3, and F1, F2, and F3 in FIG. 3 are examples of cross-sectional images of the cast product W obtained at the three positions Z1, Z2, and Z3 in the Z direction, respectively. In the present embodiment, after acquiring cross-sectional images of one cast product at a plurality of positions in the Z direction (predetermined number required to obtain a three-dimensional image described later), the required number of other cast products having the same shape are also obtained. Cross-sectional images are acquired at the same position in the Z direction.

(Position correction of cross-sectional image)
Subsequently, in step 102 of FIG. 2, the positions of the cross-sectional images F1 to F3 are corrected. This is because the position of the casting placed on the table is slightly displaced each time. In this embodiment, this deviation does not occur in the Z direction because the cast product is placed on the table, but occurs only in the XY plane. If the misalignment when the cast product is placed is small enough that there is no problem, there is no need to correct the position.

In the present embodiment, the position correction of the cross-sectional image is performed by using the CAD three-dimensional image of the cast product prepared separately. FIG. 4 shows an example of the CAD three-dimensional image W', and F1', F2', and F3'in the figure indicate the CAD three-dimensional image W'at the same Z-direction position Z1 as the above three positions Z1, Z2, and Z3. It is a CAD cross-sectional image cut into round slices by', Z2', Z3'.

The details of the position correction step 102 are shown in FIG. In FIG. 5, in step 201, CT cross-sectional image data (corresponding to the cross-sectional image) and CAD cross-sectional image data corresponding to the CT cross-sectional image data (corresponding to the CAD cross-sectional image) are cut out. In step 202, the CAD cross-sectional image data is scanned to detect the corners at the four corners. The corners C1', C2', C3', and C4' detected in the CAD cross-sectional image F1'are shown in FIG. 6 (1).

In step 203 of FIG. 5, the vertical and horizontal differential values of the CT cross-sectional image data corresponding to the periphery of the detected corners C1'to C4' are calculated, and the center of gravity whose differential value is a certain amount or more is calculated (step 204). ). FIG. 6 (2) shows the calculated center of gravity C1, C2, C3, and C4 on the cross-sectional image F1.

In step 205 of FIG. 5, the amount of translation and the amount of rotation of the CT cross-sectional image data with respect to the CAD cross-sectional image data are calculated from the relative positional relationship between the corners C1'to C4' and the center of gravity C1 to C4. This calculation uses, for example, the relative positional relationship between the corners C1'to C4'and the centers of gravity C1 to C4 detected or calculated at the multiple positions Z1 to Z3 in the Z direction of the cast product, for example, and the minimum squared error is minimized. This is done by obtaining an appropriate translation amount and rotation angle. In step 206, the CT cross-sectional image data is moved by the calculated translation amount and rotation angle to correct the image positions of the cross-sectional images F1 to F3.

(Calculation of average brightness, generation of reference cross-sectional image)
In step 103 of FIG. 2, the average brightness of each image region (pixel in the present embodiment) is calculated for the position-corrected cross-sectional images F1 to F3. This calculation of the average brightness is to calculate the average brightness of each image region of each cross-sectional image sliced at the same position in the Z direction for the required number of cast products W having the same shape. By calculating the average in this way, even if there is a cavity as a local defect inside the casting and the brightness changes significantly in this part, this averaging dilutes the change in brightness due to the cavity. Is small enough.

On the other hand, since the artifacts of each cross-sectional image always appear at the same position with the same brightness, they remain as they are even if the average of the above brightness is calculated. Therefore, when the brightness of each image region is averaged and a reference cross-section image consisting of the image regions to which such average brightness is given is generated (step 104), the reference cross-section image is a cross-section of the casting product in which the cavities are eliminated. The artifact is superimposed on the image. In the present embodiment, such a reference cross-sectional image is generated at a plurality of positions (the number required to obtain a three-dimensional image described later) in the Z direction for the cast product.

The details of the average luminance calculation step 103 (FIG. 2) are shown in FIG. In step 301 of FIG. 7, the CT value (corresponding to the luminance value) of the reference cross-sectional image data is reset to 0. Subsequently, the CT value of the CT image data is added to the reference cross-sectional image data (step 302), this addition is repeated for the required number of CT cross-sectional image data (step 303), and all the pixels of the reference cross-sectional image data are repeated in step 304. The CT value of is divided by the number of CT cross-sectional image data (necessary number).

The calculation of the average luminance is not necessarily limited to the arithmetic mean as described above, and may be a weighted average or the like, if necessary.

B. Generation of cross-sectional image for defect detection (acquisition of detected cross-sectional image)
FIG. 8 shows a procedure for generating a cross-sectional image for defect detection. First, the detected cross-sectional image is acquired in step 401. The detected cross-sectional image is a round-cut cross-sectional image at a predetermined position in the Z direction obtained by the above-mentioned CT in which a cast product to be detected for defects is placed on a table. The casting that is the target of defect detection has the same shape as the casting that generated the reference cross-sectional image.

In step 402 of FIG. 8, the position of the detected cross-sectional image is corrected by the same procedure as described above (see step 102 and steps 201 to 206).

(Calculation of differential brightness)
Subsequently, in step 403 of FIG. 8, the difference luminance between the detected cross-sectional image and the reference cross-sectional image is calculated. As the reference cross-sectional image in this case, the one obtained by slicing the cast product at the same position in the Z direction as the detected cross-sectional image is used. Here, the detected cross-sectional image is an image of the artifact and the cavities (if there is a cavities) superimposed on the cross-sectional image of the casting. On the other hand, in the reference cross-sectional image, as described above, the artifact is superimposed on the cross-sectional image portion of the cast product. Therefore, if the difference brightness between the detected cross-sectional image and the reference cross-sectional image is calculated, the artifact and the cross-sectional image portion of the cast product are removed, and a cross-sectional image for defect detection is obtained in which only the image portion of the cavities remains. Become.

The details of the differential luminance calculation step 403 are shown in FIG. In step 501 of FIG. 9, the CT value of the reference cross-sectional image data is subtracted from the CT value of the CT cross-sectional image data (corresponding to the detected cross-sectional image), and if the subtraction result is a positive value, this is set to 0 (step 502). ), If the subtraction result is a negative value, multiply by -1 to obtain a positive value (step 503). If there is a cavity, the brightness of that part will decrease, so the subtraction value will be a large negative value. Therefore, this is converted to a positive value.

(Removal of low difference brightness area)
In step 404 of FIG. 8, a portion of the defect detection cross-sectional image obtained in step 403 whose brightness is smaller than the threshold value is removed as noise.

(Removal of edge noise, generation of cross-sectional image for defect detection)
In step 405 of FIG. 8, a high-intensity portion in the defect detection cross-sectional image generated along the contour of the casting is removed as edge noise to generate a defect detection cross-sectional image (step 406). FIG. 10 shows the details of the edge noise removal step 405. In FIG. 10, the boundary distance is set in step 601 and in step 602, the difference data that is not zero within the region below the boundary distance from the contour portion of the casting that can be recognized by the CAD cross-sectional image data corresponding to the defect detection cross-sectional image data. If there is (difference brightness value), set this to 0.

(Generation of 3D image for defect detection, identification of cavities)
In step 407 of FIG. 8, a required number of the defect detection cross-sectional images are acquired at different round slice positions in the Z direction of the cast product, and a defect detection three-dimensional image is generated. In this three-dimensional image for defect detection, only the cavities (candidates) are arranged in the three-dimensional space. Therefore, in the following step 408, the cavities are specified from the cavities candidates in the defect detection three-dimensional image.

The details of the cavities specifying step 408 are shown in FIG. In step 701 of FIG. 11, the cavities determination criteria (volume, length, existence position) are set. In step 702, a casting cavity candidate which is a continuous non-zero region in the defect detection three-dimensional image data is detected. In the following step 703, the volume, length, and existing position of the detected cavities candidate are calculated, and if the cavities determination criteria are satisfied, the cavities candidates are identified as cavities (steps 704 to 705). This procedure is performed for all cavities candidates (step 706).

(Other embodiments)
In the above embodiment, the cavities are specified by the defect detection three-dimensional image, but it may be specified by the two-dimensional defect detection cross-sectional image.
In the above embodiment, if obvious defects in each cross-sectional image are removed in advance by visual inspection or the like, the remaining defects can be efficiently identified.
In the above embodiment, the cavities of the cast product are detected, but the present invention can be widely applied to the detection of defects in the product.

1 ... X-ray irradiator, 2 ... line sensor, 3 ... computer, R ... X-ray, W ... casting (product).

Claims (4)

Cross-sectional images obtained by cutting a predetermined number of products of the same shape at the same position are obtained, and the average brightness obtained by averaging the brightness of each corresponding image area of these cross-sectional images is calculated, and the average brightness is given to each of the image areas. A reference cross-sectional image is generated, and on the other hand, a detected cross-sectional image obtained by cutting a product having the same shape as the product to be detected as a defect at the same position is obtained, and the detected cross-sectional image is divided into each image area and each image area. The differential brightness obtained by taking the difference in the brightness of each image region of the corresponding reference cross-sectional image is calculated to generate a defect detection cross-sectional image consisting of the image region to which the differential brightness is applied, and the defect detection cross-sectional image is generated. A product defect detection method characterized by identifying defects from. A claim that a required number of defect detection cross-sectional images are generated at different round slice positions of the product, a defect detection three-dimensional image is generated from these defect detection cross-sectional images, and a defect is identified from the defect detection three-dimensional image. The product defect detection method described in 1. The defect detection method for a product according to claim 1 or 2, wherein obvious defects are removed in advance from each of the cross-sectional images to be acquired in a predetermined number. The method for detecting a defect in a product according to any one of claims 1 to 3, wherein the product is a casting and the defect is a cavity.

Images