Inspection device

By combining image generation and real-time image processing with machine learning, the problem of excessively long inspection times in existing technologies has been solved, achieving high-speed and high-precision quality judgment, which is suitable for inspection devices of packaged food.

CN114467023BActive Publication Date: 2025-11-07ISHIDA CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202080066160.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-09-24
Filing Date
2020-07-14
Publication Date
2025-11-07
Estimated Expiration
2040-07-14

AI Technical Summary

Technical Problem

In existing technologies, the inspection devices require a long time to judge the quality of packaged food and other inspection objects, which leads to a bottleneck problem in the production line.

Method used

The image generation unit generates a one-dimensional transmission signal of the object to be inspected, which is then unfolded into a two-dimensional image. The image processing unit processes a portion of the image in real time, and combined with a machine learning program, it enables high-speed quality judgment of the object to be inspected.

Benefits of technology

It shortens the time from when the inspected item passes through the irradiation line to when the quality inspection ends, achieving high-speed and high-precision quality judgment and preventing the inspection device from becoming a bottleneck in the production line.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114467023B_ABST
    Figure CN114467023B_ABST
Patent Text Reader

Abstract

An inspection apparatus includes an image generation section that expands a one-dimensional transmission signal of an object to be inspected that passes through an irradiation line of electromagnetic waves into a two-dimensional image on a memory, an image processing section that performs image processing on a partial image including a portion of the object to be inspected each time the partial image is generated in the image generation section, and an inspection section that inspects a quality of the partial image after the image processing based on a result of one or more processes of the image processing section.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present disclosure relates to an inspection device. BACKGROUND

[0002] As one of the inspection devices, a device that inspects an inspection target (an object to be inspected) such as a packaged food can be exemplified. For example, in an inspection device described in Patent Literature 1, a manner is disclosed in which an entire image of an inspection target is divided into a plurality of first partial images, and the entire image is divided into a plurality of second partial images different from the plurality of first partial images. In the inspection device, first, presence or absence of an abnormality is judged for the plurality of first partial images, and presence or absence of an abnormality is judged for the plurality of second partial images. Next, based on overlapping of the first partial image judged to include an abnormality and the second partial image judged to include an abnormality, the inspection device judges presence or absence of an abnormality with respect to the inspection target.

[0003] PRIOR ART DOCUMENT

[0004] PATENT LITERATURE

[0005] Patent Literature 1: Japanese Patent No. 6387477 SUMMARY

[0006] PROBLEMS TO BE SOLVED BY THE INVENTION

[0007] In the Patent Literature 1, after the entire image acquired is divided into the plurality of first partial images and the plurality of second partial images, respectively, the quality of the inspection target is determined. In such a method, there is a tendency that the time required to inspect the quality of one inspection target becomes long.

[0008] However, the inspection device that inspects the object to be inspected such as a packaged food is sometimes provided on a production line for streamlined measurement. For example, in a case where the method of the Patent Literature 1 is adopted for the inspection device provided on the production line, the inspection device becomes a bottleneck of the factory line.

[0009] An aspect of the present disclosure is to provide an inspection device capable of determining the quality of an object to be inspected at high speed.

[0010] MEANS FOR SOLVING THE PROBLEMS

[0011] The inspection device according to an aspect of the present disclosure includes an image generation unit that expands a one-dimensional transmission signal of an object to be inspected that passes through an irradiation line of electromagnetic waves as a two-dimensional image on a memory, an image processing unit that performs image processing with respect to a partial image including a part of the object to be inspected each time the partial image is generated in the image generation unit, and an inspection unit that inspects the quality of the partial image after the image processing based on a result of one or a plurality of processes of the image processing unit.

[0012] According to the inspection apparatus, when the image processing section generates a partial image indicating a portion of the object to be inspected using the image generating section, the image processing section performs image processing on the partial image as a target, and when the image processing section generates a next partial image adjacent to the partial image, the image processing section performs image processing on the next partial image. Thus, image processing is repeatedly performed on the newly generated partial image each time the image generating section generates a partial image. In this case, during the execution of the generation of the partial image by the image generating section, the image processing section can perform image processing on the already generated partial image. Thus, the time from when the object to be inspected is irradiated with the radiation line until the inspection of the quality of the object to be inspected ends can be shortened. Therefore, by using the inspection apparatus, the quality of the object to be inspected can be determined at high speed.

[0013] The image processing section can also perform image processing on at least a portion of the partial image subjected to the previous image processing, for example, a boundary region of the previous partial image that borders on the partial image that is the target of the current processing, and the partial image subjected to the current image processing. In this case, the inspection section can well inspect the boundary between the partial image subjected to the previous image processing and the partial image subjected to the current image processing. Thus, the quality of the object to be inspected can be determined with high accuracy.

[0014] The inspection section can also perform inspection on the partial image subjected to the image processing in parallel with the image processing by the image processing section. In this case, during the image processing by the image processing section on the already generated partial image, the inspection section can inspect the quality of the partial image that has been subjected to the image processing. Thus, the time from when the object to be inspected is irradiated with the radiation line until the inspection of the quality of the object to be inspected ends can be further shortened. Therefore, the quality of the object to be inspected can be determined at an even higher speed.

[0015] The electromagnetic wave is X-rays, and the quality is at least one of the presence or absence of a foreign object mixed in, the presence or absence of damage to the object to be inspected, and the presence or absence of a defect. In this case, the presence or absence of at least one of the foreign object mixed in, the presence or absence of damage to the object to be inspected, and the presence or absence of a defect in the object to be inspected can be determined at high speed using X-rays.

[0016] At least one of the image processing section and the inspection section can also have a program automatically set by machine learning. Even in the case where the inspection apparatus employs such a program, the inspection apparatus can determine the quality of the object to be inspected at high speed.

[0017] Effects of Invention

[0018] According to one aspect of the present disclosure, an inspection apparatus capable of determining the quality of an object to be inspected at high speed can be provided. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 is a configuration diagram of an X-ray inspection apparatus according to an embodiment.

[0020] Figure 2 It is shown Figure 1 A schematic top view of a portion of the internal structure of the shielding box shown.

[0021] Figure 3 This is a functional diagram of the control unit.

[0022] Figure 4 This is a schematic diagram showing a composite image of multiple partial images.

[0023] Figure 5 (a), (c), and (e) are schematic diagrams showing the state of an object as it passes through the electromagnetic radiation. Figure 5 (b) is a diagram showing a portion of the image generated during the first timing. Figure 5 (d) is a diagram showing a portion of the image generated at the second timing. Figure 5 (f) is a diagram showing a portion of the image generated at the third timing.

[0024] Figure 6 This is a schematic diagram showing multiple partial images after image processing and a composite image of those partial images.

[0025] Figure 7 (a) is a schematic diagram showing the image processing results of a predetermined portion of the image. Figure 7 (b) is shown in Figure 7 This is a schematic diagram of the image processing result of the partial image to be generated after the partial image shown in (a). Figure 7 (c) is to Figure 7 Part of the image shown in (a) and Figure 7 The figure shown in (b) is a combination of partial images. Detailed Implementation

[0026] Hereinafter, embodiments relating to one aspect of this disclosure will be described with reference to the accompanying drawings. In the drawings, the same or equivalent parts are labeled with the same reference numerals, and repeated descriptions are omitted.

[0027] Figure 1 This is a schematic diagram of the inspection device involved in this embodiment. Figure 2 It is shown Figure 1 A schematic top view of a portion of the internal structure of the shielding box shown. Figure 1As illustrated, the inspection apparatus 1 is an apparatus that inspects the quality of an article G using electromagnetic waves. The article G is a package that contains contents such as food, and is an article to be inspected that flows in a conveyance direction A of an in-feed conveyor 51. The size of each article G in the conveyance direction A is substantially the same. In the present embodiment, the electromagnetic waves are X-rays, and the quality of the article G is at least one of the presence or absence of foreign matter mixed in the article G, the presence or absence of damage to the article G, and the presence or absence of a defect. Thus, the inspection apparatus 1 according to the present embodiment is an X-ray inspection apparatus. The quality of the article G is not limited to the presence or absence of foreign matter mixed in, the presence or absence of damage, and the presence or absence of a defect. For example, as the quality inspection of the article G, inspection of the presence or absence of a cavity, the presence or absence of a shape abnormality, the presence or absence of an accessory contained, and the like can be performed. The defect of the article G also includes, for example, the article G being an empty bag.

[0028] The article G before inspection is fed into the inspection apparatus 1 by the in-feed conveyor 51. The article G after inspection is fed out from the inspection apparatus 1 by an out-feed conveyor 52. In a case where there is an article G that is determined to be a defective product by the inspection apparatus 1, the article G is excluded, for example, at a later stage (downstream) than the inspection apparatus 1.

[0029] The inspection apparatus 1 includes an apparatus main body 2, a support leg 3, a shield box 4, a conveyance section 5, an electromagnetic wave irradiation section 6, an electromagnetic wave detection section 7, a display operation section 8, and a control section 10.

[0030] The apparatus main body 2 houses the control section 10 and the like. The support leg 3 supports the apparatus main body 2. The shield box 4 is provided to the apparatus main body 2. The shield box 4 prevents leakage of electromagnetic waves to the outside. An inspection region R in which inspection of the article G based on electromagnetic waves is performed is provided inside the shield box 4. A feed-in port 4a and a feed-out port 4b are formed in the shield box 4. The article G before inspection is fed from the in-feed conveyor 51 to the inspection region R via the feed-in port 4a. The article G after inspection is fed from the inspection region R to the out-feed conveyor 52 via the feed-out port 4b. Electromagnetic wave shielding barriers (not illustrated) that prevent leakage of electromagnetic waves are provided to the feed-in port 4a and the feed-out port 4b, respectively.

[0031] The conveyance section 5 is disposed inside the shield box 4. The conveyance section 5 conveys the article G in the conveyance direction A from the feed-in port 4a via the inspection region R to the feed-out port 4b. The conveyance section 5 is, for example, a belt-type conveyor that spans between the feed-in port 4a and the feed-out port 4b.

[0032] As Figure 1 and Figure 2As shown, the electromagnetic wave irradiation unit 6 is an electromagnetic wave source disposed within the shielding box 4. The electromagnetic wave irradiation unit 6 irradiates the article G conveyed by the conveying unit 5 with electromagnetic waves. In this embodiment, the electromagnetic wave irradiation unit 6 irradiates the article G with linear electromagnetic waves that extend perpendicularly to the conveying direction A when viewed from above. The electromagnetic wave irradiation unit 6 has, for example, an X-ray tube that emits X-rays and a collimator that expands the X-rays emitted from the X-ray tube into a fan shape in a plane perpendicular to the conveying direction A. In this case, the X-rays irradiated from the electromagnetic wave irradiation unit 6 include, for example, X-rays of various energy bands ranging from low energy (long wavelength) to high energy (short wavelength).

[0033] The electromagnetic wave detection unit 7 is a line sensor that detects electromagnetic waves irradiated by the electromagnetic wave irradiation unit 6, and is disposed within the shielded enclosure 4. The electromagnetic wave detection unit 7 is positioned on the irradiation line L of the electromagnetic waves irradiated by the electromagnetic wave irradiation unit 6, and detects electromagnetic waves that have passed through the article G. The line sensor, serving as the electromagnetic wave detection unit 7, is composed of electromagnetic wave detection elements (e.g., X-ray detection elements) arranged one-dimensionally in a direction perpendicular to the conveying direction A when viewed from above. This line sensor detects electromagnetic waves that have passed through the article G and the conveyor belt of the conveyor unit 5.

[0034] like Figure 1 As shown, the display operation unit 8 is a user interface provided on the main body 2 of the device. The display operation unit 8 displays various information and accepts input of various conditions. The display operation unit 8 is, for example, an LCD display with a touch panel, which displays the operation screen for the touch panel. In this case, the operator can input various conditions through the display operation unit 8.

[0035] The control unit 10 is an integrated circuit disposed within the main body 2 of the device. The control unit 10 controls the operation of each part of the inspection device 1. The control unit 10 includes a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and memory. The memory includes, for example, ROM (Read Only Memory) and RAM (Random Access Memory). The detection result of electromagnetic waves is input from the electromagnetic wave detection unit 7 into the control unit 10.

[0036] Figure 3 This is a functional diagram of the control unit. For example... Figure 3 As shown, the control unit 10 includes a signal acquisition unit 21, an image generation unit 22, an image processing unit 23, an inspection unit 24, a judgment unit 25, an output unit 26, and a recording unit 27.

[0037] The transmission signal acquisition unit 21 acquires the detection result of the electromagnetic wave detection unit 7. The detection result of the electromagnetic wave detection unit 7 is a one-dimensional transmission signal of the article G on the irradiation line L of the electromagnetic wave. The transmission signal is a signal generated by the electromagnetic wave detection unit 7 by detecting the electromagnetic wave that has transmitted through at least a portion of the article G. The transmission signal acquisition unit 21 transmits the acquired one-dimensional transmission signal to the image generation unit 22.

[0038] The image generation unit 22, which is configured by, for example, a main GPU, expands the transmission signal of the article G on a memory as a two-dimensional image. The memory on which the two-dimensional image is expanded is, for example, a memory included in the GPU, but is not limited thereto. The image generation unit 22 generates a transmission image (image) based on, for example, the detection result of the electromagnetic wave generated by the electromagnetic wave detection unit 7. In the present embodiment, the image generation unit 22 expands the image of the article G on the memory during the passage of the article G through the irradiation line L, and after the passage of the article G through the irradiation line L. The image of the article G is a two-dimensional image generated from the transmission signal of the article G. The image generated by reading out from the end of the article G at a certain interval (predetermined period) by the image generation unit 22 is transmitted to the image processing unit 23 and the recording unit 27 as a partial image of the article G. In the present embodiment, the partial image is an X-ray transmission image in a portion of the article G, and is generated using the transmission signal acquired within a predetermined period. The predetermined period is shorter than a period obtained by dividing the length of the article G along the conveyance direction A by a value based on the conveyance speed of the article G by the conveyance unit 5. In this case, the entire article G can be confirmed by sequentially arranging a plurality of partial images.

[0039] The image generation unit 22 can also generate a composite image representing the entire article G by synthesizing a plurality of partial images. Figure 4 is a schematic view showing a composite image of a plurality of partial images. Figure 4 The composite image GP shown is generated by synthesizing a plurality of partial images P1 to P n (n is a natural number of 2 or more) and represents the entire article G. The partial images P1 to P n are sequentially arranged along the conveyance direction A and are sequentially generated. In this way, the partial images P1 to P n are sequentially read out by the image processing unit 23, and thus are sequentially moved to the image processing unit 23 from the partial image P1. In Figure 4 , the width of the partial images P1 to P n along the conveyance direction A is constant, but is not limited thereto. For example, the width of the partial image P1 along the conveyance direction A can also be greater than the width of the other partial images P2 to P n . Each partial image is also composed of, for example, 128 x 128 pixels, taking into account the capacity of the memory.

[0040] The image processing section 23 is configured by, for example, a main GPU, and performs image processing on the image generated by the image generation section 22. Specifically, the image processing section 23 performs image processing on the partial image P read out from the image generation section 22 n The image processing is performed. At this time, even for the partial image P n-1 including a part (a boundary part where the partial images P n are contiguous to each other) of the processed partial image P Thus, for the boundary part, the image processing is performed twice. Due to this, in the case where a foreign matter exists on the boundary line, the foreign matter can be detected surely. In the present embodiment, the image processing section 23 repeatedly performs the image processing with the partial image generated each time in the image generation section 22 as a subject. The image processing on the partial image includes, for example, a binarization process, an article region specifying process, an extraction process of a feature amount from the article region, and the like. The image processing section 23 transmits the partial image after the image processing (hereinafter, also referred to simply as "processed image") to the inspection section 24 and the recording section 27. The image processing section 23 can also perform the image processing on a part in which a plurality of partial images are combined. In this case, the image processing section 23 transmits the combined partial image after the image processing to the inspection section 24 and the recording section 27.

[0041] The image processing section 23 performs the image processing on the partial image by using, for example, one or a plurality of image processing algorithms. The image processing algorithm refers to a form indicating a processing order of the image processing performed on each image. The image processing algorithm is configured by one image processing filter or a combination of a plurality of image processing filters. The plurality of image processing algorithms can be acquired from the outside via a network such as the Internet. The plurality of image processing algorithms can be acquired from an external storage medium such as a USB memory or a mobile hard disk. One or more of the plurality of image processing algorithms can employ a genetic algorithm (GA = Genetic Algorithms) which is a method applied to a mechanism of inheritance and evolution in the biological world. In this case, each image can be automatically generated from the plurality of image processing filters based on the specifications of the inspection device 1, the inspection conditions, and the like. At least a part of the plurality of image processing algorithms can be appropriately set by the operator via the display operation section 8, for example.

[0042] Instead of using the above-described image processing algorithm, the image processing section 23 can also perform the above-described image processing using a program automatically set by machine learning. Such a program is an inference program in which a prediction model (learned model) generated by machine learning, a parameter (learned parameter) obtained as a result of machine learning is incorporated. As an example of machine learning for a learned model, a neural network, a support vector machine, a genetic algorithm, and the like are exemplified. The learned model can also include a convolutional neural network, and can include a neural network of a plurality of layers (for example, 8 layers or more). That is, the learned model corresponding to the above-described program can also be generated by deep learning.

[0043] The inspection section 24 inspects the quality of the article G based on the processing result of the image processing section 23. As the quality of the partial image, at least one of the presence or absence of a foreign object, the presence or absence of a damage, and the presence or absence of a defect is exemplified, but is not limited thereto. The quality inspection of the article G in the present embodiment includes, for example, a foreign object confirmation inspection, a defect confirmation inspection, a number of contents confirmation inspection, a hollow confirmation inspection, and the like on the image after the image processing. In the present embodiment, the inspection section 24 inspects the quality of the partial image after the image processing (i.e., a part of the article G) based on the processing result of the image processing section 23 once or a plurality of times. The inspection section 24 inspects the quality of the article G using, for example, at least one of one or a plurality of image processing algorithms and a program automatically set by machine learning. The inspection section 24 transmits the quality inspection result to the determination section 25 and the recording section 27.

[0044] In the present embodiment, the inspection section 24 inspects the partial image after the image processing in parallel with the image processing by the image processing section 23. During the period in which, for example, the inspection section 24 inspects the quality of a predetermined partial image, the image processing section 23 performs image processing on a partial image generated after the predetermined partial image.

[0045] The determination section 25 determines whether or not the article G is a good product based on the quality inspection result received from the inspection section 24. For example, the determination section 25 integrates the quality inspection result for each partial image, and determines the presence or absence of a foreign object in the article G, the presence or absence of a damage of the article G, the presence or absence of a defect of the article G, and the like. The determination section 25 transmits the determination result to the output section 26 and the recording section 27.

[0046] The output section 26 outputs the determination result of the determination section 25 to at least one of a portion other than the control section 10 in the inspection device 1 and a device different from the inspection device 1. Thereby, at least one of the inspection device 1 and the device different from the inspection device 1 (for example, a sorting device disposed more downstream than the inspection device 1) can perform an action when the article G is a defective product. As another example of the above-described device different from the inspection device 1, for example, the infeed conveyor 51, the outfeed conveyor 52, a notification device, and the like are exemplified.

[0047] The recording unit 27 records signals and data generated by the control unit 10. For example, the recording unit 27 records the transmission signal sent from the transmission signal acquisition unit 21, the data of each partial image and the composite image sent from the image generation unit 22, the processed partial image sent from the image processing unit 23, the processing result of that partial image and the data of the composite image after image processing, the quality inspection result sent from the inspection unit 24, and the judgment result sent from the judgment unit 25. The recording unit 27 can also establish a correlation between, for example, the composite image after image processing for a certain item G and the quality inspection result of that item G. In this case, the confirmation of the quality inspection results of each item G based on the operator can be accelerated.

[0048] Next, refer to Figure 5 While illustrating (a) to (f), this section also describes an example of a method for obtaining a partial image of item G. Figure 5 (a), (c), and (e) respectively show schematic diagrams illustrating the state of object G under the illumination of electromagnetic wave ray L. Figure 5 As shown in (a), (c), and (e), a foreign object C is mixed into the interior of item G.

[0049] exist Figure 5 In (a), the positional relationship between the object G and the electromagnetic wave detection unit 7 during the first timing is shown. During the first timing, the foreign object C is located upstream of the electromagnetic wave detection unit 7 in the transmission direction A. Figure 5 (b) shows a partial image 31 generated at the first timing. The partial image 31 includes a portion 31a representing a part of the article G, which is a two-dimensional image unfolded in memory based on the transmitted signal acquired during a predetermined period in which the first timing is set as the end point.

[0050] exist Figure 5 In (c), the positional relationship between the object G and the electromagnetic wave detection unit 7 during the second timing is shown. During the second timing, the foreign object C overlaps with the electromagnetic wave detection unit 7 when viewed from above. Figure 5 Partial image 32, generated at the second timing, is shown in (d). Partial image 32 differs from part 31a and includes part 32a representing a portion of article G and part 32b representing foreign object C. Partial image 32 is a two-dimensional image unfolded in memory based on transmitted signals acquired during a period where the first timing is set as the start point and the second timing is set as the end point. In this embodiment, partial image 32 is generated during the period from the generation of partial image 31 to the image processing of partial image 31.

[0051] exist Figure 5 In (e), the positional relationship between the item G and the electromagnetic wave detection unit 7 during the third timing is shown. During the third timing, the foreign object C is located downstream of the electromagnetic wave detection unit 7 in the transmission direction A.Figure 5 Partial image 33 generated at the third timing is shown in (f). Partial image 33 differs from parts 31a and 32a and includes part 33a representing a portion of article G. Partial image 33 is a two-dimensional image unfolded in memory based on transmitted signals acquired during the period from the second timing to the third timing. In this embodiment, partial image 33 is generated during the period from the generation of partial image 32 to the image processing of partial image 32. Alternatively, partial image 33 may also be generated during the period until the image processing of partial image 31 is completed. In other words, the predetermined partial image in this embodiment is generated during the period until the image processing of the partial image generated between these partial images is completed.

[0052] Figure 6 This is a schematic diagram showing multiple partial images after image processing and a composite image of these multiple partial images. Figure 6 Processed images 41-43, corresponding to portions of the image after image processing, are shown. Processed image 41 is generated by... Figure 5 The partial image 31 shown in (b) is obtained by image processing. Similarly, image 42 is processed by... Figure 5 Image 43 is obtained by image processing of the portion of image 32 shown in (d), and the processed image is obtained by processing the portion of image 32. Figure 5 Image 33 shown in (f) is obtained by image processing. By combining processed images 41 to 43, a composite image 40 is obtained. Processed image 42 displays part 32b corresponding to part image 32 and part 42a representing foreign object C. Processed image 42 is obtained, for example, during the inspection of processed image 41. Processed image 43 is obtained, for example, during the inspection of processed image 41 or during the inspection of processed image 42.

[0053] Next, the effect of the inspection device 1 according to the present embodiment will be described with reference to the comparative example described below. The inspection device according to the comparative example has substantially the same configuration as the inspection device 1. However, in the inspection device according to the comparative example, image processing is performed on the image including the entire object under inspection after the image is generated. That is, in the comparative example, the partial image of the object under inspection is not generated. Further, the quality of the object under inspection is inspected on the basis of the processing result of the image including the entire object under inspection. In such a comparative example, the image processing and the quality inspection are first performed after the entire object under inspection passes through the irradiation line of the electromagnetic wave. Thus, depending on the load of the image processing or the number of quality inspection items, or the like, several seconds are required from when the object under inspection passes through the irradiation line described above until the quality inspection of the object under inspection is completed. In recent years, due to the high speed of the production line, the time required for the flow measurement of the object under inspection is very short, for example, less than 1 second. Thus, in the case where the inspection device according to the comparative example described above is introduced into the production line, the inspection device becomes a bottleneck in the production line.

[0054] In contrast, according to the inspection device 1 according to the present embodiment, the image processing section 23 repeatedly performs image processing on the partial image representing a portion of the article G as the object under inspection each time the partial image is generated. In this case, during the period in which the generation of the partial image based on the image generation section 22 is performed, the image processing section 23 can perform image processing on the already generated partial image. Thus, the time from when the article G passes through the irradiation line L until the quality inspection of the article G is completed can be shortened. Therefore, by using the inspection device 1, the quality of the article G can be determined at high speed. Thus, even if the inspection device 1 is introduced into the production line, the inspection device 1 can be prevented from becoming a bottleneck in the production line.

[0055] In the present embodiment, the inspection section 24 inspects the partial image after the image processing in parallel with the image processing by the image processing section 23. Thus, during the period in which the image processing section 23 performs image processing on the already generated partial image, the inspection section 24 can inspect the quality of the already image-processed partial image. Thus, the time from when the article G passes through the irradiation line L until the quality inspection of the article G is completed can be further shortened. Therefore, the quality of the article G can be determined at higher speed.

[0056] In the present embodiment, the electromagnetic wave is an X-ray, and the quality is at least one of the presence or absence of a foreign object, the presence or absence of a breakage, and the presence or absence of a defect. Thus, the presence or absence of a foreign object, the presence or absence of a breakage, and the presence or absence of a defect in the article G can be determined at high speed using the X-ray.

[0057] In this embodiment, at least one of the image processing unit 23 and the inspection unit 24 may also be equipped with a program automatically set by machine learning. Generally, there is a tendency that the time required for image processing, etc., using a program automatically set by machine learning is longer than that required for using only an image processing algorithm. However, in the inspection device 1 described above, the processing using the above-described program is performed during the period when the article G passes through the irradiation line L. Therefore, in the inspection device 1, the amount of data processed after the article G passes through the irradiation line L is significantly reduced compared to the comparative example described above. Therefore, even when using the above-described program, the inspection device 1 can determine the quality of the article G at high speed.

[0058] The following describes variations of the above-described embodiments. In these variations, descriptions that overlap with the above-described embodiments are omitted. Therefore, the following mainly describes the differences from the above-described embodiments.

[0059] First, refer to Figure 7 In addition to (a) and (b), it is explained that judging the quality of an item based solely on the quality inspection results of a portion of the images may lead to misjudgment. Figure 7 (a) is a schematic diagram showing the image processing results of a predetermined portion of the image. Figure 7 (a) shows a processed image 46, which is a partial image after image processing. The processed image 46 includes a portion 46a representing the foreign object C. The size of portion 46a is smaller than the size that the inspection device 1 can determine as the foreign object C. Therefore, in cases where only the quality of the processed image 46 is checked, for example, it is highly likely that the quality of the processed image 46 will be determined to be good.

[0060] Figure 7 (b) is shown Figure 7 This is a schematic diagram of the image processing result of the partial image to be generated after the partial image shown in (a). Figure 7 (b) shows a processed image 47, which is a partial image after image processing. Processed image 47 includes a portion 47a representing the foreign object C. The size of portion 47a, like portion 46a, is smaller than the size that the inspection device 1 can determine as the foreign object C. Therefore, if only the quality of processed image 47 is checked, for example, it is highly likely that the quality of processed image 47 will be determined to be good. Consequently, if only the quality inspection results of each processed image are considered, there is a possibility of misjudging the quality of the item.

[0061] In contrast, in this variant, in addition to the initially generated partial image, the quality of the composite image obtained by combining partial images processed from multiple images is checked. Figure 7 (c) is to Figure 7 Part of the image shown in (a) and Figure 7The diagram shown in (b) combines a portion of the image. In this variation, image processing is performed by adding the boundary region of the previous portion of the image, which is adjacent to the portion of the image to be processed, to the portion of the image being processed. Specifically, as shown in... Figure 7 As shown in (c), a composite image 45 is generated by performing image processing (i.e., repeatedly processing) by adding a portion of the previously processed image to a portion of the current image. This composite image 45 includes a combination of... Figure 7 Parts 46a and 47a shown in (a) and (b) are part 45a. Part 45a is the size or larger that the inspection device 1 can determine is a foreign object C. Therefore, when inspecting the quality of the composite image 45, it is possible to more reliably determine whether the article contains a foreign object.

[0062] In this modified example described above, the same effects as the above-described embodiments are achieved. Furthermore, in this modified example, in addition to the initially generated partial image, the quality of the composite image obtained by combining partial images processed from multiple images is checked. Therefore, it is difficult to misjudge the quality of the item. Thus, in this modified example, both high-speed and high-precision quality inspection of the item can be achieved simultaneously.

[0063] The above describes one aspect of the embodiments and modifications involved in this disclosure, but this disclosure is not limited to the embodiments and modifications described above. In, for example, the embodiments and modifications described above, the image processing unit and the inspection unit are different components from each other, but this is not a limitation. For example, the inspection unit may also be part of the image processing unit. In any aspect of this disclosure, the control unit performs image processing and inspection on a portion of the image.

[0064] In the above variations, the synthesized image is obtained by processing a portion of the previously processed image plus a portion of the current image, but is not limited to this. The synthesized image can also be obtained, for example, by combining a portion of the previously processed image and a portion of the current image. That is, a synthesized image can be generated without further processing the previously processed image. The synthesized image can also, for example, combine the entirety of the previously processed image and the current image. Alternatively, it can combine not only the previously processed image but also previously processed images with the current image. In other words, the synthesized image can also include multiple processed images.

[0065] In the above-described embodiments and the above-described modifications, the partial image after the image processing can be divided into a plurality of regions, and the quality of each region can be inspected. In this case, the quality of the plurality of regions divided by the first method and the quality of the plurality of regions divided by a second method different from the first method can be inspected respectively.

[0066] Explanation of Reference Signs

[0067] 1: inspection apparatus; 2: apparatus main body; 3: support leg; 4: shield box; 5: conveyance section; 6: electromagnetic wave irradiation section; 7: electromagnetic wave detection section; 8: display operation section; 10: control section; 21: transmission signal acquisition section; 22: image generation section; 23: image processing section; 24: inspection section; 25: determination section; 26: output section; 27: recording section; 31 to 33, P1 to P n : partial image; 40, 45, GP: composite image; 41 to 43, 46, 47: processed image; C: foreign matter; G: article; L: irradiation line.

Claims

1.An inspection apparatus comprising: a conveyance section that has a conveyance surface and conveys an object to be inspected placed on the conveyance surface; an image generation section that, in a state in which the object to be inspected is placed on the conveyance surface, develops a one-dimensional transmission signal of the object to be inspected that passes through an irradiation line of electromagnetic waves into a two-dimensional image on a memory; an image processing section that, each time a partial image of a portion of the object to be inspected is generated in the image generation section, performs image processing with the partial image as a target; and an inspection section that inspects a quality of a composite image obtained by combining a plurality of the partial images after the image processing, wherein the image processing section performs image processing on at least a portion of a partial image of a previous image processing and a partial image of a current image processing. 2.The inspection apparatus according to claim 1, wherein the inspection section inspects the partial image after the image processing in parallel with the image processing by the image processing section. 3.The inspection apparatus according to claim 1 or 2, wherein the electromagnetic waves are X-rays, and the quality is at least one of presence or absence of a foreign object and presence or absence of a defect. 4.The inspection apparatus according to claim 1 or 2, wherein at least one of the image processing section and the inspection section is provided with a program automatically set by machine learning. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​

Citation Information

Patent Citations

  • Method and device for inspecting sheetlike article

    JP2010101692A

  • Optical inspection device

    JP2017125800A