JPH0454899B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to inspecting and screening defects in the sidewalls of transparent containers, and in particular to inspecting and screening defects in refractive sidewalls such as air bubbles or opaque sidewalls such as stones. The present invention relates to the selective sorting of glass containers having defects detected in the following manner.

Various types of defects can occur in the manufacture of glass containers. To this end, optical scanning techniques have conventionally been proposed for inspecting containers with defects that affect the optical transmission properties of the side walls of the container. U.S. Patent No. 4378493, U.S. Patent No. 4378494, and U.S. Patent No.
No. 4,378,495 discloses a method and apparatus for transporting glass containers through multiple stations for physical and optical inspection. In one optical inspection station, a glass container is held vertically and rotated about its central axis. 1
A light source emits light energy that is diffused through the side walls of the container. A camera having a plurality of light sensitive elements (hereinafter referred to as pixels) linearly arranged parallel to the axis of rotation of the container is arranged to observe the light transmitted through the vertical stripes of the side wall.
The output of each pixel is sampled at increments of vessel rotation, and a rejection signal is generated if the difference in magnitude of the signals from adjacent pixels exceeds a predetermined threshold level. Appropriate rejection signals are thus generated and defective containers are separated from the conveyor line.

The method and apparatus described in the above US patent is commonly referred to as a sidewall inspection device (hereinafter simply referred to as SID) and is considered to be very effective and efficient for automatic inspection and classification of glass containers. However, there are some problems in using this SID for detecting certain types of defects. For example, when inspecting for refractive defects in the direction perpendicular to the axis of rotation of a transparent container (hereinafter referred to as ribbon tears),
A glass container defect detection device has been proposed. In this defect detection device, a diffused light source with a filter is directed at the side wall of the container, and this light source emits light having an intensity gradient approximately parallel to the axis of the transparent container to the photosensitive area of the camera. irradiate. The light intensity is detected and a defect signal is generated as a function of the magnitude of the light intensity gradient at a pixel within the camera. Defective containers are then separated from the conveyor line. This technique has been successfully used for the reliable and efficient detection of lateral ribbon tear defects and for the classification of containers containing such defects.

Differentiating between defect type and size using SID is similarly problematic. For example, SID uses a wide range of light energy sources, the energy range of which is wide enough that most refractive defects do not refract light enough to be visible as dark spots in the bright background of the wide light source. . In other words, SID is effective for opaque defects, but generally ineffective for refractive defects, since opaque defects absorb light energy so they can be seen as dark spots on a bright background. It is. The present invention addresses the problem of detecting opaque defects and at the same time also enables the detection of refractive defects by optical processing, and allows for the detection of small refractive defects such as small air bubbles in the container side wall and large refractive defects such as large air bubbles. It is intended to enable differentiation between defects and defects. In the manufacture of glass containers, it is important to be able to detect refractive defects, such as large air bubbles, which propagate initial cracks and lead to container failure, as well as detect opaque defects. On the other hand, refractive defects such as small air bubbles in the sidewalls are acceptable in the market and should not be excluded. These problems are particularly important in the manufacture of tapered neck containers such as beer bottles. Segregating and rejecting containers with market-acceptable defects is inappropriate and increases manufacturing costs. Therefore, there is a need for a technique for reliably detecting unacceptable defects while distinguishing between acceptable defects, and classifying and eliminating only those defects.

It is, therefore, an object of the present invention to utilize the techniques disclosed in the aforementioned U.S. patents and patent applications to provide an economical and easily distinguishable sidewall defect characteristic and dimension, and to pass through the defect within an acceptable range. It is possible to effectively separate containers that are outside the allowable range while
It is an object of the present invention to provide a method and apparatus for the inspection and sorting of transparent containers, especially glass containers. It is therefore or another object of the invention to provide characterization of individual defects.

According to the present invention, a transparent container is rotated around its axis, and a predetermined function is applied to the side wall of the transparent container to vary the intensity of light depending on the position in a direction perpendicular to the rotation axis of the container. Therefore, transparent containers are inspected and sorted by irradiating them with diffused light with varying light intensity. A camera is positioned to monitor the energy of light passing through the sidewall in a narrow strip-shaped field of view parallel to this axis of rotation. The type of defect in the sidewall is determined as a function of the magnitude of the signal generated by the camera pixels, and the size of the defect is determined as a function of the position of that signal. Containers with unacceptable defects, such as large air bubbles or stones, are rejected and automatically separated from acceptable containers.

In detail, the intensity of the light directed through the container and towards the camera is controlled as a function of the position of the diffuse light source, spread laterally on either side of the center of illumination, so that providing first and second outer regions each having a substantially uniform light intensity and a third central region between the outer regions having a diffused light intensity different from the light intensity of the outer region; . The light intensities of these outer regions may be different from each other or at approximately the same level. If the intensities in these outer regions are at different levels, the intensity in the central region varies uniformly, linearly or logarithmically between these levels, or is approximately uniform in intensity between these levels. In a preferred embodiment, the intensity in the first and second regions is at approximately equal levels, and that in the central region is also approximately uniform, but at a different level than in the outer regions.

The present invention will be explained below based on the drawings.

FIG. 1 is a schematic plan view of a container inspection system 10 including a star wheel 12 coupled to a drive hub 14 for counterclockwise step rotation as indicated by arrow 16. The feed conveyor 18 includes a driven screw 20 for feeding a transparent container 22, such as a rigid glass container, around the periphery of the wheel 12. The wheels 12 are adapted to carry the container 22 along an arcuate path through a plurality of inspection stations, only one of which is shown at 24. The discharge conveyor 26 includes a driven screw 28 adapted to receive containers 22 from the periphery of the star-shaped wheel 12 after each container has been moved by the star-shaped wheel 12 through a plurality of inspection stations. A plunger 30 is coupled to a solenoid 32 and positioned to block the passage to the discharge conveyor 26 when activated. When an out-of-tolerance defect is detected in a particular container at any of these inspection stations, the plunger 30 activates the solenoid 3.
2 to prevent the passage of the conveyor 26 as the container rotates out to a position adjacent the outlet of the conveyor 26, thereby allowing defective containers to be conveyed by the wheels 12 to the reject discharge chute 34. do.

The present invention relates to an examination station, designated station 24. Station 24 includes a drive roller 36 coupled to container 22 and arranged to rotate it counterclockwise while holding it in an axially fixed position by idler roller 38. A light source 40 is positioned within the periphery of the wheel 12 below the plane of the hub 14 to provide diffused light radially to illuminate the entire height and width of the sidewall of the container 22 while the drive roller 36 rotates the container 22 about its axis. outwardly oriented. A camera 42 is positioned radially outward of the light source 40 and wheel 12 by a bracket 44. Camera 42 includes a plurality of light sensitive elements or pixels 43 (third
256 pixels 43 are preferably arranged in a linear array parallel to the axis of rotation of the container 22 in the station 24 and aligned with that axis and the vertical centerline of the light source 40. A lens 46 images a vertical strip-like area of the side wall of the container onto this array of elements. Inspection system 10 as described above
and the testing station are similar to those disclosed in the aforementioned US patent.

This light source 40 consists of a plurality of incandescent lamps arranged in three rows parallel to the axis of rotation of the container. A diffuser plate is placed between the lamp and the container to allow diffused light from a relatively wide range of light sources to pass through the side walls of the container and into the lens 46 and camera 42.
Sampling electronics compare signals from adjacent pixels in the linear array and provide an abnormal signal to the information processor when the difference between the signals is greater than a predetermined threshold level. The processing device thereby operates the solenoid 32 and plunger 30 as the associated defective container moves to the exit position of the inspection station, whereby the defective container is sorted into the rejection chute 34 as described above.
As mentioned above, such inspection and screening techniques, although effective, have problems in distinguishing between defect types and sizes. The present invention aims to solve these problems.

2 and 3 are electro-optical and functional block diagrams of a preferred embodiment of the present invention at inspection station 24. FIG. A diffuser plate 48 and an intensity filter plate 50 are disposed within the light source 40, shown in top view in FIG. 1 and in side view in FIG. pass through the camera 42.

The sampling circuit 54 receives signals from each pixel 43 (FIG. 3) in the camera 42 and provides a serial signal string to the information processing device 56. Container 22 is rotated in the direction of arrow 57 at a speed that is controlled as a function of the rotational speed of inspection system 10 to ensure that camera 42 provides a predetermined number of scans, such as 300 scans per container. The sampling within circuit 54 can be adjusted accordingly. Processor 56 issues an anomaly signal when the magnitude of signals from adjacent pixels in one scan differs by more than a predetermined threshold. By "adjacent pixels" is intended individual elements within the linear array of cameras 42 that are physically proximate to each other. Processor 56 performs defect continuity analysis by evaluating the location of multiple anomalous signals to determine whether a defect is present. Based on this analysis, the processing device 56
(FIG. 2), the operation of the plunger 30 is controlled to classify defective containers.

FIG. 3 is a side view showing the relationship of filter plate 50 and camera 42 thereto at line 3--3 of FIG. According to the invention, a filter plate 5 is used for detecting and differentiating the size and type of side wall defects.
The optical density of 0 spans the width SW of the light source 40, i.e. it is varied transversely to the axis of rotation of the container to create an anisotropic light intensity distribution along its lateral width. . In the particular embodiment described herein, the optical density of the filter plate 50 includes at least two laterally adjacent illumination areas 60, 62;
Preferably three such areas 58, 60, 62
across the width SW of the light source 40. Adjacent areas 58 and 60 and 60 and 62
have different optical densities. The centerline of camera 42 is aligned with the centerline of central region 60 between outer regions 58 and 62. If there is no container 22 between the lens 46 and the light source 40, or if a container with no sidewall defects is rotated between the lens and the light source,
The intensity of light entering camera element 43 is primarily a function of the optical density at the centerline of central region 40. However, when a refractive defect, such as a bubble, moves in front of the camera, such defect refracts the field of view of pixel 43 within camera 42 toward outer region 58 or 62. The present invention utilizes controlled variations in lateral light intensity for differentiation of defect type and size.

Figures 4A and 4B, which have the same horizontal axis, illustrate one embodiment of the invention. FIG. 4A is a diagram similar to FIG. 3, and FIG. 4B illustrates the optical density of filter plate 50a (FIG. 4A) as a function of lateral position. 4A and 4B, the optical attenuation density within the outer region 58a of the filter plate 50a is substantially zero or transparent;
The attenuation density of region 62a is approximately 100%, ie, opaque. In the central region 60a, the optical source extinction density varies linearly from substantially zero at the lateral edges continuous with the outer regions 58a to substantially 100% at the edges continuous with the regions 62a.

Sampling circuit 54 receives signals from each pixel 43 in camera 42 and provides signals from each pixel sequentially to generate one column of pixel signals during each scan of the linear array of camera 42 at time t. A bubble defect 64 is shown within the field of view of pixel 43n during several scans from time t ₂ to t ₆ . The horizontal axis of FIG. 4C shows the scan time increments t ₁ -t ₇ and relates them to actual bubble position. Thus the scan at t ₁ is the bubble 64
occurs just before the bubble 64 rotates into the field of view of pixel 43n, and the scan at t ₇ causes the bubble 64 to rotate into the field of view of pixel 43n.
Occurs immediately after exiting the field of view. The vertical axis of FIG. 4C shows the actual light intensity seen by pixel 43n for each scan as bubble 64 passes through the field of view of pixel 43n. The sampling circuit 54 provides a sequence of pixel signals for each scan to a processing unit 56, which detects an abnormal signal when the difference in magnitude between adjacent pixel signals during the scan, such as from pixels 43n and 43m, is greater than a predetermined threshold. occurs. FIG. 4D, which has the same horizontal axis as FIG. 4C, shows the corresponding difference signal generated by processor 56 and used to determine the presence of an anomaly. The anomaly occurs at a specific location on the side wall of the container. This abnormal position is defined by the scanning signal giving an indication of the angular position of the container and the pixel signal corresponding to the longitudinal position along the sidewall, as described above. In the present invention
SID also analyzes the continuity of defects by analyzing the locations of multiple anomalies to determine whether or not there are defects.

In operation, the light intensity detected by pixel 43n changes as its field of view is refracted from the centerline of central region 60a by bubble 64.
Scanning begins at scanning time t1 just before the bubble 64 enters the field of view of the pixel 43, and the intensity detected by the pixel 43n at that time is at about the 50% level since its field of view is at the center line of the central region 60a. Thus, the corresponding difference signal at scan time t1 is essentially zero since the intensities detected by adjacent pixels 43n and 43m are substantially the same. However, at scan time t2, the leading edge of bubble 64 moving in direction 66 with respect to stationary filter 50a and camera 42 is at pixel 43.
n field of view and detected by pixel 43n is at a level of approximately 100% because the leading edge of bubble 64 refracts the field of view of pixel 43n into optically substantially transparent outer region 58a. The corresponding difference signal at scan time _t2 has a substantially positive value because the intensity detected by pixel 43n is substantially greater than that by pixel 43m.
At time _t3 , the leading edge of the bubble 64 is located at the pixel 43n.
further into the field of view of pixel 43, the intensity detected thereby becomes weaker due to the refractive properties of its leading edge.
Since the field of view of n is refracted only to the extent that it becomes a portion closer to the outside of the central region 60a, it increases only to about a 75% level. Thus, the difference signal at time _t3 is positive because the intensity due to pixel 43n is still greater than that due to pixel 43m. bubble 4
At time _t4 , just before the refractive leading edge of bubble 64 exits the field of view of pixel 43n, the intensity detected by pixel 43n is such that the essentially parallel planes of bubble 64 no longer refract the field of view of pixel 43n. It will return to about 50% level. Therefore, the difference signal at time _t4 becomes 0. Scan time t ₅
At , the trailing edge of bubble 64 begins to enter the field of view of pixel 43n. The detected intensity is reduced to about 25% level because the refractive characteristic of the trailing edge is not so great and refracts the field of view of the pixel 43n only to the extent of the outer part of the central region 60a. Therefore, the difference signal at time _t5 has a negative value because the intensity detected by pixel 43n is somewhat less than that by pixel 43m. At scan time _t6 , the trailing edge of bubble 64 is completely within the field of view of pixel 43n, so the detected intensity is zero because the trailing edge refracts the field of view into outer region 62a, which is substantially opaque. Therefore, the difference signal at that time has a large negative value because the intensity detected by pixel 43n is much smaller than that by pixel 43m. The trailing edge of the bubble 54 is the pixel 43
The intensity detected at scanning time _t7 after exiting the field of view of pixel 43n becomes approximately 50% level again because the field of view of pixel 43n is not refracted by the defect and coincides with the center line of central region 60a. Thus, the difference signal at scan time _t7 is between pixels 43n and 43m.
Since the detected intensities are substantially the same, they become 0.

As described above, the difference signal associated with a bubble or other refractive defect provides a computer-recognizable signature from a positive peak value to a negative peak value. Processor 56 provides an indication of the lateral dimension of bubble 64 by counting the number of scans between the positive and negative peak values of the difference signal corresponding to the leading and trailing edges of bubble 64, respectively. The size of the bubble 64 is determined by using a continuity analysis of . The processor 56 is programmed to issue a reject signal 32 when the number of scans exceeds a predetermined threshold indicating the presence of large bubbles, so that containers with large bubbles are rejected and containers with small bubbles are accepted. Not excluded as being within the scope.
This analysis can be used to obtain the same feature if the bubble 64 is large enough to block the field of view of several pixels 43. Similar features can be obtained for each pixel that varies only in lateral length.

The present invention not only provides SID with the ability to detect refractive defects by optical manipulation and even distinguish between small and large refractive defects;
It also retains the ability to detect opaque defects. Opaque defects absorb light energy and can be seen in the present invention as dark spots on a "gray" background. Opaque defects inherently create different signal characteristics than refractive defects. In Figures 5A-5D, an opaque defect or stone 68 is shown within the field of view of pixel 43n. That is, the light intensity detected by pixel 43n changes as its field of view is blocked by stone 68 from the centerline of central region 60a. At scan time _t2 , when the leading edge of stone 68 enters the field of view of pixel 43n, the intensity detected thereby is zero because the stone is opaque. Therefore, the difference signal at time _t2 has a large negative value because the intensity detected by pixel 43n is much smaller than that by pixel 43m. The same thing happens at time t ₃
−applies to the intensity due to pixel 43n at t ₆ . The features created by stones 68 are therefore easily distinguishable from those created by air bubbles 64. In other words, bubbles produce positive and negative peak values, while stones produce only negative peak values. Thus, in this embodiment, the SID can detect opaque defects, and at the same time, optical processing can detect refractive defects and distinguish their sizes.

Figures 6A-6D and 7A-7D, which are similar to Figures 4A-4D and 5A-5D, respectively, illustrate other embodiments of the invention. 6th A, 6B
In the illustration, the optical attenuation density in outer region 58b of filter 50b is substantially zero or transparent, and that of outer region 62b is substantially 100% or opaque. In the central region 60b, the optical density is approximately 50%.

Sampling circuit 54 receives signals from each pixel 43 in camera 42 and serially applies signals from each pixel 43 to generate a column of pixel signals during each scan of the linear array of camera 42 at time t. Air bubble 64 is shown within the field of view of pixel 43n during the scan at times t ₂ -t ₆ . The horizontal axis of FIG. 6C shows the scan time increments t ₁ -t ₇ and relates them to the actual bubble position. Thus,
In the scanning at time _t1 , the bubble 64 is located at the pixel 43n.
The scan at time _t7 occurs just after bubble 64 leaves the field of view of pixel 43n. The vertical axis of FIG. 6C indicates that the bubble 64 is the pixel 43n.
shows the actual light intensity seen by that pixel for each scan as it passes through the field of view. Sampling circuit 54 provides one column of pixel signals for each scan to processing unit 56, which detects when the difference between adjacent pixel signals, e.g., signals from pixels 43n and 43m during one scan, exceeds a predetermined threshold. sends out an abnormal signal. FIG. 6D, which has the same horizontal axis as FIG. 6C, shows the difference signal generated by processor 56 and used to determine the presence of an anomaly. The anomaly occurs at a specific location on the side wall of the container. This position is limited by the number of scans at the angular position of the container and the number of pixels corresponding to the longitudinal position along the side wall of the container, as described above. Here, SID performs defect continuity analysis by analyzing the locations of multiple anomalies to determine the presence or absence of defects.

The light intensity detected by pixel 43n changes as its field of view is refracted from the centerline of central region 60b by bubble 64. bubble 64 is pixel 4
Starting from scanning time t ₁ just before entering the field of view of 3n,
The intensity due to pixel 43n has a field of view of area 60.
Since it coincides with the center line of b, the level is about 50%.
Thus, the difference signal at time t ₁ is at pixel 43
Since the intensity due to n and 43m are almost the same, it becomes 0. However, at time t ₂ the filter 50
If the leading edge of the bubble 64 enters the field of view of the pixel 43n while moving in the direction 66 with respect to the camera 42, the intensity detected thereby is such that the leading edge of the bubble essentially optically blocks the field of view of the pixel 43n. Since the light is refracted to the transparent outer region 58b, the level is about 100%.
The difference signal at this time has a large positive value because the detected intensity by the pixel 43n is considerably larger than that by the pixel 43m. Air bubble 6 at scanning time t ₃
When the leading edge of 4 further enters the field of view of pixel 43n,
The resulting strength drops to the 50% level. Therefore, the difference signal at that time is approximately zero. This is the same at times _t4 and _t5 . At time t ₆ , bubble 6
When the trailing edge of pixel 43n completely enters the field of view of pixel 43n, the intensity becomes zero as the field of view is refracted into the substantially opaque outer region 62b. Therefore, the difference signal at this time has a large negative value because the difference signal due to the pixel 43n is much smaller than that due to the pixel 43m. After the trailing edge of the bubble exits the field of view of pixel 43n at scan time _t7 , the detection intensity is at about the 50% level because the field of view of pixel 43n coincides with the center line of central region 60b rather than the defect. The difference signal at this time is pixels 43n and 43m.
Since the detection intensities of are almost the same, it becomes 0. Fourth
As in the case of Fig. A-4D, the processing device 56 removes the bubbles 6
Determine the size of bubble 64 using defect continuity analysis, which indicates the lateral dimension of the bubble by counting the number of scans between the positive and negative peaks of the difference signal corresponding to the leading and trailing edges of 4, respectively. do. The processor 56 is programmed to issue an exclusion signal 32 when the number of scans exceeds a predetermined threshold indicating the presence of large bubbles, so that containers with large bubbles are excluded and those with small bubbles are not excluded. Make it. Similar features can be obtained using this analysis if the bubble is large enough to span several pixels 43. A similar feature can be obtained for each pixel that changes only in lateral length.

7A-7D, an opaque defect or stone 68 is shown within the field of view of pixel 43n. The light intensity detected by the pixel 43n has a field of view of stone 68.
changes when it is blocked from the center line of central region 60b by. When the leading edge of stone 68 enters the field of view of pixel 43n at scanning time _t2 , the detection intensity by that pixel becomes zero. At this time pixel 4
Since the signal due to 3m is larger than that due to 43n, the difference signal has a large negative value. This is time t ₃ −
The same applies to _t6 . The features created by stones 68 are easily distinguishable from those created by air bubbles 64. That is, bubbles are represented by positive peak values and negative peak values, while stones are represented by only negative peak values. This embodiment uses SID to enable detection of opaque defects and detection of refractive defects including their size.

8A similar to Figures 4A-4D and 5A-5D
Figures -8D and 9A-9D show other embodiments of the present invention. 8A-8D, the optical attenuation density in the outer regions 58c, 62c of filter 50c is approximately zero, ie, transparent. In the central region 60c it is approximately 50%.

In this embodiment, the intensity detected by the pixel 43n is such that its field of view is in the central region 60c due to the air bubble 64.
changes when refracted from the center line of bubble 6
Scanning time just before 4 enters the field of view of pixel 43n
Starting from t ₁ , the field of view of the pixel 43n coincides with the center line of the central region 60c, so that the detection intensity is at about the 50% level. Therefore, the difference signal at this time is approximately 0 because the detected intensities by the pixels 43n and 43m are approximately the same. A time t ₂ when the leading edge of the bubble enters the field of view of pixel 43n while moving in the direction of arrow 66 relative to filter 50c and camera 42.
In this case, the field of view of the pixel 43n is refracted to the transparent outer region 58c by the leading edge, so that the field of view is approximately 100%.
At this time, the detection intensity of pixel 43n is considerably greater than that of 43m, so the difference signal has a large positive value. At scanning time t ₃ -t ₅ when the leading edge of bubble 64 enters further into the field of view of pixel 43n, the detected intensity of pixel 43n is at about the 50% level and the difference signal is approximately zero. When the trailing edge of the bubble enters the field of view of pixel 43n at scanning time _t6 , it refracts the field of view into the transparent outer region 62c, so that the detection intensity is
Return to 100%. The difference signal at this time has a large positive value. When the trailing edge of bubble 64 leaves the field of view of pixel 43n at time _t7 , pixel 4
Since the field of view 3n is not refracted by a defect but coincides with the center line of the central region 60c, it becomes the 50% level, and the difference signal at that time becomes almost 0.

The difference signal associated with a bubble or other refractive defect thus creates a computer-processable feature consisting of a pair of positive peak values. Processor 56 similarly determines the size of bubble 64 using defect continuity analysis by counting the number of scans between the positive peak values of the difference signal corresponding to the leading and trailing edges of bubble 64.
The processor is programmed to issue a reject signal 32 when the number of scans exceeds a predetermined threshold indicating the presence of large bubbles, thereby causing only containers with large bubbles to be rejected. The case where the bubble spans several pixels 43 is similar to the previous embodiment.

In Figures 9A-9D, an opaque defect or stone 68 is shown within the field of view of pixel 43n. The intensity detected by pixel 43n changes as its field of view is blocked by stone 68 from the centerline of central region 60c. Stone 6 at scanning time t ₂
When the leading edge of the pixel 43n enters the field of view of the pixel 43n, the detection intensity by the pixel 43n becomes zero. The difference signal at this time has a large negative value. Scanning time t ₃ − t ₆
The same is true for . Therefore, the characteristics of the stone consist only of negative peak values, whereas the bubbles 64 consist of positive peak values, so they can be easily distinguished. This embodiment allows SID to detect not only opaque defects but also refractive defects and their sizes.

Thus, the present invention allows not only the detection of opaque and refractive defects, but also the differentiation of these defects both in terms of type and size. Also, bubbles 64 in Figures 4A, 6A, and 8A and bubbles 64 in Figures 5A and 7
The stones in Figures A and 9A are shown as having the same size for comparison. Further, the varying defect attitude of embodiments of the present invention may be reflected, for example, in areas 58 and 58 of FIG.
changing the widths D, W, E of 60, 62; changing the dimension C between the centerline of the camera and the continuous edges of regions 58, 60; and/or changing the total effective width SW of the light source 40. The scanning increment may be changed and controlled by. Each filter area 58, 60, 6
2 preferably has a uniform attenuation concentration in the longitudinal direction, ie perpendicular to the axis of the container. However, such concentration may be varied in the longitudinal direction if desired in accordance with the content of the aforementioned US patent application. For containers with narrow necks (Figure 2), the width W of the intermediate region opposite the neck
may increase with increasing neck diameter.

The sampling circuit 54 and the information processing device 56 (FIG. 2) are disclosed in U.S. Pat. Nos. 4,378,494 and 4,378,495.
No.

[Brief explanation of the drawing]

FIG. 1 is a plan view of a container inspection system to which the present invention is applied, FIG. 2 is an electro-optical functional block diagram of the present invention, FIG. 3 is a side view taken along line 3-3 in FIG. 2, and FIG. 4A. and FIG. 5A is a diagram similar to FIG. 3 showing an embodiment of the present invention, FIG. 4B, FIG. 4C,
Figures 4D, 5B, 5C, and 5D are
Figures A and 5A are graphs for explaining the configuration and operation of the embodiment; Figures 6A and 7A are diagrams similar to Figure 3 showing other embodiments of the present invention; Figure 6C, Figure 6D, Figure 7B, Figure 7C
FIG. 7D is a graph for explaining the configuration and operation of the embodiment of FIGS. 6A and 7A, and FIG.
Figures 8B, 8C, 8D, and 9B are diagrams showing still other embodiments of the present invention.
9C and 9D are graphs for explaining the operation. DESCRIPTION OF SYMBOLS 10... Container inspection system, 12... Star wheel, 14... Drive hub, 18... Feed conveyor, 20
... Driven screw, 22 ... Container, 24 ... Inspection station hand, 26 ... Discharge conveyor, 28 ... Followed screw, 30 ... Plunger, 32 ... Solenoid, 34 ... Exclusion discharge shot, 36 ... Drive roller, 38 ... Idler roller, 40 ... Light source, 42...
Camera, 43...pixel, 44...bracket, 4
6... Lens, 50... Filter plate, 54... Sampling circuit, 56... Information processing device, 58, 62... Outer region, 60... Center region, 64... Air bubble, 68...
stone.

Claims (1)

[Claims] 1. Apparatus 12, 14, 36, 38 for positioning the transparent container 22 and rotating it around its central axis
and a light irradiation device 40 that irradiates diffused light that passes through the side wall of the container 22 that is positioned and rotated.
a light detection device 42 arranged to receive light energy transmitted through the side wall of the container 22; comprising devices 54, 56 for detecting defects 64, 68, and devices 30, 34 for removing the containers 22 in which defects have been detected;
The light irradiation device 40 includes an illumination source 52 and a filter 50.
The filter 50 is configured such that the intensity of the diffused light transmitted through the container 22 becomes a predetermined intensity determined by a predetermined function corresponding to the position of the illumination source 52 in the width direction. A defect detection and sorting device for transparent containers, characterized in that the light from the illumination source 52 is distributed in a direction perpendicular to the central axis of the container 22. 2 The filter 50 of the light irradiation device 40 has at least two regions (58 and 6) adjacent to each other in the lateral direction.
0 or 58a and 60a or 58b and 60b or 58c and 60c), and the intensity of the diffused light transmitted through each region is determined by a different function depending on the position of the illumination source 52 in the width direction. A defect detection and sorting device for transparent containers according to claim 1, characterized in that the device is configured as follows. 3 The filter 50 of the light irradiation device 40 has three regions (58 to 62 or 58a to 62a or 5
8b to 62b or 58c to 62c), each of the regions is configured such that the intensity of transmitted light energy is determined by a predetermined function depending on the lateral position, and each of the regions 3. The transparent container defect detection and sorting device according to claim 2, wherein the predetermined function of is different from the predetermined function of an adjacent region. 4 Outer areas (58a,
62a or 58b, 62b or 58c, 62
The intensity of the diffused light transmitted through c) is approximately constant with respect to the lateral position, and the intensity of the diffused light transmitted through the intermediate region (60a or 60b or 60c) is a constant of the diffused light of the outer region. 4. The defect detection and sorting device for transparent containers according to claim 3, wherein the defect detection and sorting device is different from the strength. 5 Both outer regions (58a, 62a or 58
The almost constant intensity of the diffused light transmitted through the intermediate region (60b, 62b) is different from each other, and
5. The defect detection and sorting device for transparent containers according to claim 4, wherein the intensity of the diffused light transmitted through the a or 60b is between the different intensities of the diffused light in the both outer regions. 6. The intensity of the diffused light in the intermediate region 60a is between the substantially constant intensities of the diffused light in the outer regions 58a and 62a, and varies according to a constant function corresponding to the lateral position across the intermediate region. Claim 4 characterized in that the changes are as defined by
A defect detection and sorting device for transparent containers according to item 1 or 5. 7. The intensity of the diffused light in the intermediate region 60a varies as determined by a substantially linear function in response to lateral position across the intermediate region. The transparent container defect detection and sorting device described above. 8. The intensity of the diffused light in the intermediate region 60b or 60 is substantially uniform with respect to the lateral position, and
5. The defect detection and sorting device for transparent containers according to claim 4, wherein the intensity of the diffused light of rays c and 62c is not equal to each other. 9. A transparent container according to claim 8, characterized in that the diffused light in the outer regions 58b, 62b has a substantially constant intensity with respect to the lateral position, and the intensities are not equal to each other. defect detection and sorting equipment. 10. The defect detection and sorting device for transparent containers according to claim 9, wherein the intensity of the diffused light in the intermediate region 60b is an intermediate value of the unequal diffused light intensities in both the outer regions. 11 The substantially constant diffused light intensity of both outer regions 58c, 62c is substantially equal to each other, the diffused light intensity of the intermediate region 60c is substantially uniform with respect to the lateral position, and the diffused light intensity of both outer regions 9. The defect detection and sorting device for transparent containers according to claim 8, characterized in that the light intensity is not equal to the light intensity. 12. The transparent container defect detection and sorting device according to claim 11, wherein the diffused light intensity of the intermediate region 60c is smaller than the diffused light intensity of the outer regions 58c and 62c. 13 The photodetection device 42 comprises a plurality of photoresponsive elements 43 arranged in a linear array substantially parallel to the central axis, and the defect detection devices 54, 56
a comparison device for comparing the intensity of light energy detected by the photoresponsive element 43 with the intensity of light detected by an adjacent photoresponsive element at each rotational increment of the container; A transparent container according to any one of claims 1 to 12, characterized in that it has an identification device for identifying defects 64, 68 in the side wall of the container by a function of the difference. Defect detection and sorting equipment. 14. A defect in a transparent container according to claim 13, characterized in that the identification device has means responsive to the comparison device to distinguish between the types of defects 64, 68 as a function of the difference therebetween. Detection device. 15 Device 1 for positioning and rotating the container
a conveyor device 18 for sequentially feeding and unloading said containers 22 to and from 2, 14, 36, 38;
26 and a device 32 for selectively activating the ejector device 30 to remove the container from the conveyor device when the defect is detected. A defect detection and sorting device for transparent containers according to any one of the items. 16. A method for sorting transparent containers having defects in their side walls that affect optical transparency, including the steps of: (a) rotating the transparent container around its central axis; and (b) adjusting the intensity of light transmitted through the container to directing through the side wall of the container diffused light whose intensity is determined as a value of a predetermined function in response to the position of the illumination source in a direction substantially orthogonal to the central axis of the container; d) positioning a camera having a plurality of photoresponsive elements arranged in a linear array extending in a direction parallel to the central axis to form a parallel field of view; identifying a defect in the side wall of the container as a function of the difference in light energy detected; and (e) rejecting containers in which a defect has been identified and sorting defective containers from acceptable containers. Defect detection and sorting method for transparent containers. 17 In step B, the light from the illumination source directed through the side wall of the container is divided into at least two regions by filtering the light, and the intensity of the diffused light in each region is determined by the difference in position in the direction transverse to the central axis of the container. 17. The method for detecting and sorting defects in transparent containers according to claim 16, further comprising the step of changing the value of the function so as to have a value of a function. 18 In step B, the light from the illumination source directed to pass through the side wall of the container is divided into at least three regions adjacent to each other in the width direction of the illumination source, and the intensity of the diffused light in each region is adjusted to the widthwise direction. Defect detection and sorting of transparent containers according to claim 17, characterized in that the method is determined as a value of a function of position, and includes the step of changing the function of each region to be different from the function of an adjacent region. Method. 19 The step B includes (a) filtering the diffused light so that the intensity of the diffused light in the outer region of the region is approximately uniform with respect to the horizontal position; b) filtering the diffused light such that the intensity of the diffused light in the intermediate region of the regions is determined as a function different from the substantially uniform light intensity in each of the outer regions; The method for detecting and sorting defects in transparent containers according to claim 18. 20. Claim 19, wherein the respective substantially uniform light intensities of the outer regions differ from one another at intensity levels, and the light intensity of the intermediate region lies between the different intensity levels of the outer regions. The method for detecting and sorting defects in transparent containers as described in 2. 21. Claims characterized in that the light intensity in the intermediate region is between substantially constant intensities in each of the outer regions and varies as a value of a uniform function with respect to the lateral position of the intermediate region. The method for detecting and sorting defects in transparent containers according to any one of Item 19 and Item 20. 22. The method for detecting and sorting defects in transparent containers according to claim 21, wherein the light intensity in the intermediate region changes as a value of a substantially linear function with respect to the lateral position of the intermediate region. 23 In the step B, the diffused light transmitted through the container forms three light regions adjacent in the width direction of the illumination source that crosses the central axis of the container, and the diffused light intensity of the outer region crosses the central axis of the container. The intensity of the diffused light in the intermediate region is substantially uniform with respect to the position in the direction transverse to the central axis, and the intensity of the diffused light is unequal to the intensity of the diffused light in both the outer regions. 20. The method for detecting and sorting defects in transparent containers according to claim 19, further comprising the step of applying a filter. 24. The method of claim 23, wherein the step (b) includes the step of filtering the diffused light so that the outer regions are not equal to each other and have substantially uniform diffused light intensity. Defect detection and sorting method for transparent containers. 25. Claim 25, characterized in that step (B) includes an additional step of filtering the diffused light so that the intensity of the diffused light in the intermediate region is an intermediate value between the intensities of the diffused light in both the outer regions. 25. The method for detecting and sorting defects in transparent containers according to item 24. 26. Said step (b) comprises the step of filtering the diffused light of said outer region so that the diffused light intensities of said outer regions are equal to each other and have substantially uniform intensities. The method for detecting and sorting defects in transparent containers according to item 23. 27. The defect in the transparent container according to claim 26, wherein step (B) further includes a step of filtering so that the intensity of diffused light in the intermediate region is smaller than the intensity of diffused light in the outer region. Detection and selection method. 28 additional control for adjusting and controlling the width dimension of the region transverse to the central axis and the position of the photoresponsive element with respect to the region so as to detect defects in the container sidewall of a predetermined minimum size and type; A defect detection and sorting method for transparent containers according to any one of claims 16 to 27, which comprises steps.