Inspecting The Bottom Wall Of Hollow Open-ended Containers

Abstract

A method and apparatus for inspecting the bottom wall of hollow open-ended containers wherein a beam of radiant energy to which the container is transparent is directed through the open end of the container to illuminate the bottom wall. The beam is further directed against a screen to produce a darkened image of the defect on the screen. The image of the screen is then analyzed by an image scanning device to produce a reject signal when a defect is present.

Description

This invention relates to a method and apparatus for inspecting hollow containers and articles for defects.

BACKGROUND OF THE INVENTION

In the inspection of hollow articles which have open ends, such as glass containers, for defects such as chips, seeds, or bubbles, inclusions and the like, a major problem that occurs is that certain marks on the surface of the container, such as mold marks and the like, produce false signals even though they may not affect either the strength or appearance of the article and therefore should not be used as a basis for rejecting the article. A further problem in connection with such articles is that it is difficult to make an inspection when the articles are moving at high speeds on the order of 800 to 1,000 containers or articles per minute. Further, various attempts that have been made to make such inspection have necessitated the use of rapidly moving parts that are obviously difficult to align and maintain and produce possible sources of error in the inspection.

Among the objects of the invention are to provide a method and apparatus for effectively inspecting the bottom wall of hollow open-ended articles such as containers for defects such as chips, seeds, and bubbles; wherein the method and apparatus is relatively unaffected by other unobjectionable optical markings such as mold marks and the like; wherein the articles can be inspected while being moved continuously at relatively high speeds; and wherein the apparatus does not require the use of moving parts.

SUMMARY OF THE INVENTION

In accordance with the invention, a beam of radiant energy to which the article is transparent is directed through the open end of the article to illuminate the bottom wall. The beam is further directed against a screen to produce a darkened image of the defect on the screen which is enlarged for some types of defects. Successive portions of the image are then analyzed by an image scanning device to produce a reject signal when a defect is present.

DESCRIPTION OF THE DRAWINGS

FIG. 1 is a part-sectional side elevational view of an apparatus embodying the invention.

FIG. 2 is a schematic diagram of the method and apparatus utilized in the invention.

FIG. 3 is a fragmentary diagram on an enlarged scale of a portion of the diagram shown in FIG. 2.

FIG. 4 is a bottom plan view of the schematic shown in FIG. 3 showing the image of a defect on a screen.

FIG. 5 is a fragmentary part-sectional view taken along the line 2--2 in FIG. 1, parts being broken away.

FIG. 6 is an electrical schematic of the apparatus embodying the invention.

FIG. 7 is a schematic of the scanning portion of the apparatus.

FIG. 8 is a scanning diagram of the portion of the apparatus shown in FIG. 7.

FIG. 9 is a block diagram of the circuitry shown in FIG. 6.

FIG. 10 shows a series of pulse diagrams as they are modified in various portions of the circuitry.

DESCRIPTION

Referring to FIG. 1, the method and apparatus embodying the invention is shown as particularly adapted for use in connection with the manufacture of hollow glass containers C which have an open upper end and are generally bulbous in shape. In one method of manufacture of such containers, they are made on what is commonly known as a glass ribbon machine wherein a ribbon of glass is deposited onto a conveyor made of a plurality of interconnected orifice plates 10 that are moved beneath blow heads and in synchronism with molds to blow and form the glass ribbon into the articles.

In accordance with the invention, the articles C are moved successively and continuously by orifice plates 10 through an inspection station 11 where the bottom wall of the containers is inspected and thereafter to a reject station 12 where a reject device such as a cylinder is operated to reject any containers that have defects in the bottom wall. The defects which are to be detected in the case of glass comprise chips, seeds or bubbles or the like.

The basic method and apparatus for making the inspection at the inspection station 11 is shown diagrammatically in FIG. 2 and comprises a source 13 of radiant energy to which the article is transparent, such as a helium-neon laser, which is emitted in a collimated manner and then caused to form a point image by means of lens 22. The point image of lens 22 is located within the open end of the article or container C to be inspected. From this point image the radiant energy diverges to completely illuminate the plane 15 of the bottom wall of the container. Defects, such as a chip, seed, bubble, inclusion, or the like located in the bottom wall of the container are projected as dark shadows within this diverging beam. In this way the provision of a point source of radiant energy allows the formation of a sharp and distinct shadow of a defect within the illuminated bottom image on screen 15a. The image on the screen 15a is then analyzed for changes in illumination by an image dissector camera which analyzes successive portions of the entire illuminated image as the container moves continuously through the inspection station.

A combination of collecting lens group 23 and an imaging lens group 25 is utilized to form image 15a. This optical arrangement allows the influence of mole marks and other non-critical optical imperfections to be minimized. Without this arrangement of lenses 23 and 25, such non-critical imperfections would also be imaged as dark shadows and could not be distinguished from real defects. Lenses 23 and 25 tend to minimize the relatively smaller images that might be created by optical deviations from mold marks and the like which are not objectionable and are not to be detected.

Defects such as bubbles or seeds, as shown in FIGS. 3 and 4, cause a redirection of light and an enlarged darkened image I of the defect on screen 15a. Such defects produce a lens effect caused by the pocket of gas in the bubble or seed that functions to focus the light and contributes to the enlarged darkened image of this defect type.

A laser source of radiant energy is preferred since it provides sharp, well-defined shadow images of defects, has high output power, long life and is relatively cool.

The specific apparatus for making the inspection is shown in FIG. 5 wherein identical reference numerals are utilized for purposes of clarity. The source of radiant energy comprising laser 13 is mounted in a housing 17a at one side of the path of the articles C and is directed horizontally through a spatial filter 18. The spatial filter is essentially a beam diverging device and a pinhole. A standard high power microscope lens is used for the beam diverger. The pinhole serves to remove interference in the beam caused by stray light diffracted from dust, etc., that is part of any laser optical system. Diffraction interference degrades the laser beam by producing phase and amplitude variations or modulation across the uniphase output of the laser beam. Unless these are removed, Fresnel zone patterns will occur in the laser beam which will represent excessive noise to the electronic system.

The beam then passes through a collimating lens 19 and an iris 20 placed immediately after the collimating lens 19. This provides control of the output beam diameter and directly affects the cone angle of the laser beam which examines the bottle bottom. The size of the iris is adjusted by a knob located on the underside of the top section just behind the diverging lens group 22. This adjustment controls the area of examination of the bottle bottom within the limitations of the acceptance angle of the condensing lens group 23. The beam is then reflected by a mirror 21 downwardly through a lens assembly 22 that produces the diverging beam.

A second housing 17b is positioned at one side and below the path of the containers or articles C and includes a lens assembly 23 which directs the light that has passed through container C against a mirror 24 which, in turn, reflects the light through a lens 25 which images the bottom of the container onto a ground glass screen 15a. Screen 15a is preferably made of red filter glass to prevent stray light from causing unwanted noise in the electronics system. The image dissector camera 16 is mounted within the housing 17b.

Sensors such as magnetic sensors 30 are provided adjacent the inspection station to sense the presence of a support with a container thereon for purposes presently described to synchronize the operation of the electronics associated with the apparatus, which are shown schematically in FIG. 6.

Referring to FIG. 7, the image dissector camera 16 consists of an image dissector tube 31 with its high voltage power supply and a set of imaging optics. The lens forms an image of the target on the photocathode surface 31a of the tube, where the incident photons cause the emission of photo-electrons.

The photocathode surface 31a is maintained at a high negative potential with respect to the anode 32. Therefore, as the photo-electrons are emitted, they are accelerated towards the anode through a focusing coil and deflecting coils 31b. The focusing coil 31b forms an electron image of the photocathode on a rectangular aperture plate 33. The aperture plate 33 is located between the photocathode 31a and the photomultiplier section 34, and only those photo-electrons passing through this aperture 33 are amplified by the photomultiplier.

The output current from the anode 32, representing a highly amplified version of the photo-electron current emitted by the cathode 31a, is monitored by a small resistor in series with the anode lead. The voltage across this resistor, the video output, is buffered by an operational amplifier 34a and directed to the main system. X and Y deflection coils 31b allow any portion of the target image to be monitored by deflecting the photo-electrons from that section through the photomultiplier aperture.

In order for the camera to observe the complete image of the bottle bottom, a scan pattern or raster must be generated which permits each section of the image to be scanned in turn.

Generation of such a pattern (FIG. 8) requires the simultaneous application of suitable drive voltages to the X and Y deflection coils 31b of the image dissector camera 16. It is easily seen that such a pair of waveforms are; a triangular waveform for the vertical drive and a staircase waveform for the horizontal drive. The horizontal scan of the camera will be against the direction of apparent motion of the article or bottle.

A typical image dissector camera which produces satisfactory results is Model F5005 Vidissector Camera with an F4011 tube, manufactured by ITT Industrial Laboratories. The camera is equipped with an S25 photocathode with a 0.005 .times. 020.0 inch aperture. The photocathode has a 1.1 inch diameter active area.

Referring to the block diagram FIG. 9, the signal from oscillator wave shaper divider 35 is fed to an integrator 36 and, in turn, to an analog switch 37 to the vertical drive 38 for the camera.

The oscillator wave shaper divider 35 also supplies a gate 39 and, in turn, a counter 40, and successively a digital analog converter 41, an analog switch 42, and the horizontal drive 43 for the camera. A start pulse begins the start circuitry 44 and operates the synchronizing circuitry 45. A gate pulse actuates "get ready" circuitry 46 controlling analog switches 37 and 42. A three-element band pass circuit including a band pass circuit 48, a high pass circuit 49 and a large defect circuit 54 is provided. Threshold circuitry 47, actuated by counter 40, operates band pass circuit 48 and together with high pass circuit 49 provides signals to a comparator 50 and, in turn, to a defect detector 51 and through a gate 52 to a defect storage 53 for subsequent rejection. The gate 52 is controlled by the large defect circuit 54.

The corresponding elements of the block diagram (FIG. 9) are shown in broken lines in FIG. 6.

Referring to the schematic diagram FIG. 6, the oscillator wave shaper divider 35 comprises a clock 55 and flip flops 56, 57. Integrator 36 comprises an amplifier 58. Analog switch 37 comprises switch 59 while vertical buffer amplifier and gain adjuster 38 comprise an amplifier 60 and associated circuitry. The gate circuit 39 comprises gates 61, 62. Analog switch 42 comprises switch 63 and horizontal buffer amplifier and gain adjusting circuit 43 comprises an amplifier 64.

The start circuitry 44 comprises gates 65, 66 and a flip flop 67. The threshold circuitry 47 comprises count detector 68, gate 69 and threshold adjuster 70. The get ready circuit 46 comprises gates 71 and flip flop 72. Band pass circuit 48 comprises amplifiers 73, 74 and high pass circuit 49 comprises an amplifier 75. Comparator 50 comprises an amplifier 76. Defect detector 51 comprises a gate 77, one-shots 78, 79 and gate 80. Defect storage 53 comprises gate 81, gate 82 and flip flop 83. Large defect circuit 54 comprises amplifiers 84, 85, 86 and a gate 87.

Referring to the schematic diagram FIG. 6, the output pulses from 20 KHz clock 55 are shaped by flip flop 56 and the frequency is halved. The frequency is halved again by flip flop 57 and the output is fed to an integrator 36 formed by amplifier 58. The output of amplifier 58 is a 5 KHz triangular wave, symmetrical about zero volts, the result of integrating the 5 KHz square wave from flip flop 57. Assuming that analog switch 37 is closed, the signal from amplifier 58 is amplified and buffered by amplifier 60 for application to the vertical drive input of the camera 16.

The output of flip flop 56 is fed via gates 61 and 62 to seven bit ripple counter 40. Digital to analog converter 41 follows the counter 40, and its output is a positive going staircase starting at zero volts when the counter 40 is in the zero state. The output of the digital to analog converter 41 is summed by amplifier 43 with an adjustable bias and inverted, thus generating a bipolar, negative going staircase waveform for application to the horizontal drive input to the camera 16.

In operation, an article coming down the line initiates a get ready pulse through a remote sensor 30 positioned on the line which sets flip flop 72. The output of flip flop 72 closes switch 37 which starts the vertical scan and opens a switch 42 which removes the bias from amplifier 43, thus driving the scan to its extreme horizontal position.

The output of the counter 40 was set to zero at the end of the previous raster and the digital to analog converter 41 output is also zero. When the article is in position at the inspection station, a "start" pulse is initiated by sensor 30 which sets flip flop 67, opening gate 61 and allowing clock pulses from flip flop 56 into the counter 40.

Thus, the camera scans the image vertically and steps across horizontally each time the count increases by one. When the preset number of pulses have been counted, the end of count detector 68 resets the counter and also the start and get ready flip flops 67, 72. As soon as flip flop 72 is reset, it opens switch 37, removing the vertical drive signal and closing switch 42, reapplying the bias to amplifier 43, thus setting its output to zero which places the scanning aperture in the middle of the field.

Referring to FIG. 6, the video signal is fed simultaneously to three channels. The first channel 48 consists of amplifier 73 which together with the network following it has a specific band pass transfer function and amplifier 74 which level shifts the conditioned video signal and applies it to one side of differential comparator 50. The second channel 49 is formed by amplifier 75 and the network at its output which has specific high pass transfer function.

Differential comparator 50 compares the outputs of amplifiers 74 and 75 and triggers one-shots 78 and 79 via gate 77 if a difference (defect) occurs. The third channel 54 is formed by amplifiers 84 and 85 which feed the video signal to comparator 86 which provides a positive output when the input goes more negative than a preset reference level.

The output of comparator 86 provides a means of blanking the defect logic via gate 52 between scans. However, if a very large defect should occur which happened to be more positive than the reference level, it would cause false blanking to occur. One-shots 78, 79 overcome this problem by generating a defect output in the form of a pulse 1 .mu.s wide but delayed by 2 .mu.s. This allows any large defect spike time in which to disappear after, which the defect signal can pass through the gate and set defect flip flop 83 which in turn signals the data logger. Gate 82 controlled by the start flip flop 67 blanks the defect logic between bottles.

Examination of a typical trace shows that the sensitivity of the system is a maximum in the center of the scan pattern and decreases or falls off toward the edges. This is due to the combined effect of a Gaussian laser output, the imaging optics, and uneven camera sensitivity. In order to compensate for these effects, two different methods are employed in the detection circuitry. One method acts to give improved uniformity of detection as the scan format steps across the bottle bottom image. The other applies to give an adaptive threshold for each scan line. Each method acts in an orthogonal direction to the other. The total result is an improved uniformity of detection sensitivity across the bottle area investigated, even for conditions of nonuniform illumination or uneven camera sensitivity.

For improved detection sensitivity in the scan step direction the electronics signal at the output of amplifier 74 is weighted, the weight being controlled by the scan count. Essentially the threshold adjustment circuit monitors the counter output and delivers a preset bias to amplifier 74 whose value depends on the actual scan count. To accomplish this, the pulses from counter 40 are fed to the threshold adjuster 70 which furnishes a changing bias voltage between selectable counts. For example, the scan pattern may be divided into five segments for weighing purposes. These may be counts 0 to 12, 12 to 24, 24 to 40, 40 to 52 and 52 to 64, but could in general be extended to include each scan increment if desired.

In the direction along each scan line, another approach is employed. The pulse representing a given defect is higher when it appears near the center of the sweep and lower near the ends. To provide more uniform detection sensitivity across the scan sweep, an adaptive threshold is utilized. The incoming video signal is processed in two separate stages, one stage represents the defect information and the other serves to provide an adaptive threshold. The defect information is processed by selective filtering and differentiation at the output of amplifier 73 and is then inverted by amplifier 74 and sent to comparator 76. In the comparator, the processed defect signal from the band pass circuit 48 is referenced to the output of amplifier 75 of the high pass circuit 49. The reference signal is generated by filtering the video input to decrease its high frequency content and is then adjusted in gain to represent the general signal envelope across the scan. This envelope serves as the adaptive threshold which duplicates the general contour of the video signal. In this way, the signal variation across the scan which is due to the nonuniform illumination and uneven camera photocathode sensitivity is employed to give greater detection sensitivity where the defect signal amplitude is lower at each end of the scan. This results in an overall improvement in uniformity of detection sensitivity. The manner in which the signal is modified is shown in FIG. 10.

Claims

We claim:

1. An apparatus for inspecting the bottom wall of hollow open-ended articles for defects which comprises

means for moving the articles successively and continuously in a predetermined path past an inspection station,

means at the inspection station for directing a beam of radiant energy to which the article is transparent against the bottom wall of the article,

a screen positioned adjacent the bottom wall of the article for creating an image of the bottom wall on the screen with an image of the defect present in the bottom wall,

and means for analyzing successive portions of said image to determine the presence of a defect.

2. The combination set forth in claim 1 including means for creating a reject signal when said analyzing means detects the presence of a defect.

3. The combination set forth in claim 1 wherein said image creating means comprises a lens system interposed between said articles and said screen.

4. The combination set forth in claim 3 wherein said lens system comprises collecting lens means and imaging lens means.

5. The combination set forth in claim 1 wherein said means for directing a beam of radiant energy includes a laser.

6. The combination set forth in claim 1 wherein said means for creating an image includes means for collimating said beam and thereafter creating a divergent beam.

7. The combination set forth in claim 1 wherein said means for directing a beam includes a source positioned adjacent the path of the article at one side thereof and mirror means for redirecting the beam into the open end of the article.

8. The combination set forth in claim 1 wherein said means for analyzing said image includes collecting lens means.

9. The combination set forth in claim 1 wherein said means for analyzing said containers includes an image dissecting camera.

10. The combination set forth in claim 1 including lens means interposed in the path of said image between said screen and said analyzing means for minifying the image of said screen.

11. The combination set forth in claim 1 wherein said means for analyzing said image is positioned at one side of the path of the container, and mirror means for redirecting the image from said screen toward said analyzing means.

12. The combination set forth in claim 1 wherein said means for analyzing said image comprises raster scanning means and means for compensating for the effect that the scan pattern signals are a maximum in the center of the scan pattern and decrease toward the edges by increasing the sensitivity of the scan pattern signals toward the edges of the scan.

13. The combination set forth in claim 12 wherein said compensating means functions to count the scan pattern and modify the scan pattern in accordance with the count.

14. The combination set forth in claim 12 including means for producing a threshold signal that has a greater magnitude adjacent the center of the scan pattern than at the edges and means for comparing the scan pattern signal with the threshold signal to produce a reject signal when any scan pattern signal exceeds the threshold signal.

15. The combination set forth in claim 14 wherein said comparing means comprises filtering means for filtering the scan pattern signal and means for adjusting the gain thereof.

16. An apparatus for inspecting the bottom wall of hollow open-ended articles for defects which comprises

means for directing a beam of radiant energy to which the article is transparent against the bottom wall of the article,

a screen positioned adjacent the bottom wall of the article to create an image of the bottom wall with an image of the defect present in the bottom wall,

and means for analyzing successive portions of said image to determine the presence of a defect.

17. The combination set forth in claim 16 including means for creating a reject signal when said analyzing means detects the presence of a defect.

18. The combination set forth in claim 16 wherein said image creating means comprises a lens system interposed between said article and said screen.

19. The combination set forth in claim 18 wherein said lens system comprises collecting lens means and imaging lens means.

20. The combination set forth in claim 16 wherein said means for directing a beam of radiant energy includes a laser.

21. The combination set forth in claim 16 wherein said means for creating an image includes means for collimating said beam and thereafter creating a divergent beam.

22. The combination set forth in claim 16 wherein said means for directing a beam includes a source positioned adjacent the path of the article at one side thereof and mirror means for redirecting the beam into the open end of the article.

23. The combination set forth in claim 16 wherein said means for analyzing said image includes collecting lens means.

24. The combination set forth in claim 16 wherein said means for analyzing said containers includes an image dissecting camera.

25. The combination set forth in claim 16 including lens means interposed in the path of said image between said screen and said analyzing means for minifying the image of said screen.

26. The combination set forth in claim 16 wherein said means for analyzing said image is positioned at one side of the path of the container, and mirror means for redirecting the image from said screen toward said analyzing means.

27. The combination set forth in claim 16 wherein said means for analyzing said image comprises raster scanning means and means for compensating for the effect that the scan pattern signals are a maximum in the center of the scan pattern and decrease toward the edges by increasing the sensitivity of the scan pattern signals toward the edges of the scan.

28. The combination set forth in claim 27 wherein said compensating means functions to count the scan pattern and modify the scan pattern in accordance with the count.

29. The combination set forth in claim 27 including means for producing a threshold signal that has a greater magnitude adjacent the center of the scan pattern than at the edges and means for comparing the scan pattern signals with the threshold signal to produce a reject signal when any scan pattern signal exceeds the threshold signal.

30. The combination set forth in claim 29 wherein said comparing means comprises filtering means for filtering the scan pattern signals and means for adjusting the gain thereof.

31. The method of inspecting the bottom wall of hollow open-ended articles which comprises

moving the articles successively and continuously in a predetermined path past an inspection station,

directing a beam of radiant energy to which the article is transparent through the open end of the container against the bottom wall of the container as the container moves continuously through the inspection station to illuminate the bottom wall,

causing a defect in the bottom wall to redirect a portion of the beam,

creating an image of the bottom wall whereby a defect produces a localized change in the illumination of the image,

analyzing said image,

and producing a reject signal when said analysis evidences the presence of a defect,

said step of analyzing said image comprising analyzing successive portions of said image to produce scan pattern signals,

the step of analyzing successive portions of the image is by a raster scan and including the step of compensating for the effect that the scan pattern signals are a maximum in the center of the scan pattern and decrease toward the edges by increasing the sensitivity of the scan pattern signals toward the edges of the scan,

the step of producing a threshold signal that has a greater magnitude adjacent the center of the scan pattern than at the edges and comparing the scan pattern signals with the threshold signal to produce said reject signal when any scan pattern signal exceeds the threshold signal.

32. The method set forth in claim 31 wherein said comparing step is achieved by filtering the scan pattern signals and adjusting the gain thereof.

33. The method of inspecting the bottom wall of hollow open-ended articles which comprises

moving the articles successively and continuously in a predetermined path past an inspection station,

directing a beam of radiant energy to which the article is transparent through the open end of the container against the bottom wall of the container as the container moves continuously through the inspection station to illuminate the bottom wall,

causing a defect in the bottom wall to redirect a portion of the beam,

creating an image of the bottom wall whereby a defect produces a localized change in the illumination of the image,

analyzing said image,

and producing a reject signal when said analysis evidences the presence of a defect,

said step of analyzing saidimage comprising analyzing successive portions of said image to produce scan pattern signals,

the successive portions of said image being scanned transversely to the direction of movement of the articles.

34. The method of inspecting the bottom wall of hollow open-ended articles which comprises

directing a beam of radiant energy to which the article is transparent through the open end of the container in a diverging path against the bottom wall of the container,

causing a defect in the bottom wall to redirect a portion of the beam,

creating an image of the bottom wall whereby a defect produces a localized change in the illumination,

and producing a reject signal when said localized change evidences the presence of a defect,

said step of analyzing said image comprising analyzing successive portions of said image to produce scan pattern signals,

the step of analyzing successive portions of the image is by a raster scan and including the step of compensating for the effect that the scan pattern signals are a maximum in the center of the scan pattern and decrease toward the edges by increasing the sensitivity of the scan pattern signals toward the edges of the scan,

the step of producing a threshold signal that has a greater magnitude adjacent the center of the scan than at the edges and comparing the scan pattern signals with the threshold signal to produce said reject signal when any scan pattern signal exceeds the threshold signal.

35. The method set forth in claim 34 wherein said comparing step is achieved by filtering the scan pattern signals and adjusting the gain thereof.