JP2002048723A - Method and device for inspecting flaw in the vicinity of edge - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inspection method and its device for surely detecting a flaw in the vicinity of an edge without misdetection of a flaw which does not actually exist in the edge. SOLUTION: This inspection method comprises a photographing process of taking a picture of an object for providing a photographed image, an edge direction differentiation process of performing differentiation in the edge direction of the object in the photographed image for generating an edge direction differentiation image, and a flaw detection process in the vicinity of the edge of detecting a flaw in the vicinity of the edge based on the edge direction differentiation image. This method is applicable to the flaw detection device in the vicinity of the edge.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention belongs to the technical field of automatically inspecting the appearance and surface condition of an article. In particular, the present invention relates to a method and an apparatus for detecting a defect near an edge which can detect a defect near the edge of an article.

[0002]

2. Description of the Related Art There is known an apparatus for performing inspection by performing image processing based on a captured image of an article to be inspected. At that time, it is necessary to perform imaging in consideration of the positioning accuracy and the like in the imaging of the article so that the article does not come off the captured image. Generally, imaging is performed such that an image slightly larger than the size of the article is obtained. In the captured image, an inspection range is designated by an operator's instruction input, or image processing is performed to determine the article and automatically set the inspection range. Image processing for inspection is performed only for the inspection range.

[0003] A plastic card will be described as an example of an article. Defects in the appearance and surface state of the plastic card include scratches, dents (dents), and protrusions (peaks). Since this defect occurs at an unspecified portion of the plastic card, it is necessary to inspect the entire surface of the plastic card.

[0004]

However, if an edge of a plastic card is included in the above-described inspection area, it is erroneously detected that a defect that does not actually exist exists at the edge. This is because the pixel value changes discontinuously at the boundary (edge) between the plastic card portion and the background portion in the captured image. So, conventionally,
By setting the inspection area inside the outer diameter of the plastic card, erroneous detection is prevented. Therefore, naturally, a defect generated near the edge cannot be detected.

The present invention has been made to solve the above problems. An object of the present invention is to provide an inspection method and apparatus capable of reliably detecting a defect existing near an edge without erroneously detecting a defect that does not actually exist at the edge.

[0006]

The above object is achieved by the present invention described below. That is, the inspection method according to claim 1 of the present invention includes an imaging step of imaging an article to obtain a captured image, and an edge direction of performing differentiation in an edge direction of the article in the captured image to generate an edge direction differential image. The method includes a differentiation process and a near-edge defect detection process of detecting a defect near an edge based on the edge direction differential image.

According to the present invention, an article is imaged in an imaging process to obtain a captured image, and in the edge direction differentiation process, differentiation is performed in the edge direction of the article in the captured image to generate an edge direction differential image, and In the defect detection process, a defect near the edge is detected based on the edge direction differential image. That is, in the captured image, the pixel value changes discontinuously at the boundary (edge), but in the edge direction differential image generated by performing differentiation in the edge direction, the pixel value does not change discontinuously. Moreover, the defect can be emphasized in the edge direction differential image as compared with the captured image. Accordingly, there is provided an inspection method capable of reliably detecting a defect existing in the vicinity of an edge without erroneously detecting a defect that does not actually exist in the edge.

[0008] The inspection method according to claim 2 of the present invention comprises:
2. The inspection method according to claim 1, wherein a plurality of edge direction differential images are generated in accordance with a plurality of edge directions of the article in the edge direction differentiation step, and the plurality of edge direction differential images are generated in the edge vicinity defect detection step. A defect near the plurality of edges is detected based on the detected defect. ADVANTAGE OF THE INVENTION According to this invention, the defect which exists in the edge vicinity of several directions in an article is detected.

Further, the inspection method according to claim 3 of the present invention,
3. The defect inspection method according to claim 1, wherein an arrangement direction of pixels in the captured image matches an edge direction of the article. According to the present invention, differentiation in the edge direction of an article in a captured image can be performed with high accuracy with a small differential error. as a result,
Even a defect located very close to the edge is detected.

[0010] The inspection method according to claim 4 of the present invention comprises:
In the edge vicinity defect inspection method according to any one of claims 1 to 3, the article is a plastic card. According to the present invention, a defect near the edge of a plastic card is detected.

Further, the inspection apparatus according to claim 5 of the present invention comprises:
Imaging means for imaging an article to obtain a captured image; edge direction differentiating means for differentiating the edge direction of the article in the captured image to generate an edge direction differential image; and an edge neighborhood based on the edge direction differential image And edge near-edge defect detection means for detecting a defect in (1).

According to the present invention, the article is imaged by the image pickup means to obtain a picked-up image, and the edge direction differentiating means performs differentiation in the edge direction of the article in the picked-up image to generate an edge direction differential image. A defect near the edge is detected by the defect detection means based on the edge direction differential image. That is, in the captured image, the pixel value changed discontinuously at the boundary (edge),
In the edge direction differential image generated by performing differentiation in the edge direction, the pixel value does not change discontinuously. Moreover, the defect can be emphasized in the edge direction differential image as compared with the captured image. Therefore, there is provided an inspection apparatus capable of reliably detecting a defect existing in the vicinity of an edge without erroneously detecting a defect that does not actually exist in the edge.

[0013]

Next, an embodiment of the present invention will be described. FIG. 1 shows an example of a data processing process in the inspection method and apparatus according to the present invention. FIG. 2 is an explanatory diagram showing a process of detecting a defect existing near an edge in the inspection method and apparatus according to the present invention. Figure 1,
A description will be given based on FIG. First, in step S1 of FIG. 1, an article to be inspected, for example, a plastic card is imaged to obtain a captured image (imaging process).

An example of an imaging method will be described. The plastic cards are stacked and placed on the supply unit, and the supply unit separates the plastic cards in that state one by one and supplies them to the imaging unit. At this time, positioning is performed so as to be a predetermined position in the imaging unit. The imaged plastic cards are stacked and placed in the discharge unit.

As the imaging means, any of a line sensor camera having a one-dimensional imaging area and an area sensor camera having a two-dimensional imaging area can be used. In the line sensor camera, imaging is performed while transporting a plastic card. In area sensor cameras, a plastic card is kept stationary, or an electronic shutter,
The image is taken while being substantially still by flash light emission or the like.

As shown in FIG. 2, an image is picked up so as to include a plastic card 11 and a background 12. In addition, the illuminating unit, the imaging unit, and the like are arranged so that defects, such as a defect, a dent (a dent), and a projection (a pit), which are defects in the appearance and surface condition of the plastic card, are captured in the captured image 10. Done. Further, imaging is performed such that the arrangement direction of the pixels in the captured image 10 matches the edge direction of the plastic card 11.

The ordinary plastic card 11 has a rectangular shape whose sides are orthogonal to each other. Also, a normal captured image 1
At 0, the pixels in the captured image 10 are arranged in a grid in the top-left direction. Therefore, as shown in FIG. 2, the side (edge) of the plastic card 11 in the up-down direction coincides with the up-down direction of the captured image 10, and the left-right side (edge) of the plastic card 11.
The imaging is performed so that the left and right directions of the captured image 10 coincide with each other.

The arrangement direction of the pixels in the picked-up image 10 and the edge direction of the plastic card 11 are matched by the positioning of the plastic card 11 in the image pickup unit. Although it is difficult to make this degree of agreement perfect, this degree of agreement is improved. Thereby, the differentiation in the edge direction of the article in the captured image 10 described later is
It can be performed with high accuracy with a small differential error. As a result, it is possible to detect even a defect that is very close to the edge.

Next, in step S2 of FIG. 1, the portion of the plastic card 11 to be inspected is
Extract from 0 (test object extraction process). For example, binarization is performed based on a difference in pixel value between the plastic card 11 and the background 12 in the captured image 10 to separate the plastic card 11 portion and the background 12 portion in the captured image 10.

Next, in step S3 of FIG. 1, an inspection area in the captured image 10 is set (inspection area setting). For example, the entire area of the plastic card 11 in the captured image 10 is set as the inspection area. Further, for example, a plurality of inspection areas are set in order to set different inspection levels depending on the portion of the plastic card 11.

Next, in step S4 of FIG. 1, the vicinity of an edge in the captured image 10 is set (edge vicinity area setting). The area near the edge is determined in the next step (S
This is the area where edge direction differentiation is performed in 5). This area includes not only the part of the plastic card 11 set as the inspection area, but also the part of the background 12 not set in the inspection area. FIG. 2 shows an edge near area 1 for one of the edges of the plastic card 11 in the vertical direction.
3a, and an edge vicinity region 13b for one of the edges in the left-right direction is indicated by a broken line.

Next, in step S5 of FIG. 1, differentiation is performed in the edge direction of the plastic card 11 in the captured image 10 to generate an edge direction differential image (edge direction differentiation process). In step S2, a plurality of edge direction differential images are generated according to a plurality of edge directions of the article to be inspected. As shown in FIG. 2, in the plastic card 11, the edge direction is
There are a total of four in the left and right directions. However, in FIG. 2, in order to avoid complication, an edge direction differential image 20a for one of the top and bottom edges of the plastic card 11 and an edge direction differential image 20b for one of the left and right edges are shown. Is illustrated.

Here, the edge vicinity area 13 in FIG.
a is compared with the edge direction differential image 20a. The difference in pixel value between the plastic card 11 and the background 12 (the difference in shading in FIG. 2) and the boundary, which existed in the edge vicinity region 13a, disappeared in the edge direction differential image 20a. This is the result of edge direction differentiation,
This is because the difference between pixel values in the edge direction and the direction perpendicular to the edge direction disappears, and only the difference in pixel value in the edge direction is emphasized. The same applies to the comparison between the edge vicinity area 13b and the edge direction differential image 20b.

Next, in step S6 of FIG. 1, a defect near an edge is detected based on the edge direction differential images 20a and 20b (edge near-edge defect detection process). For example, binarization is performed using a predetermined threshold to separate a defective portion in the edge direction differential images 20a and 20b from a normal portion.

FIG. 2 shows a detected defect 24a in the edge direction differential image 20a for the defect 14a near the top and bottom edges of the plastic card 11, and an edge direction differential image 20b for the defect 14b near the left and right edges. Are shown in FIG. As shown in FIG. 2, the defects 14a and 14b
Are differentiated in the edge direction, so that the defect 14a and the defect 1
As an image in which the edge in the specific direction (the direction perpendicular to the edge direction) of FIG.
4b is obtained.

Next, in step S7 in FIG. 1, in the inspection area (see step S3) set in the picked-up image 10, other defects other than the defect near the edge are detected (other defect detection process). In addition, in the defect detection process, a known inspection method can be applied. For example, a Laplacian differential image is obtained in an inspection area where the vicinity of an edge is set as a non-inspection area. The Laplacian differential image is binarized by a predetermined threshold to separate a defective portion from a normal portion.

For example, a reference image, which is a captured image of the plastic card 11 serving as a reference, is compared with an inspection target image, which is a captured image of the plastic card 11 to be inspected, and the vicinity of the edge is set as a non-inspection area. Defect detection is performed from the difference in the inspected area.

Next, in step S8 of FIG. 1, the quality of the plastic card 11 is determined by integrating the detected edge vicinity defect and other defects (overall quality determination process). At this time, the number of defects, the size of the defect, the position of the defect, and the like can be considered.

Next, in step S9 of FIG. 1, it is determined whether or not the inspection has been completed for all the plastic cards 11 to be inspected. If it has been completed, the inspection process is completed. If the inspection has not been completed, the process returns to step S1, and the subsequent steps are repeated for the next plastic card to be inspected.

[0030]

According to the inspection method of the first aspect of the present invention, a defect that does not actually exist at an edge is not erroneously detected, and a defect existing near the edge can be reliably detected. it can. Further, according to the inspection method according to the second aspect of the present invention, it is possible to detect defects existing near edges in a plurality of directions in the article. Further, according to the inspection method of the third aspect of the present invention, it is possible to detect even a defect existing very near an edge. Further, according to the inspection method of the fourth aspect of the present invention, it is possible to detect a defect near the edge of the plastic card. Further, according to the inspection apparatus of the fifth aspect of the present invention, it is possible to reliably detect a defect existing near the edge without erroneously detecting that a defect that does not actually exist at the edge exists.

[Brief description of the drawings]

FIG. 1 is a diagram showing a process of data processing in an inspection method and apparatus according to the present invention.

FIG. 2 is an explanatory diagram illustrating a process of detecting a defect existing near an edge in the inspection method and apparatus according to the present invention. DESCRIPTION OF SYMBOLS 10 Captured image 11 Plastic card 12 Background 13a, 13b Edge vicinity area 14a, 14b Defect 20a, 20b Edge direction differential image 24a, 24b Detection defect

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Yasutaka Fujii 1-1-1 Ichigaya-Kaga-cho, Shinjuku-ku, Tokyo F-term in Dai Nippon Printing Co., Ltd. (reference) 2G051 AA73 AB07 AC01 CA03 CB01 DA01 DA05 EA08 5B057 AA01 BA02 CA12 CB12 CE20 CH08 DA03 DB02 DC16 DC32

Claims (5)

[Claims] 1. An imaging process of imaging an article to obtain a captured image, an edge direction differentiation process of differentiating an edge direction of the article in the captured image to generate an edge direction differential image, and the edge direction differential image An edge vicinity defect detection step of detecting a defect near an edge based on the following. 2. The inspection method according to claim 1, wherein a plurality of edge direction differential images are generated in accordance with a plurality of edge directions of the article in the edge direction differentiation step, and the plurality of edge direction differential images are detected in the edge vicinity defect detecting step. An inspection method comprising: detecting a defect near the plurality of edges based on an edge direction differential image. 3. The inspection method according to claim 1, wherein an arrangement direction of pixels in the picked-up image coincides with an edge direction of the article. 4. The inspection method according to claim 1, wherein the article is a plastic card. 5. An image pickup means for picking up an article to obtain a picked-up image, an edge direction differentiating means for differentiating an edge direction of the article in the picked-up image to generate an edge direction differential image, and the edge direction differential image. And an edge near-edge defect detecting means for detecting a defect near the edge on the basis of the above.