JP3284463B2 - Bottle body defect detection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a defect in a bottle body by imaging, and more particularly to a method suitable for detecting a refractive defect.

[0002]

2. Description of the Related Art Defects in the body of a glass bottle (referred to as a bottle body) are classified into light-blocking defects that block incident light and refractive defects that refract incident light. As a method for detecting a refractive defect as well as a light-shielding defect, the present applicant discloses a method disclosed in Japanese Patent Publication No. 6-90150. In the method of the publication, a light-shielding plate (described as an oblique slit plate) having a plurality of oblique gaps is arranged between the diffused illumination and the bottle,
It scans a striped image formed by light transmitted through the bottle body and compares the brightness of three adjacent points.

[0003]

However, in the method disclosed in the above-mentioned publication, if the thickness of the bottle is large, the stripe pattern may be disturbed and the defect detection may be difficult. Therefore, the present inventor paid attention to a defect called a vertical streak, which is one of the refractive defects, and studied an image processing method of a disturbed stripe pattern. Vertical streaks are linear depressions that appear partially in the longitudinal direction of the bottle. The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a bottle body defect detection method capable of detecting a defect called a vertical streak among the bottle body defects.

[0004]

SUMMARY OF THE INVENTION In order to achieve the above-mentioned object, the present invention obtains an image of a bottle body with a CCD camera using light transmitted through a light shielding plate having a plurality of oblique gaps. In the method of detecting a defect by processing a striped image, a step 1 of obtaining a binary image consisting of light and dark striped patterns in a window of a screen, and scanning the image of the step 1 in the vertical direction of the bottle, Step 2 for obtaining the number of continuous bright or dark pixels on the line (L); and when the number of pixels in step 2 exceeds the set number (K1), the scanning line (L ') preceding by the set number (K2) In the above, the set number (K)
Step 3 for calculating the number of transition points from light to dark or from dark to light between the same start point and end point as pixels continuous beyond 1), and the number of transition points in step 3 is set to a set number (K3 )
The above case is characterized by comprising step 4 of determining that there is a defect.

According to the present invention, an image is taken by a CCD camera,
By processing the obtained striped image, a binary image composed of bright and dark striped patterns is obtained. In the obtained binary image, it is noted that the vertical streak portion is an image in which bright or dark pixels are continuous in the vertical direction. If the calculated number of pixels exceeds the set value (K1), the number of pixels on the scan line (L ') preceding the set value (K2) exceeds the set number on the scan line (L). The number of light-to-dark or dark-to-light transition points between the same start and end points of the pixel is determined. Accordingly, when the number of change points is equal to or more than the set number (K3), it means that the interval between the stripes in the scanning line (L ') is normal, and the vertical stripes in the scanning line (L). It turns out that there is a defect. When the number of change points is less than the set number (K3), the intervals between the stripes of the scanning line (L ') and the scanning line (L) are both normal, and there is no vertical streak defect. I understand.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for detecting a defect in a bottle body according to the present invention will be described below with reference to FIGS. FIG. 1 is a front view showing an inspection apparatus for implementing the method for detecting a defect of a bottle body according to the present invention. As shown in FIG. 1, the inspection apparatus includes a diffused illumination 3 including a light source 3 a and a light diffusion plate 3 b for projecting light onto a glass bottle 2 placed on a turntable 1, a diffused illumination 3 and a glass bottle 2. And a CCD camera 4 disposed on the opposite side of the diffused illumination 3 across the glass bottle 2 and taking an image of the glass bottle 2 from the side (horizontal direction). And an image processing device 5 for processing images captured by the CCD camera 4. The optical axis of the CCD camera 4 is set to be horizontal or substantially horizontal.

FIG. 2 is a view taken in the direction of arrows II-II in FIG. As shown in FIG. 2, a plurality of gaps 10 s in an oblique direction are formed in the light shielding plate 10. In the configuration described above, while the bottle 2 is rotated by the turntable 1, images are taken from the side of the glass bottle 2 by the CCD camera 4 to obtain a large number of screens. In this case, the light from the diffused illumination 3 passes through the gap 10s of the light shielding plate 10 and
Is illuminated, part of the illumination light passes through the glass bottle 2 and enters the CCD camera 4.

FIG. 3 is a view showing a screen obtained by the CCD camera 4. As shown in FIG. 3, when the bottle body is imaged by the CCD camera 4 using the light transmitted through the light shielding plate 10,
A screen 7 having a striped image is obtained. Since the glass bottle 2 has an uneven thickness, a stripe pattern is partially deformed. A window 8 is set at the center of the screen 7, and image data in the window 8 is processed by the image processing device 5. Image processing is performed in the following procedure.

In the image processing, as step 1, a binary image consisting of bright and dark stripes is obtained in a window 8 of a screen 7. 4A and 4B are diagrams for explaining a stage of obtaining a binary image. FIG. 4A shows an original image in a window 8 obtained by a CCD camera, and FIG. 4B shows a binary image obtained from the original image. 3 shows a part of a value image. In the original image captured by the CCD camera, as shown in FIG. 4A, assuming that a light-shielding portion (indicated by a cross in a pixel □) is dark and a light-transmitting portion (indicated only by a pixel □) is bright. The boundary (indicated by oblique lines in the pixel □) has a brightness between light and dark. Therefore, when the brightness data is binarized using a certain brightness as a threshold, a portion brighter than the threshold among the brightnesses of light and dark becomes light (for example, 1), and a dark portion becomes dark (for example, 0). Therefore, an image having an array of 1s or 0s in a matrix is obtained. The change from light to dark is a change from 1 to 0.

FIG. 5 is a diagram showing a binary image obtained by the above-described binarization. In FIG. 5, scanning is performed downward from the origin at the upper right side, and scanning is performed further leftward than the right side. In FIG. 5, diagonal lines indicating darkness are omitted in the middle part. Next, in the image processing, as the step 2, the image of the step 1 is scanned in the vertical direction of the bottle, and the number of continuous light or dark pixels on the scanning line (L) is obtained. In the original image captured by the CCD camera, vertical streaks are pixels in which brightness between light and dark is continuous beyond the normal interval of stripes. In a binary image, as shown in FIG. Depending on the degree of brightness in the original image, bright or dark pixels such as part A or part B are continuous in the vertical direction.

Therefore, if there is a scanning line having pixels continuous beyond the normal interval of stripes, there will be vertical streaks, so that the number of continuous light or dark pixels on the scanning line (L) is reduced. Ask. However, as in the case of the portion C, there are cases where bright pixels continue beyond the normal interval of the stripes as in the case of the portion A, although there are no vertical streaks. Therefore, the image processing is performed in step 3
As shown in FIG. 5, when the number of pixels in step 2 exceeds the set number (K1), the set number of scan lines (L) in the previous scan line (L ') by the set number (K2) Between the start point (Ps) and the end point (Pe), which are the same as the pixels continuous beyond the threshold value, are determined from light to dark or from dark to light.
As a result, in the example shown in FIG. 5, the number of change points of the part A is five, and the number of change points of the part C is two.

Next, in the image processing, as step 4, if the number of change points obtained in step 3 is equal to or larger than the set number (K3), it is determined that there is a defect. In the example shown in FIG. 5, if the set number (K3) is set to 3, it is possible to distinguish the portion A (with defect) from the portion C (without defect). Since the image changes depending on the bottle, the set number is determined based on the results of a number of sample tests and is changed as necessary.

[0013]

As described above, according to the present invention,
Since the characteristic of the vertical streak is extracted in distinction from the disorder of the stripe pattern, the vertical streak defect can be reliably detected.

[Brief description of the drawings]

FIG. 1 is a front view showing an inspection apparatus for carrying out a method for detecting a defect of a bottle body according to the present invention.

FIG. 2 is a view taken in the direction of arrows II-II in FIG. 1, and is a view showing a configuration of a light shielding plate.

FIG. 3 is a diagram showing a screen obtained by a CCD camera.

FIG. 4 is a diagram for explaining a step of obtaining a binary image.
4A shows an original image in a window obtained by a CCD camera, and FIG. 4B shows a part of a binary image obtained from the original image.

FIG. 5 is a diagram showing a binary image obtained by binarization.

[Explanation of symbols]

 Reference Signs List 1 turntable 2 glass bottle 3 diffused illumination 3a light source 3b light diffuser 4 CCD camera 5 image processor 7 screen 8 window 10 light shield 10s gap L, L 'scanning line

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-6-58733 (JP, A) JP-A-63-149547 (JP, A) JP-A-2-49148 (JP, A) 90150 (JP, B2) (58) Field surveyed (Int. Cl. ⁷ , DB name) G01N 21/84-21/958

Claims (1)

(57) [Claims] 1. A method for detecting a defect by taking an image of a bottle body with a CCD camera using light transmitted through a light shielding plate having a plurality of oblique gaps and processing an obtained striped image to detect a defect. Step 1 of obtaining a binary image consisting of light and dark stripes in a window, and Step 2 of scanning the image of Step 1 in the vertical direction of the bottle to obtain the number of continuous light or dark pixels on the scanning line (L). If the number of pixels in step 2 exceeds the set number (K1),
In the scanning line (L ') preceding by the set number (K2), between the same start point and the same end point as the pixels continuing over the set number (K1) on the scan line (L), from light to dark or dark. 3. A bottle body, comprising: a step 3 for calculating the number of transition points from the color to light; and a step 4 for determining that there is a defect when the number of transition points in the step 3 is equal to or more than a set number (K3). Defect detection method.