JP5372612B2 - Glass product inspection equipment - Google Patents

Description

  The present invention is a glass product inspection for inspecting the presence or absence of defects such as scratches, bubbles, foreign matter, joint lines, defective wall thickness, and defective shape before slowly cooling a glass product molded by a molding machine in a slow cooling furnace. Relates to the device.

Inspection of defects such as scratches, bubbles, foreign matter, joint lines, defective thickness, and defective shape of glass products such as glass bottles is performed with visible light after being gradually cooled in a slow cooling furnace (so-called cold end).
However, such cold-end inspection is about an hour after the glass product is molded, so if a defect is caused by a gob molding or a modulation during glass molding, the modulation is recognized. However, there is a problem that a time lag of at least about 1 hour occurs, and many defective products are manufactured.

  If such a defect inspection can be performed on the upstream side of the slow cooling furnace (so-called hot end), if the defect is caused by modulation during gob molding or glass product molding, the countermeasures can be taken several minutes after molding. This can be done later, and can prevent wasteful production of defective products.

  In the following patent document, before still cooling a still-hot glass product molded by a molding machine in a slow cooling furnace, the thermal radiation emitted from the glass product is captured by an infrared sensor, and the thickness of the product is measured. Techniques for finding defects are disclosed.

  Patent document 1 measures the intensity | strength of the infrared rays radiated | emitted from a glass product in two different wavelengths, and calculates | requires the thickness of a glass product quantitatively from the intensity | strength of the infrared rays. That is, a relationship (function) between the thickness and the infrared intensity of two wavelengths is obtained in advance, and the thickness is obtained by applying the measured infrared intensity at the two wavelengths to an equation.

  Patent Document 2 records the distribution (temperature distribution) of infrared rays radiated from glass products, statistically processes the temperature of each part of the glass product that is continuously molded, and forms molding abnormalities in the gob and molding machine Anomalies are discovered.

JP 2000-205827 A JP 2005-534904 A

The above-mentioned patent document discloses that the thickness of a product is measured by infrared rays radiated from a still hot glass product, or a molding abnormality of a gob molding or a molding machine is discovered.
Therefore, replace the conventional cold end visible light inspection with infrared rays emitted from still hot glass products, and perform hot end inspection for defects such as scratches, bubbles, foreign objects, joint lines, wall thickness defects, and shape defects. Can be considered.
However, even if the conventional camera of the inspection device using visible light is replaced with an infrared camera and inspection at the hot end is performed, erroneous determination is serious and it is not practical.

  An object of the present invention is to enable a glass product manufactured by a molding machine to be accurately inspected for defects such as scratches, bubbles, and foreign matters at a hot end before entering a slow cooling furnace.

As a result of studying a phenomenon in which misjudgment increases in an inspection using infrared rays at a hot end, the present inventors have found that the cause is mainly due to the following two causes.
The first cause is that the intensity of infrared rays emitted from the contour portion of the still hot glass product is unstable due to various factors. Even if the image taken with the infrared camera is simply compared with the reference image to determine pass / fail, the non-defective product is erroneously determined to be defective or vice versa due to the instability of the infrared intensity at the contour. Cheap.
The second cause is that glass products molded by the molding machine are transported in a line on the conveyor, but when the glass products are imaged with an infrared camera, the glass products of the next (before and after on the conveyor) Thermal radiation is reflected on a glass product to be imaged, and this reflection tends to cause a non-defective product to be erroneously determined as a defective product, or vice versa.
The present invention has been completed in view of these points.

[Claim 1 ]
The present invention has an infrared camera that images a glass product molded and conveyed by a molding machine in front of a slow cooling furnace, and processing means for processing the captured image, and the processing means is a glass that is currently inspected. Create a difference image between the current image of the product and the past image of the glass product that was captured in the past, find the number of pixels that deviate from the predetermined brightness threshold in the difference image, and the number of pixels exceeds the predetermined pixel number threshold in Ruga Las product inspection apparatus it is determined to be defective if the said previous image, characterized in that an average image of a plurality of non-defective glass products of an image captured immediately before the glassware under test It is a glass product inspection device .

  The difference image is an image created by using the difference between the brightness of each pixel of the current image and the brightness of each corresponding pixel of the past image as the brightness of each pixel.

  The difference image that is the difference between the current image and the past image is an image in which the reflection of the heat radiation of the adjacent glass product is almost deleted, and in this difference image, the pass / fail judgment is performed by the brightness threshold value and the pixel number threshold value. The pass / fail judgment that is hardly affected by the reflection can be made, and the accuracy of the pass / fail judgment is improved.

[Claim 2 ]
The present invention, the current image and the past image is an inspection apparatus according to claim 1 which is the image in the inspection area set in the inside of the outline of an image captured by an infrared camera.

The inspection area is set inside the contour of the image captured by the infrared camera, and the pass / fail judgment is made based on the brightness threshold and the pixel count threshold in the inspection area. The accuracy of pass / fail judgment is improved without affecting the result.

The reflection of the next glass product changes from moment to moment due to various factors (for example, changes in mold temperature and misalignment of the supply position of the gob to the mold), so the immediately preceding glass product is as close as possible to the glass product being inspected. If the image is a past image, the influence of the reflection can be eliminated more accurately.
In addition, by setting the past image as an average image of two or more images, if there is a specific image, the influence can be reduced.

  An average image is an image in which the average brightness of each corresponding pixel of a plurality of images is configured as the brightness of that pixel.

[Claim 3 ]
Further, the present invention is a glass product molding machine in which the molding machine has a plurality of sections, and the past image is an average image of images of a plurality of non-defective glass products in the same section taken immediately before the glass product under inspection. The inspection apparatus according to claim 1 or 2.

  If the molding machine has multiple sections, as in the so-called IS molding machine of glass bottles, the temperature of the glass product differs depending on the molded section, so the past image was molded in the same section as the glass product being inspected. By adopting the product image as the past image, it is possible to accurately eliminate the influence of the reflection, and it is possible to make a quality determination with high accuracy.

[Claim 4 ]
The present invention, the infrared camera is an inspection apparatus according to any one of claims 1 to 3, intended to capture infrared wavelength of 0.8～1.7Myuemu.

In the present invention, a commercially available infrared camera that captures infrared rays having a wavelength of 0.8 to 1.7 μm can be used.
By inspecting with infrared rays having this wavelength, it becomes possible to inspect for defects in the shape of the inner surface of the glass product and defects in colored glass such as black and green.

[Claim 5 ]
The present invention has a rejecter downstream of the infrared camera, there glassware said processing means has determined to be defective by the inspection in the inspection apparatus according to any one of claims 1 to 4 eliminated by the rejector .

By the rejector, defective products are automatically removed from the conveyor.
As the rejector, a well-known type such as a type in which glass products on a conveyor are blown away with air or a type to extrude can be used.

[Claim 6 ]
Further, the present invention is the inspection apparatus according to any one of claims 1 to 5 , wherein the processing unit includes a storage unit, and records an image captured by the infrared camera in the storage unit.

[Claim 7 ]
Further, the present invention is the inspection apparatus according to claim 6 , wherein the processing unit includes a display unit, and an image recorded in the storage unit is displayed on the display unit.

  By recording captured images in storage means (memory, hard disk, etc.), these images can be called up on display means (CRT, liquid crystal monitor, etc.) as necessary to analyze the cause of defective products. Can be used.

[Claim 8 ]
The present invention, the molding machine is a glassware forming machine having a plurality of sections, according to claim 7, wherein the processing means it is history defective displayed on said display means together with the section number that is molded Inspection equipment.

  By displaying the history of defective products together with the section number in which they were molded, for example, if there are many defective products in a product molded in a specific section, it is considered that something abnormal has occurred in that section, and you should deal with it Can do.

[Claim 9 ]
Further, the present invention is a glassware molding machine in which the molding machine has a plurality of sections, the processing means divides an image captured by the infrared camera into a plurality of areas, obtains an average temperature for each area, and inspects The inspection apparatus according to claim 7 or 8 , wherein a history of temperature information is displayed on the display means for each section of the glass product for each section that has been completed.

  For example, if the temperature of the neck of the product molded in the 10th section is displayed for multiple glass bottles (graph display etc.) and there is little variation in the temperature, molding in that section is smooth and there are many variations In such a case, it can be determined that some abnormality has occurred, or that an abnormality may occur soon, and this can be dealt with.

The present invention can accurately inspect defects in glass products at a so-called hot end.
In addition, the information acquired in the inspection can be used for analyzing the gob molding process and the defective operation of the molding machine.
Inspection results and analysis results at the hot end can be promptly fed back to the gob molding process and molding machine to improve them, so that many defective products are not formed easily and the glass product manufacturing efficiency is improved.

It is a schematic explanatory drawing of a glass product shaping | molding process. It is a schematic explanatory drawing of the test | inspection apparatus of an Example. It is explanatory drawing of the image imaged with the infrared camera. It is explanatory drawing of a present image, a past image, and a difference image. It is explanatory drawing of a defective article log | history display screen. It is explanatory drawing of the average temperature history of a certain area | region of the glassware shape | molded by a certain section.

FIG. 1 shows an outline of a hot end of glass product (glass bottle) molding by an IS molding machine.
In this case, the IS molding machine 6 has ten sections from the first section to the tenth section (circle numbers in FIG. 1 represent section numbers), the first section being the most downstream side of the conveyor, the tenth section. Is the most upstream side.
The glass product G formed in each section of the molding 6 is sent out on the conveyor 7 as indicated by the arrow in FIG. 1 and is conveyed in a row toward the downstream side, and is put into the slow cooling furnace 8 by the pusher 9. It is sent.
The glass product is coated (so-called hot-end coating) in the coating chamber 10 as necessary before entering the slow cooling furnace.

In the inspection apparatus of the present invention, the infrared camera 1 is installed so as to take an image of the glass product G being conveyed on the conveyor 7 between the IS molding machine 6 and the slow cooling furnace 8.
Since the environment in which the infrared camera 1 is installed is high temperature, it is desirable to cool the camera by storing it in a cooling box (not shown).
On the downstream side of the infrared camera 1, a rejecter 3 and a discard unit 5 are provided. The rejector 3 blows glass products determined to be defective toward the disposal unit 5 and excludes them from the conveyor 7.

FIG. 2 shows an outline of the configuration of the inspection apparatus according to the embodiment.
The inspection apparatus includes an infrared camera 1, a processing unit 2, a rejecter 3, a display unit 4, and an input unit.
The infrared camera 1 is a commercially available product that captures infrared rays having a wavelength of 0.8 to 1.7 μm, and a controller is attached.
The processing means 2 is a commercially available computer (personal computer), to which a display means 4 (monitor) and input means (keyboard, mouse) are connected.

The control system attached to the IS molding machine 6 has a section signal, a timing signal, and a reset signal.
The section signal is a signal representing a section number (including a station symbol when there are a plurality of stations in a section), which is output every time a bottle formed in each section is sent out on the conveyor.
The timing signal is a pulse signal synchronized with the operation speed of the IS molding machine. For example, 1800 pulses are output while completing one cycle of operation (while completing molding once in all sections).
The timing of shear cutting for cutting the gob, the operation of all parts of the molding machine, the conveyance speed of the conveyor 7 and the like are synchronized with this timing signal.
The reset signal is one pulse output every time the IS molding machine 6 completes one cycle of operation.

  The processing means 2 receives a section signal, a timing signal, and a reset signal from the control system of the IS molding machine 6 so that a certain glass product G is transported to the front of the infrared camera 1 and the shutter is released. Which section (and station) is formed, and when the glass product G is defective, can be grasped when it is conveyed to the front of the rejector 3 to operate the rejector.

  When the glass product G comes to the front of the infrared camera 1, the processing means 2 sends an imaging signal to the controller of the infrared camera. As a result, the infrared camera releases the shutter and sends the image data to the processing means 2.

FIG. 3 shows the imaged glass product G, and the thick solid line is the contour 11.
The processing means 2 sets the inspection area 12 inside the contour 11.
The inspection area is determined in advance so as to exclude the contour portion where the infrared radiation is unstable in consideration of the shape of the glass product. In this case, the glass product diameter is reduced to about 90% in the horizontal direction. In some cases, it may be a region contracted only in the vertical direction or horizontally and vertically.

[Judgment method 1 : Reference example ]
The processing means 2 obtains the number of pixels that deviate from the predetermined brightness threshold in the inspection area 12, and determines that the product is defective when the number of pixels exceeds the predetermined pixel number threshold.
The brightness threshold value can be determined for each area by dividing the inspection area into a plurality of areas. FIG. 3 shows an example in which the inspection area is divided into five areas, that is, the mouth, neck, upper trunk, lower trunk, and bottom.
The brightness threshold value can be determined as follows, for example. When the brightness gradation is 14 bits (2 to the power of 14 = 16384) and the average brightness of the upper part of the cylinder is 4000, the threshold value is within the range of 10%, and the number of pixels outside 3600-4400 is the defective portion. Is determined. Other areas can be determined similarly.
The pixel number threshold value may be arbitrarily determined depending on how many defects are present as non-defective products.

[Judgment Method 2 : Example ]
The “current image” in the upper left of FIG. 4 is an image of the inspection area 12 of the glass product set in the same manner as described above. Reflections 13 of adjacent glass products (front and rear glass products on the conveyor 7) appear slightly bright on the left and right sides.
The “past image” in the upper right of FIG. 4 is an average image of the inspection area of two glass products molded in the same section taken immediately before the glass product in the current image. The average image is created by averaging the brightness of the corresponding pixels of a plurality of (in this case, two) images.
The “difference image” at the bottom of FIG. 4 is created by subtracting the brightness of the corresponding pixels of the current image and the past image. In the difference image, the reflection 13 is almost erased, and only the portion of the defect 14 is displayed brightly.
The processing unit 2 obtains the number of pixels that deviate from the predetermined brightness threshold in the difference image, and determines that the product is defective when the number of pixels exceeds the predetermined pixel number threshold.
The brightness threshold value in this case is brightness that does not determine that the normal part is a defective part, for example, the brightness gradation is 14 bits (2 14 = 16384), and is 500, and deviates from 0 to 500. The number of pixels is determined as a defective part.
The pixel number threshold value may be arbitrarily determined depending on how many defects are present as non-defective products.

In the present invention, it is possible to inspect for defects such as scratches, bubbles, foreign matters (such as brick pieces of a kiln), joint lines (those with a strong finish joint line), defective thickness, and poor shape.
Scratches, bubbles, foreign matter, and joint lines are all brightly displayed on the captured image because the infrared radiation of the portion becomes strong.
In the case of poor thickness (local thickness disturbance), the thick part is bright and the thin part is dark.
In the case of a shape defect (such as a dent or protrusion), the portion or its periphery becomes bright.
The inspection apparatus of the present invention detects such a disturbance in the brightness of the defective portion, and determines that the defective product is defective when the degree of the defect exceeds a predetermined level.

When the processing means 2 determines that the glass product G is defective, the processing means 2 sends the reject signal to the rejector 3 at the timing when the glass product is conveyed to the front of the rejector 3 to operate the glass product G. To eliminate.
You can also sound an alarm when there is a major defect. This is because a second pixel number threshold value that is considerably larger than the normal threshold value is provided as the pixel number threshold value, and an alarm is issued when the number of pixels that deviate from the brightness threshold value exceeds the second pixel number threshold value. Just make it sound.

The image data captured by the inspection apparatus of the present invention can be used for analysis of whether or not the glass product molding process is normally performed.
FIG. 5 shows an example of a screen displayed by the processing unit 2 by sending a display signal to the display unit 4.
In this screen, the history of glass products determined to be defective is shown in order from the top to the bottom. The history shows the section number (section where the glass product was molded), the date of inspection, and the time. When each defective product is marked with a circle in “Original Image”, the original image is displayed in the original image display column on the left side. The number of “section number” represents a section. In this case, there are two stations in one section, and two glass products are formed simultaneously in one section. Section numbers “A” and “B” represent molded stations.
In this way, when the history of defective products is displayed on the display means together with the section number where the defective product is molded, it is possible to know the molding abnormality of the section.
For example, in the case of FIG. 5, since many defective products are generated with the section number “1A”, it is understood that there is some abnormality in the A station in the first section. A mark is put on the “original image 3” formed with the section number “1A”, and the image is displayed in the “original image display” portion. Thus, the specific state of the defect of a glass product can be confirmed visually.

  In the screen of FIG. 5, it is possible to further display areas where defective products are defective (mouth, neck of FIG. 3). For example, the number of pixels that deviate from a predetermined brightness threshold may be totaled for each area, and the area with the largest number of pixels may be displayed as “an area where a defect has occurred”.

  FIG. 6 is an example in which a history of the average temperature of the upper part (FIG. 3) of each glass product formed in the 10th section B station is displayed. In FIG. 6, the newer glass product (formed later) is shown on the left side. In the upper part of FIG. 6, since the variation in the average temperature of the upper part of the glass product is large, it can be seen that there is an abnormality in molding at the 10th section B station. As a result of the inspection, it was found that the gob did not fall to the center of the rough mold, and as a result of improving it, as shown in the lower part of FIG. Thereby, it turns out that the shaping | molding abnormality in the 10th section B station was eliminated.

  FIG. 6 shows an example of displaying the temperature information history at the B station in the tenth section, but the temperature information history at any station in any section can be displayed.

  In this way, the image captured by the infrared camera is divided into a plurality of areas, the average temperature for each area is obtained, and the history of the temperature information is displayed for each area of the glass product for each section that has been inspected. By displaying the symbol, it becomes easy to find the molding abnormality of the section (and station).

DESCRIPTION OF SYMBOLS 1 Infrared camera 2 Processing means 3 Rejector 4 Display means 5 Disposal part 6 IS molding machine 7 Conveyor 8 Slow cooling furnace 9 Pusher 10 Coating chamber 11 Contour 12 Inspection area 13 Reflection 14 Defect G Glass product

Claims (9)

An infrared camera that images a glass product that is molded and conveyed by a molding machine in front of the slow cooling furnace, and a processing means that processes the captured image,
The processing means creates a difference image between the current image obtained by imaging the glass product currently being inspected and the past image of the glass product taken in the past, and obtains the number of pixels that deviate from a predetermined brightness threshold in the difference image. in Ruga Las product inspection apparatus be determined to be defective if the number of pixels exceeds a predetermined pixel number threshold value, the past image is an image of a plurality of non-defective glass product picked up just before the glassware under test A glass product inspection device characterized by being an average image of the above . The inspection apparatus according to claim 1 , wherein the current image and the past image are images in an inspection region set inside an outline of an image captured by the infrared camera.   The molding machine is a glassware molding machine having a plurality of sections, and the past image is an average image of a plurality of non-defective glassware images of the same section taken immediately before the glassware under inspection. 2. The inspection apparatus according to 2. The infrared camera, inspection apparatus according to any one of claims 1 to 3, intended to capture infrared wavelength of 0.8～1.7Myuemu. The inspection apparatus according to any one of claims 1 to 4 , further comprising: a rejector on a downstream side of the infrared camera, wherein the glass product that is determined to be defective by the processing unit is excluded by the rejector. Inspection device according to any one of claims 1 to 5, wherein said processing means comprises a memory means, for recording an image captured by the infrared camera in the storage means. The inspection apparatus according to claim 6 , wherein the processing unit includes a display unit, and the image recorded in the storage unit is displayed on the display unit. 8. The inspection apparatus according to claim 7 , wherein the molding machine is a glass product molding machine having a plurality of sections, and the processing means displays a history of defective products on the display means together with a section number where the molding is formed. The molding machine is a glass product molding machine having a plurality of sections, and the processing means divides an image captured by the infrared camera into a plurality of sections, obtains an average temperature for each section, and finishes the inspection. The inspection apparatus according to claim 7 or 8 , wherein a history of temperature information is displayed on the display means for each area of each glass product.

Images