KR102260734B1 - Device for inspecting products and method using the same - Google Patents

Abstract

The present invention relates to a product inspection device for inspecting the surface and shape of a product, comprising a first analysis unit and a second analysis unit performing a surface and shape inspection of a product by matching and comparing an image of the inspected product with various data obtained through a non-defective image. Through the automatic product inspection characteristics of the first analysis unit and the second analysis unit, the consumption of manpower is minimized during product inspection and the inspection accuracy is further improved compared to the visual and human judgment. In addition, the product inspection method according to the present invention uses the product inspection device. Accordingly, the product inspection method has an effect corresponding to the effect of the product inspection device described above.

Description

Device for inspecting products and method using the same

The present invention relates to a product inspection apparatus and a product inspection method, and more particularly, to a product inspection apparatus and a product inspection method for performing surface and shape inspection of a final product.

The final product produced through a series of processes is tested to determine whether defects exist in its surface and shape as the final process.

However, when such product inspection is performed by the naked eye and human judgment, the process efficiency is inevitably reduced due to the accumulation of worker's fatigue, and in the case of a precision product, the inspection reliability is inevitably low.

As part of solving the above problems, an object of the present invention is to provide a product inspection apparatus and a product inspection method for automatically performing product inspection by replacing human resources.

The product inspection method according to the present invention includes an inspection image acquiring step in which at least one first inspection image that is a surface image of an inspection target product, which is an inspection target product, is acquired, and the first inspection image is analyzed in the inspection image acquiring step, so that the It characterized in that it comprises a quality determination step of determining whether the surface defect of the inspection item.

Also, in the quality determination step, when a distribution of a gray level indicating less than a predetermined brightness is detected in the first inspection image, it is characterized in that the inspected product is determined to be defective.

In addition, the product inspection method is performed before the inspection image acquiring step, and is performed before the reference image acquiring step in which at least one first reference image that is a surface image of a good product is acquired, and the inspection image acquiring step, In the image acquisition step, the first reference image is partitioned into at least one first region of interest corresponding to the gray level distribution, and determining whether the surface of the inspection object is good or bad The method further includes a reference setting step of setting at least one first reference, which is a brightness value of a gray level, for each of the first regions of interest as a reference for performing the examination, wherein the obtaining the examination image includes: the first examination image is obtained corresponding to the first reference image, and in the determining step, the first inspection image is divided into at least one or more first corresponding ROIs corresponding to the first ROI, and then each of the first When the corresponding ROI matches each of the first criteria, and there is the first corresponding ROI that satisfies the first criterion, it is characterized in that it is determined that the corresponding object to be inspected is defective.

In addition, in the reference setting step, the first reference is characterized in that the size of the gray level is further included.

In addition, the step of acquiring the reference image further includes a process of acquiring at least one second reference image that is a shape image of a good product, and the step of dividing the region of interest includes a predetermined degree of similarity with respect to shape distribution in the second reference image. The step of dividing each of the sections in which is present into a second region of interest is further included, and the step of setting the reference includes a dimension range for each of the second regions of interest as a criterion for judging the shape or quality of the inspected object. The method further includes a process of setting a second reference, which is a reference, in the inspection image acquiring step, at least one second inspection image that is a shape image of the inspected object is acquired to correspond to the second reference image, In the step of dividing the second examination image into at least one or more second corresponding ROIs corresponding to the second ROI, each of the second corresponding ROIs is matched with the corresponding second criterion, and the second inspection image is divided into one or more second corresponding ROIs. When the second corresponding ROI that does not satisfy the second criterion is detected, the corresponding inspection object is determined to be defective.

The product inspection apparatus according to the present invention comprises a first determination module configured to analyze at least one first inspection image, which is a surface image of an inspection target product, to determine whether the inspection target product has a surface defect. It is characterized in that it contains a part.

In addition, the first determination module is characterized in that when a distribution of a gray level indicating less than a predetermined brightness in the first inspection image is detected, it is determined that the inspected product is defective.

In addition, the first analysis unit divides a first reference image, which is a surface image of a non-defective product, into at least one first region of interest corresponding to a distribution of a gray level having a predetermined similarity. A first division module, and a first criterion setting module for setting at least one first criterion, which is a brightness value of a gray level, for each first region of interest as a criterion for judging the quality of the surface of the inspected object; , a first corresponding partitioning module for partitioning the first examination image into at least one first corresponding ROI corresponding to the first ROI, wherein the first determination module is configured to divide each of the first corresponding ROIs. When the first corresponding region of interest that matches the corresponding first criterion and satisfies the first criterion exists, the inspected object is determined to be defective.

In addition, the first reference setting module, when setting the first reference, it characterized in that it further includes the size of the gray level to consider.

In addition, the product inspection apparatus includes a second analysis unit configured to include a second determination module that analyzes at least one second inspection image that is a shape image of an inspection target product and determines whether the inspection target product has a shape defect. It is characterized in that it further comprises.

In addition, the second analysis unit includes a second partitioning module for partitioning each of the sections having a predetermined similarity with respect to shape distribution in at least one second reference image that is a shape image of a good product into a second region of interest; a second standard setting module for setting a second standard, which is a standard including a dimensional range for each of the second areas of interest, as a standard for judging whether the shape of the inspection item is good or bad; and the second examination image corresponding to the second area of interest and a second corresponding partitioning module for partitioning the ? into at least one or more second corresponding ROIs, wherein the second determining module is configured to match each of the second corresponding ROIs with the corresponding respective second criteria to match the second criteria. When the second corresponding region of interest that is dissatisfied exists, the inspected object is determined as defective.

The product inspection apparatus according to the present invention includes a first analysis unit and a second analysis unit for performing a surface and shape inspection of a product by matching and comparing an image of the inspected product with various data obtained through the non-defective image, and such Through the automatic product inspection characteristics of the first analysis unit and the second analysis unit, the consumption of manpower is minimized during product inspection and the inspection accuracy is further improved compared to the visual and human judgment.

In addition, the product inspection method according to the present invention uses the product inspection apparatus. Accordingly, the product inspection method has an effect corresponding to the effect of the product inspection apparatus described above.

1 is a block diagram showing a product inspection apparatus according to the present invention.
FIG. 2 is a photograph showing an example of an image acquired by the surface image acquisition member shown in FIG. 1 .
FIG. 3 is a detailed configuration diagram illustrating the first analysis unit shown in FIG. 1 .
4A is a schematic diagram showing an example of use of the first partition module shown in FIG. 3 .
FIG. 4B is a schematic diagram illustrating an example in which a first criterion is set in the first region of interest shown in FIG. 4A .
4C is a schematic diagram illustrating an example of matching the first ROI and the first criterion shown in FIG. 4B with respect to the first corresponding ROI.
FIG. 5 is a photograph showing an example of an image acquired by the shape image acquiring member shown in FIG. 1 .
FIG. 6 is a detailed configuration diagram illustrating a second analysis unit illustrated in FIG. 1 .
7A is a schematic diagram showing an example of use of the second partition module shown in FIG. 6 .
FIG. 7B is a schematic diagram illustrating an example in which a second criterion is set in the second region of interest shown in FIG. 7A .
FIG. 7C is a schematic diagram illustrating an example of matching the second ROI and the second criterion shown in FIG. 7B with respect to the second corresponding ROI.
8 is a flowchart illustrating a product inspection method according to the present invention.
9A is a detailed flowchart illustrating a product surface inspection process of the first analysis unit shown in FIG. 8 .
9B is a detailed flowchart illustrating a product shape inspection process of the second analysis unit shown in FIG. 8 .

Prior to the detailed description of the present invention, specific details for carrying out the present invention are included in the embodiments and drawings to be described below, and the same reference numerals described throughout the specification refer to the same components.

In addition, the singular expressions in this specification will also include the plural unless otherwise specified in the phrase.

Hereinafter, a product inspection apparatus and a product inspection method according to the present invention will be described with reference to the drawings.

1 is a block diagram showing a product inspection apparatus according to the present invention, and first, with reference to this, a product inspection apparatus according to the present invention will be described.

As shown in FIG. 1 , the product inspection apparatus 1000 according to the present invention includes an image acquisition unit 100 , a rotation unit 200 , a lighting unit 300 , a first analysis unit 400 , and a second analysis unit 500 . ) may be included.

In the case of the image acquisition unit 100 , it is a means for acquiring the surface and shape images of the inspected product 1 , which is a product to be determined as good or bad, and may be composed of various image acquisition means such as a camera, a scanner, a camcorder, and the like.

And, the rotating unit 200 is a means for rotating the inspection target (1) about a predetermined rotation axis.

In this case, while the object 1 to be inspected 1 is rotated by the rotating unit 200 , the image acquisition unit 100 takes the product a plurality of times.

Corresponding to the rotation of the rotating unit 200, the entire structure of the inspection object 1 can be exposed to the image acquisition unit 100, and through this, even if a plurality of image acquisition units 100 are not provided, the It becomes possible to photograph the whole structure of the inspection object 1 .

In addition, at least one of the rotation speed of the rotating unit 200 and the photographing speed of the image acquisition unit 100 may be variously manipulated.

In addition, the lighting unit 300 performs a function of creating brightness suitable for photographing by irradiating a predetermined light around the object 1 to be inspected.

In addition, the first analysis unit 400 analyzes the image of the inspection object 1 provided through the image acquisition unit 100 to determine the quality of the product surface.

In this case, various types of defects such as scratches, foreign matter adhesion, and black spots may exist as surface defects of the product.

In addition, when light is irradiated using the lighting unit 300 or the like on the surface of the product on which such product defects exist, the periphery of the defective portion has a predetermined value compared to the normal surface in response to the shape, such as the area, height, or depth of the defective portion. shading occurs.

In addition, due to the shade composition, a gray level indicating a darker value is formed around the defective portion compared to the normal surface portion.

In this way, the section in which the above-described gray level difference occurs can be referred to as a section in which product surface defects exist, and the first analysis unit 400 determines the surface brightness of the product surface, that is, the gray level. Analyze whether the product surface is defective or not.

In addition, the second analysis unit 500 analyzes the image of the inspected product 1 provided through the image acquisition unit 100 to determine whether the product has a shape defect.

In order to determine such a product shape defect, various factors such as dimensions, curvature, thickness, and inclination of each part of the product may be considered.

Next, each configuration of the product inspection apparatus 1000 will be described in detail with further reference to other drawings.

FIG. 2 is a photograph showing an example of an image acquired by the surface image acquisition member shown in FIG. 1 .

FIG. 3 is a detailed configuration diagram illustrating the first analysis unit shown in FIG. 1 .

And, FIG. 4A is a schematic diagram illustrating an example of using the first partition module shown in FIG. 3 , and FIG. 4B is a schematic diagram illustrating an example in which a first criterion is set in the first region of interest shown in FIG. 4A , FIG. 4C is a schematic diagram illustrating an example of matching the first ROI and the first criterion shown in FIG. 4B with respect to the first corresponding ROI.

The image acquisition unit 100 and the first analysis unit 400 will be described in detail with further reference to FIGS. 2, 3, and 4A to 4C.

First, as shown in FIG. 1 , the image acquisition unit 100 may include a surface image acquisition member 110 .

The surface image acquisition member 110 is a means for acquiring a surface image of a product to be analyzed by the first analysis unit 400 outlined above, and may be configured as a line camera or the like.

Such a line camera or the like is a means for imaging the shape of a photographed product into a two-dimensional line shape, as illustrated in FIG. 2 .

For example, in FIG. 1, a predetermined cylindrical structure is shown as the object to be inspected 1, and when the cylindrical structure is photographed with a line camera or the like, a kind of flat image is obtained as shown in FIG. 2, not the three-dimensional shape of the photographing part. can be

In this way, by using a line camera or the like, the section where the surface defect exists and the section around it can appear as if they exist on the same plane.

Thus, the gray level distribution on the surface of the product can appear on the same plane, and accordingly, the gray level distribution on the surface of the product can be grasped more clearly.

And, as shown in FIG. 3 , the first analysis unit 400 includes a first division module 410 , a first reference setting module 420 , a first corresponding division module 430 , and a first determination A module 440 may be included.

The first analysis unit 400 sets standards for product surface inspection and performs product surface inspection according to the set standards.

At this time, in order to set the standard for determining the quality of the product surface, it is possible to refer to the surface image data of the non-defective product, and the first partition module 410 and the first standard setting module 420 are connected to the surface image of the non-defective product through the surface image of the non-defective product. Inspection criteria can be set.

More specifically, as shown in FIG. 4A , the first partitioning module 410 identifies sections in which gray levels are formed within a predetermined distribution in the first reference image is, which is a surface image of a non-defective product.

Also, the first partition module 410 designates each of the sections having a uniform gray level distribution as a first region of interest (roi, rs).

For example, in the first reference image (is) shown in FIG. 4A , as gray level distribution ranges are 145 to 125, 105 to 85, 65 to 45, each of the sections in which the gray level distribution is within 20 is roi 1, It is exemplified as being partitioned into the first region of interest rs such as roi 2 and roi n.

In this case, the above-mentioned matters are merely examples for explanation, and a criterion for partitioning at least one first region of interest rs may be devised through various parameters as well as the aforementioned gray level distribution map.

In addition, the above-described first region of interest (rs) partitioning process may be performed by a user directly manipulating the first partitioning module 410 .

And, as shown in FIG. 4B , the first standard setting module 420 sets a first standard p1 that is a standard for judging the quality of the inspected product 1 .

In detail, as described above, when a defect exists on the surface of the product, it appears darker than in the surrounding section. In this case, the first reference setting module 420 recognizes a surface defect for each first region of interest rs. Executes the function to set the gray level range to be used.

For example, as shown in FIG. 4B , the first reference p1 in the first region of interest rs corresponding to roi 1 is set to have a gray level of 115 or less.

In addition, in the above-described example, the first reference p1 is set to have a gray level of 115 or less and a size of 10 or more, and various types such as a pixel may be set as the size unit at this time.

That is, in this case, whether or not the surface defect is determined according to whether a section having a gray level of 115 or less and a size of 10 or more is detected within the first region of interest rs.

As such, the first criterion p1 means the gray level range and the size to be determined as a surface defect in the image acquired by the line camera, and this first criterion p1 setting is determined by the user as the first criterion This may be performed by manipulating the setting module 420 .

Next, as shown in FIG. 4C , the first corresponding partitioning module 430 describes the first inspection image ic, which is the surface image of the inspection target object 1 , the first region of interest rs ) corresponding to the first corresponding region of interest (rc).

At this time, the first reference image (is) and the first inspection image (ic) are the arrangement of the lighting unit 300 and the image acquisition unit 100 , the number of rotations of the rotation unit 200 , and the image acquisition unit 100 . It was obtained based on a case in which all conditions for product inspection, such as shooting speed and the overall lighting condition of the inspection site, correspond to each other.

Assuming that the same photographing conditions are satisfied, as illustrated in FIG. 4C , the first reference image is and the first inspection image ic have the same overall size.

And, by performing a simple process such as symmetrically moving the first region of interest rs to the first inspection image ic through the size matching action between the above-described both images (is, ic), The position and size of the first relative region of interest rc on the first examination image ic can be precisely set by matching the position and size of the relative first region of interest rs on the first reference image is.

In addition, through the above-described image acquisition condition matching, the gray level distribution in the first inspection image ic is formed to be the same as that of the first reference image is except for the defective part, so that the first inspection image ( ic) and the first reference image (is), the defect part can be clearly identified at a glance.

Next, as described above, after the size matching between the images (is, ic) and the size and arrangement matching between the two regions of interest (rs, rc) are performed, the first determination module 440 performs the first criterion ( p1) are custom-matched to a first corresponding ROI rc corresponding to each of the first ROIs rs.

At this time, if there is a first corresponding region of interest rc that satisfies the first criterion p1 , the first determination module 440 determines that the inspected object 1 is defective.

For example, referring to FIG. 4C , a gray level value of 32 and a size of 13 are detected in the first corresponding region of interest rc corresponding to roi 2 .

In this case, when the gray level of the first reference p1 for the corresponding first ROI rc and the corresponding first ROI rs is 75 or less, but a section whose size is 10 or more is detected It is set to recognize this as a surface defect.

That is, in the case of the first corresponding ROI rc corresponding to roi 2, the corresponding first criterion p1 is satisfied, so that a surface defect exists on the corresponding first ROI rc. can be judged as

Therefore, in this case, it is finally determined that a surface defect exists in the product including the first corresponding region of interest rc corresponding to roi 2 .

The image division and individual standard setting characteristics for each division of the first analysis unit 400 prevent misjudgment of surface defects of the product due to various factors, such as uneven illumination of the product surface and a draft phenomenon that may occur when photographing a rotating product. is to prevent

For example, when a first inspection image ic corresponding to the first reference image is shown in FIGS. 4A to 4C is obtained, if an accurate product surface inspection is performed, the first inspection image must be recognized as a good product. .

However, in this case, as shown in FIGS. 4A to 4C , due to various factors such as non-uniformity of illumination, the average gray level distribution of the first reference image (is) is somewhat like the roi 1, roi 2 to roi n sections. may be formed differently.

In this case, if the gray level reference value for determining whether a surface defect exists is simply set in a single way, the product may be recognized as defective by the gray level distribution diagram described above.

As a detailed example of this, if it is assumed that the surface defect criterion is set to a gray level of 110 or less, and a first inspection image ic corresponding to the first reference image is is obtained, the average gray level distribution diagram All of the sections roi 2 to roi n in which is 110 or less are inevitably recognized as surface defects.

In this way, if the surface defect criteria are simply and unitarily set, the roi 2 to ro n sections in which the surface defects do not actually exist are recognized as defective by other factors such as non-uniformity of lighting, and misjudgment will occur.

In order to prevent such a misjudgment, a separate lighting unit 300 number and arrangement structure are devised so that the lighting for shooting is evenly distributed on the surface of the product, and the surface image acquisition member 110 is used as a somewhat high-performance product to prevent draft phenomena. There may be other ways to prepare.

However, in order to carry out the design of the lighting unit 300 arrangement plan and the replacement plan of the surface image acquisition member 110 , considerable equipment cost and research and development period are inevitably consumed, which may cause a significant economic burden.

In addition, even if the above-described facility arrangement and replacement plan is performed, errors such as occurrence of uneven distribution of lighting within a significant level are still unavoidable problems, and thus, there is still a possibility of erroneous judgment of surface defects.

However, if the first analyzer 400 is provided, as described above, each of the sections having the same gray level distribution can be individually inspected according to the principle of setting the first reference p1 described above. The possibility of that misjudgment can be significantly reduced.

In addition, even if an algorithm capable of finely detecting the change rate of the gray level compared to the surrounding pixels is provided in addition to such a high-level lighting arrangement, misjudgment problems may still occur.

More specifically, since the gray level change rapidly occurs near the branch point of the gray level change amount generated due to the non-uniformity of illumination, the algorithm may erroneously judge the aforementioned branch point section as defective.

For example, taking the case shown in FIGS. 4A to 4C as an example, at the boundary point between each first region of interest rs or the first corresponding region of interest rc, an abrupt change in roi is caused due to the non-uniform distribution of illumination. , in this case, the algorithm may misjudg these boundary point sections as defective.

Therefore, if the image acquisition conditions such as lighting brightness are kept constant, the lighting distribution change point, which is a problem common to other surface inspection methods, through the image section unique to the first analysis unit 400 and individual standard setting characteristics for each section There is an effect that the high probability of misjudgment in the section is remarkably reduced.

Next, FIG. 5 is a photograph showing an example of an image acquired by the shape image acquiring member shown in FIG. 1 .

And, FIG. 6 is a detailed configuration diagram showing the second analysis unit shown in FIG. 1 .

And, FIG. 7A is a schematic diagram illustrating an example of use of the second partition module shown in FIG. 6 , and FIG. 7B is a schematic diagram illustrating an example in which a second criterion is set in the second region of interest shown in FIG. 7A , and FIG. 7C . is a schematic diagram illustrating an example of matching the second ROI and the second criterion shown in FIG. 7B with respect to the second corresponding ROI.

The image acquisition unit 100 and the second analysis unit 500 will be described in detail with further reference to FIGS. 5, 6 and 7A to 7C .

First, as further referenced to FIG. 1 , the image acquisition unit 100 may further include a shape image acquisition member 120 .

The shape image acquisition member 120 is a means for acquiring a shape image of a product to be analyzed by the second analysis unit 500 outlined above, and may be configured as an area camera or the like.

At this time, as an example of use of the shape image acquisition member 120 , the side shape of the ampoule is shown in FIG. 5 .

And, referring to FIG. 6 , the second analysis unit 500 includes a second division module 510 , a second reference setting module 520 , a second corresponding division module 530 , and a second determination module. 540 may be included.

As described above, the second analysis unit 500 is similar to the first analysis unit 400 described above, and sets an inspection standard through acquisition of a non-defective product image and performs a product shape inspection according to the set inspection standard.

More specifically, as shown in FIG. 7A , the second partitioning module 510 divides each of the sections in which the shape is identical in the second reference image ih, which is a shape image of a good product, into a second region of interest ( Region of interest, roi, rh).

As an example of this, as shown in FIG. 7A , in the second reference image ih showing the ampoule, the area corresponding to the body is designated as roi 1, and the area corresponding to the neck is designated as roi 2. .

In addition, the above-described second region of interest (rh) partitioning process may be performed by a user directly manipulating the second partitioning module 510 .

And, as shown in FIG. 7B , the second reference setting module 520 sets a second reference p2 that is a reference for determining the quality of the inspected product 1 .

In detail, as shown in FIG. 7B , at least one parameter for shape inspection, such as a dimension range, a length range, a radius range, and a curvature range, of each part may be assigned to each second region of interest rh.

In addition, the above-described setting of the second reference p2 may be performed by the user operating the second reference setting module 520 .

And, as shown in FIG. 7C , the second corresponding division module 530 corresponds to the second inspection image ik, which is a shape image of the inspected object 1, which is the inspection object, to the second region of interest rh. to perform a function of partitioning into the second corresponding region of interest (rk).

At this time, the second reference image (ih) and the second inspection image (ik) are the arrangement of the lighting unit 300 and the image acquisition unit 100 , the number of rotations of the rotation unit 200 , and the image acquisition unit 100 . It was obtained based on the case where all conditions for product inspection, such as shooting speed and the overall lighting condition of the inspection site, correspond to each other.

In this case, similarly to the principle corresponding to the setting of the first corresponding ROI rc described above with reference to FIGS. 4A to 4C , the preset second ROI rh is symmetrical to the second examination image ik. By moving, the second corresponding region of interest rk may be set.

Through such an image acquisition condition and the ROI (rh, rk) matching characteristic, even if the second corresponding ROI rk is matched with each of the second criteria p2 corresponding thereto, the probability of an erroneous judgment is remarkably high. is reduced, and this matching check is performed by the second determination module 540 .

In detail, as shown in FIG. 7C , the second determination module 540 assigns each of the second corresponding regions of interest rk to the preset second reference p2 for each corresponding second region of interest rh. match with

At this time, when the second corresponding ROI rk that does not satisfy the second criterion p2 is detected, the corresponding inspected object 1 is recognized as defective.

In addition, although FIGS. 5 and 7A to 7C are illustrated based on the side shape of the product, this is only an example for description.

In more detail, various positions of the inspected object 1 such as the front, plane, and bottom surfaces can be photographed through the inspection article rotation function of the rotating unit 200 described above, and at this time, the second analysis described above for each position image It will be noted that the functions of the unit 500 may be performed.

Next, Figure 8 is a flow chart showing a product inspection method according to the present invention, with reference to this will be described as a whole for the product inspection method according to the present invention.

8, the product inspection method according to the present invention uses the above-described product inspection apparatus 1000, and includes a reference image acquisition step (S100), a region of interest partitioning step (S200), and a reference setting step (S300). , including an inspection image acquisition step (S400) and a quality determination step (S500).

First, in the reference image acquisition step ( S100 ), an image of a good product is acquired through the surface image acquisition member 110 and the shape image acquisition member 120 described above.

Next, in the region of interest segmentation step ( S200 ) and the reference setting step ( S300 ), a predetermined inspection standard is set through the first analysis unit 400 and the second analysis unit 500 through the non-defective image. .

Next, in the inspection image acquiring step ( S400 ), an image of the inspection target object 1 to be inspected is acquired through the surface image acquiring member 110 and the shape image acquiring member 120 .

Finally, in the quality determination step (S500), the inspection standard set in the region of interest partitioning step (S200) and the reference setting step (S300) is applied to the image of the object to be inspected 1, and the first analysis unit 400 ) and the quality of the surface and shape of the product to be inspected is determined through the second analysis unit 500 .

As such, the product inspection method is intended to determine whether the surface and shape of the product is defective.

9A is a detailed flowchart illustrating a product surface inspection process of the first analysis unit shown in FIG. 8 , and the above-mentioned surface inspection method of the product will be described in detail with reference to this.

First, as further referred to in FIG. 9A , in the reference image acquisition step S100 , the above-described first reference image is is acquired through the surface image acquisition member 110 .

Next, in the region of interest segmentation step S200 , a process of setting at least one of the first regions of interest rs described above in the first reference image is through the first segmentation module 410 proceeds. do.

As described above with respect to the product inspection apparatus 1000 , the first region of interest rs is set in the first reference image is based on the gray level distribution.

Next, in the reference setting step S300 , the aforementioned first reference p1 is set for each of the first region of interest rs through the aforementioned first reference setting module 420 .

As described above, the first reference p1 represents the numerical value and size of a gray level to be determined as a surface defect.

Next, in the inspection image acquisition step S400 , the first inspection image ic, which is a surface image of the object 1 to be inspected, is acquired through the surface image acquisition member 110 .

In this case, the first inspection image ic is acquired under the same inspection condition as the first reference image is.

Next, in the quality determination step ( S500 ), at least one first corresponding region of interest rc is set in the first inspection image ic through the above-described first corresponding partition module 430 , which is described above. As described above, the number, interval, and size corresponding to the first region of interest rs on the first examination image ic are set.

Finally, the first criterion p1 corresponding to each of the first corresponding ROIs rc is matched through the above-described first determination module 440 .

In this case, if there is a first corresponding region of interest rc that satisfies at least one first criterion p1 in the first inspection image ic, it is determined that the corresponding inspection object 1 is defective.

9B is a detailed flowchart illustrating a product shape inspection process of the second analysis unit shown in FIG. 8 , and the above-mentioned product shape inspection method will be described in detail with reference to this.

First, as shown in FIG. 9B , in the reference image acquiring step S100 , the above-described second reference image ih is acquired through the shape image acquiring member 120 .

Next, in the region of interest partitioning step S200 , a process of setting at least one of the aforementioned second regions of interest rh in the second reference image ih through the aforementioned second partitioning module 510 proceeds. do.

As described above with respect to the product inspection apparatus 1000 , the second region of interest rh is designated as a region of interest in each of the sections in the product image in which the same structural shape exists.

Next, in the reference setting step S300 , the aforementioned second reference p2 is set for each second region of interest rh through the aforementioned second reference setting module 520 .

As described above, the second reference p2 may include at least one or more of various parameters for inspecting the shape of a product, such as a range of dimensions of each part, a range of curvature, and a radius range.

Next, in the inspection image acquiring step ( S400 ), the above-described second inspection image ik , which is a shape image of the inspected object 1 , is acquired through the surface image acquiring member 110 .

In this case, the second inspection image ik is obtained under the same inspection condition as the second reference image ih.

Next, in the quality determination step S500 , at least one second corresponding region of interest rk is set in the second inspection image ik through the above-described second corresponding division module 530 , which is described above. As described above, the number, interval, and size corresponding to the second region of interest rh on the second examination image ik are set.

Finally, the second criterion p2 is matched for each second corresponding region of interest rk through the above-described second determination module 540, and at least one second criterion in the second inspection image ik is matched. If there is a second corresponding region of interest rk dissatisfied with (p2), it is determined that the inspected object 1 is defective.

In addition, the method for inspecting the surface and shape of the product described above with reference to FIGS. 9A and 9B may be performed simultaneously in response to each step, and only one of them may be selectively performed.

As described above, the present invention has a main technical idea to provide a product inspection apparatus and a product inspection method, and the embodiment described above with reference to the drawings is only one embodiment, and the true scope of the present invention is The scope of the claims and equivalent embodiments that may exist in various ways.

1: Item to be inspected
1000: product inspection device according to the present invention
100: image acquisition unit
110: surface image acquisition member
120: shape image acquisition member
200: rotating part
300: lighting unit
400: first analysis unit
410: first compartment module
420: first reference setting module
430: first corresponding partition module
440: first determination module
500: second analysis unit
510: second compartment module
520: second reference setting module
530: second corresponding partition module
540: second determination module

Claims (11)

an inspection image acquisition step of acquiring at least one first inspection image that is a surface image of an inspection target product; and
a quality determination step of determining whether a surface defect of the inspected object is present by analyzing the first inspection image in the inspection image acquiring step;
The quality determination step is,
When a gray level distribution indicating less than a predetermined brightness is detected in the first inspection image, it is determined that the inspection object is defective,
The product inspection method is,
a reference image acquisition step performed before the inspection image acquisition step, wherein at least one first reference image that is a surface image of a good product is acquired;
It is performed before the inspection image acquisition step, and the first reference image in the reference image acquisition step is divided into at least one first region of interest corresponding to a distribution of a gray level. zone division step; and
a reference setting step of setting at least one first criterion, which is a brightness value of a gray level, for each of the first regions of interest as a criterion for judging the quality of the surface of the inspected object; further comprising,
The inspection image acquisition step includes:
The first inspection image is obtained to correspond to the first reference image,
The quality determination step is,
After the first examination image is partitioned into at least one first corresponding ROI corresponding to the first ROI, each of the first corresponding ROIs is matched to the corresponding first criterion, and the first inspection image is divided into one or more first corresponding ROIs. If the first corresponding region of interest that satisfies the criteria exists, it is determined that the inspected item is defective,
The standard setting step is
The first criterion further includes the size of the gray level,
The reference image acquisition step is,
The process of acquiring at least one or more second reference images, which is a shape image of a good product, is further included,
The segmentation step of the region of interest comprises:
In the second reference image, each of the sections having a predetermined similarity with respect to the shape distribution is partitioned into a second region of interest;
The standard setting step is
The step of setting a second standard, which is a standard including a dimensional range for each of the second regions of interest, is set as a standard for determining whether the shape of the inspected object is good or bad
The inspection image acquisition step includes:
At least one second inspection image that is a shape image of the object to be inspected is obtained to correspond to the second reference image,
The quality determination step is,
After the second examination image is divided into at least one or more second corresponding ROIs corresponding to the second ROI, each of the second corresponding ROIs is matched with the corresponding second criterion to determine the second criterion The product inspection method, characterized in that when the unsatisfactory second corresponding region of interest is detected, the inspected object is determined to be defective.
delete delete delete delete a first analysis unit configured to include a first determination module that analyzes at least one first inspection image that is a surface image of an inspection target product and determines whether the inspection target product has a surface defect;
The first determination module,
When a gray level distribution indicating less than a predetermined brightness is detected in the first inspection image, it is determined that the inspection object is defective,
The first analysis unit,
A first partitioning module for partitioning a first reference image, which is a surface image of a non-defective product, into at least one first region of interest corresponding to a distribution of a gray level having a predetermined degree of similarity; a first reference setting module for setting at least one first criterion, which is a brightness value of a gray level, for each first region of interest as a criterion for judging the quality of the surface of the inspected object; and corresponding to the first region of interest and a first corresponding partitioning module for partitioning the first inspection image into at least one or more first corresponding ROIs,
The first determination module,
If there is a first corresponding ROI satisfying the first criterion by matching each of the first corresponding ROIs to each of the corresponding first criteria, the corresponding inspection object is determined as defective;
The first reference setting module further includes the size of the gray level when setting the first reference,
The product inspection device,
a second analysis unit configured to include a second determination module that analyzes at least one second inspection image that is a shape image of an inspection target product and determines whether the inspection target product has a shape defect;
The second analysis unit,
a second partitioning module for partitioning each of sections having a predetermined similarity with respect to shape distribution in at least one second reference image that is a shape image of a non-defective product into a second region of interest; a second reference setting module for setting a second reference, which is a reference including a dimensional range for each second region of interest, as a reference, and at least one second corresponding interest in the second inspection image corresponding to the second region of interest Comprising a second corresponding partitioning module partitioning into regions,
The second determination module,
Product inspection apparatus, characterized in that by matching each of the second corresponding regions of interest to each of the corresponding second criteria, and determining that the inspected object is defective when there is a second corresponding region of interest that does not satisfy the second criterion .
delete delete delete delete delete

Images