JPS61107144A - Method for inspecting container - Google Patents

Abstract

PURPOSE:To enable certain high speed inspection and to facilitate handling, by comparing the generation signal corresponding to the rotary position of a container by using a unidimensional image sensor. CONSTITUTION:The emitted light from the diffusion plate 5 of an illumination apparatus 3 passes through a rotating glass bottle 1 and the linear part corresponding to a part of the optical image of the bottle 1 is formed on the image sensor 8 of a camera 6. Said linear part is converted by an A/D converter 12 in a signal processing system 9 on the output of the sensor 18 and four sets of picture element signals are stored by a shift register 17. One picture element signal is obtained at the time of one scanning of the sensor 8. The output of the register 17 and A/D conversion output enter a subtraction part 14 and the subtraction output is compared with the output signal of a setting part 16 by a comparing part 15 and, when the subtraction output is larger, a signal 15a is applied to an exclusion apparatus 10 as an exclusion order signal. The inspected bottle is excluded.

Description

TECHNICAL FIELD OF THE INVENTION The present invention relates to a method for inspecting transparent or translucent containers, such as glass bottles.

(Technical Background of the Invention and Problems with the Background Art) Bottle inspection can be broadly divided into i15! bottle inspection in the filling process for beverages and foods and inspection of new bottles in the bottle manufacturing process.In the former, In recent years, with the advent of superior imaging devices such as COD and improved video signal processing techniques, high-speed and highly accurate inspection technology has been realized (Japanese Patent Application Laid-Open No. 57
-137844, JP-A-58-132650). - way,
In the latter bottle manufacturing process, the bottle processing speed is 10 per minute.
Since the speed is relatively low (0 to 200 bottles), a method is adopted in which each part of the bottle is inspected by temporarily stopping the bottle at an inspection position and then applying spin to the bottle. Conventionally, bottle inspection in the loading process focused on defects such as chips and rips, so modulated light was emitted by multiple projectors, and the reflected light was received by a number of corresponding receivers, resulting in changes in the amount of light. Detecting defects.

However, recently, demands on the quality of bottles have become more diverse, and in addition to chips and scratches on the bottle body, it is also necessary to detect foreign objects such as minute stones and metals, as well as minute air bubbles called seeds, and eliminate such bottles. be. Therefore, efforts are being made to develop this inspection technique, but a highly accurate and reliable inspection technique has not yet been disclosed. Among the recently disclosed devices is one using a one-dimensional image sensor (linear array). This uses a relatively inexpensive linear array, since using a two-dimensional image sensor (for example, a COD camera) would make the device expensive. The techniques disclosed in JP-A-58-99738 and JP-A-58-71442 use a single row or two rows of diode arrays, respectively.
The former is an attempt to reduce the effect of the dead zone, which is a defect in diode arrays, by narrowing the width of the dead zone, making the unit element a parallelogram, and covering it by combining two rows. A method is adopted in which the entire inspection area of the bottle body is scanned by linear arrays from two directions by changing the direction of the bottle.

However, with such a method, it is difficult to reliably detect minute defects (about 0.5 mm) which are currently desired.

[Purpose of the invention]

The purpose of the present invention is to use an easy-to-handle and inexpensive device to
An object of the present invention is to provide a method for inspecting containers that can be inspected quickly and reliably.

[Summary of the invention]

The present invention employs a one-dimensional image sensor (linear array) that has a higher resolution than a two-dimensional image sensor (solid-state image sensor) when viewed only in the -dimension, and rotates the container in a fixed position. A set of signals is sequentially generated by scanning a one-dimensional image sensor, and a set of signals is generated when the container is at a certain rotational position, and a set of signals is generated when the container is at another rotational position slightly different from the above rotational position. The presence or absence of a defect is determined by comparing these with each other.

[Invention Examples]

FIG. 1 schematically shows a system used to carry out the inspection method of the present invention. In the same figure, reference numeral 2 denotes a rotating table, and containers are carried in by a transport mechanism (not shown).
For example, the glass ring 1 is rotated about its central axis. A lighting device 3 includes a lamp 4 and a diffuser plate 5. A camera 6 includes a lens 7 and a one-dimensional image sensor (for example, a solid-state image sensor such as a COD) 8.

The light emitted from the diffuser plate 5 of the illumination device N3 passes through the bottle 1 and is converged by the lens 7, and a part of the optical image of the bottle 1, ie, 1a in FIG. 2(a), is displayed on the image sensor 8. A linear part corresponding to the part shown by is formed. This image is bright in areas where there is a large amount of light transmitted, and dark in areas where there is little.The vertical axis is the vertical position (height position) of the bottle.
If a graph is drawn with brightness as the horizontal axis, it will be as shown in FIG. 2(b), for example. That is, above the top of the bottle, the light from the illumination (diffusion plate) enters the image sensor directly, so the brightness and therefore the magnitude of the electrical signal are maximum and constant.

Since the area around the mouth of the bottle has a complex shape, the intensity of light passing through that area also changes in a complex manner. The brightness gradually increases from the mouth to the body, and there is no major change in the body. The brightness rapidly decreases at the bottom of the bottle, and reaches its lowest level at the base on which the bottle is placed.

The image sensor 8 using COD has a light-receiving surface at the imaging position, a light-receiving section consisting of 2,000 to 3,000 light-receiving elements arranged in a straight line, and a plurality of light-receiving elements provided corresponding to each of the light-receiving elements. and a storage section consisting of a storage element,
An amount of charge corresponding to the brightness of each pixel of the image is generated by the light receiving element, and after this charge is transferred to the corresponding storage element, it is sequentially shifted and output. Since the sequential output (reading) of signals by this shift is equivalent to the sequential output of signals by scanning the storage element, hereinafter the sequential output of signals from the storage element may be referred to as scanning.

If we plot the signal magnitudes of a series of signals sequentially generated from the image sensor that is imaging the bottle 1 in the order in which the signals are generated (therefore, with respect to the elapse of time during signal generation), we get Figure 2 ( A graph like b) is obtained.

FIG. 3 shows a signal processing system for processing the output signal of the image sensor 8.

Reference numeral 11 denotes an amplifier circuit that amplifies analog pixel signals ES sequentially output from the image sensor 8 into VSA. 1
2 is an A/D (analog-digital) converter which converts the amplified analog pixel signal VSA into, for example, 64 levels (64 levels).
bit) into a digital signal.

Of the above, the camera 6, the amplifier circuit 11, and the A/D converter 12 constitute a pixel signal generation device that sequentially and repeatedly generates signals of pixels of an image of a straight line portion of the bottle.

The shift register 13 stores four sets of pixel signals output from the A/D converter 12. Here, one set of pixel signals means a collection of signals sequentially generated when the image sensor 8 is scanned once. The shift register 13 therefore outputs the signal input four scans ago.

The subtraction unit 14 calculates the difference between the output of the shift register 13 and the output of the A/D converter 12. In other words, the difference between the signal when the bottle is at each rotational position and the signal at a time four scans earlier is determined.

Since the bottle is rotated about its axis in the inspection position, the signal obtained from each scan represents the intensity of light passing through the bottle when the bottle is in a different rotational position, The signal obtained by each scan is shown as s in Fig. 4(a).
o, si.

82, 83, and 84 represent the intensity of light passing through the bottle wall at different angles. Comparing each signal with the signal from four scans ago means, for example, the line S
This means comparing the intensity of light passing through the bottle along line O with the intensity of light passing through the bottle along line S4. (The amount of concave rotation of the bottle during one scan period is ignored and is assumed to be zero.) If we connect the points where 1i1S0.84 intersects with the surface of the bottle, we get 80.84 as shown in Fig. 4(b). becomes. The signals for one scan obtained in each case are as shown in FIGS. 4(c) and 4(d). In the illustrated example, there is a defect on the line S4 in FIG. 4(b), and therefore the signal is small in the corresponding portion, as shown by S4D in FIG. 4(d).

Comparing each signal with the signal from four scans ago also means comparing the intensities of light received by the same light-receiving element of the image sensor, and therefore the signals at the same height position on the bottle. means.

As mentioned above, the intensity of light passing through a bottle differs depending on the height of the bottle, but this difference can be canceled out by comparing signals at the same height, and it is possible to detect defects in the bottle. Only if the defect is small and is on one line, e.g. S4 and not on the other mso, will the difference between the two signals be large.

The comparing section 15 compares the magnitude of the output of the subtracting section 14 and the signal preset in the setting section 16, and if the output of the subtracting section 14 is larger, a signal 15a1 is generated. In this case, a signal may be output if the output of the subtraction unit 14 is larger even once, and a signal may be output when it is determined that the output of the subtraction unit 14 is larger at least a predetermined number of times during one scan. The signal 15a may be outputted, or the signal 15a may be outputted when the determination that the output of the subtraction unit 14 is larger is made consecutively at least a predetermined number of times during one scan. The signal 15a is given to the exclusion device 10 as an exclusion command signal. The rejecting device 10 receives this signal and rejects the inspected bottle.

In the above embodiment, each signal is compared with the signal from four scans ago, but the number of scans before the signal to be compared can be changed depending on the size of the defect to be detected, the rotational speed of the bottle, etc. sell. Also, each signal may be replaced by a plurality of other signals, e.g.
Compare not only the signals before scanning but also the signals before 3 or 5 scanning, and synthesize these comparison results (for example, logical OR,
The presence or absence of a defect may be determined based on logical product (logical product, combination of these).

Furthermore, instead of determining the difference between the signals, a ratio may be determined, and the presence or absence of a defect may be detected based on the result of determining whether the ratio is greater than or equal to a predetermined value.

FIG. 5 shows another example of the signal processing system. This example differs from the example shown in FIG. 3 in that a shift register 17 and a subtraction section 18 are inserted. The shift register 17 stores the pixel signal output from the A/D converter 12 for several pixels, for example, 5 pixels. The subtraction unit 18 is the previous one.
The purpose is to find the difference between the output of the A/D converter 12 and the output of the A/D converter 12. By finding the difference in this way, an operation similar to differentiation is performed.

Shift register 13 and subtraction section 14 receive the output of subtraction section 18 instead of the output of A/D converter 12. Therefore, the subtraction unit 14 calculates the difference between the differential signal of the pixel signal in each scan and the differential signal of the pixel signal a predetermined number of scans earlier, for example, 4 scans, and the comparison unit 15 calculates the difference between the differential signals. The presence or absence of a defect is determined based on whether the value is larger than a set value.

The reason for comparing the differential signals in this way is as follows. In other words, since a two-piece mold is used to form the bottle,
A seam appears in the formed bottle, as shown by a series of rXJ marks in FIG. 6(a). This tends to appear strongly in the body of the bottle, and the signal due to light passing through this area (indicated by It is smaller than the signal (indicated by S in FIG. 6(b)). Therefore, if the signal is processed by the circuit shown in FIG. 3, there is a risk that the bottle will be determined to be defective. On the other hand, since the influence of the seam is not constant for the entire signal for one scan and only changes little by little, the differential signal (indicated by dX in Fig. 6(C)) is different from that of the normal part (
(shown as ds in FIG. 6(C)). Therefore, the difference between the two is small, and there is no possibility that the bottle will be judged as defective due to the seam.

In the above embodiment, the signal from the image sensor 8 is converted into a digital signal and then stored in a digital register, but an analog register (for example, one containing a combination of a resistor and a capacitor) is used. You can also do that.

Also, instead of using registers that store several sets of signals (four sets in the example) as described above, a storage device with a large storage capacity, for example, that can store data over the entire circumference of the bottle (
It is also possible to use a device in which data can be written or read by specifying an address. If you do this,
Data for the same rotational position can be used repeatedly.

For example, you may want to compare not only the magnitudes of signals from image sensors, but also their differential signals, or just compare data at rotational positions that differ by a certain number of scans (for example, 4 scans). However, when comparing data at different rotational positions with a different number of scans (for example, 5 scans), the data at the same rotational position will be used repeatedly, so it is convenient to use the above storage device. be.

FIG. 7 shows another embodiment of the invention. The camera 6' of this embodiment has a built-in two-dimensional image sensor 31 in addition to the one-dimensional image sensor 8, and the light passing through the lens 7 is transmitted between the one-dimensional image sensor 8 and the two-dimensional image sensor by a half mirror 32. -31. The output of the dimensional image sensor 8 is processed by a signal processing system 9 similar to the embodiment shown in FIGS. 1 and 3 or 5.
and undergo similar treatment.

The two-dimensional image sensor 31 has 400 to 500 pixels in the horizontal direction.
About 300 to 400 light-receiving elements are arranged in a matrix in the vertical direction, and an image of the entire bottle (shown in Figure 8) can be imaged on the light-receiving surface. The output of 31 is supplied to a monitor television 33 via an amplifier 32. The synchronization signal generator 34 synchronizes the operations of the image sensors 8 and 31 and the signal processing system 9. The inspection gate setting device 35 sets the upper and lower limits of the inspection area according to the shape of the bottle and the inspection conditions.Inspection gate markers (33a, 33a for bottle 1A in FIG. 8, 33b
33c and 33b for bottle 1B).

Cameras that only incorporate a one-dimensional image sensor require the use of measuring instruments such as an oscilloscope to align the position (up and down, left and right) and the distance to the lens (focus). Scanning is complicated because it is required every time there is a change in type or shape. On the other hand, if you use a camera that also incorporates the two-dimensional image sensor 31 as described above, it will be easy to not only align and focus, but also set the inspection gate, making preparations for inspection quick and easy. obtain.

Note that even when a two-dimensional image sensor is installed, the inspection is performed using the one-dimensional image sensor 8 because the one-dimensional image sensor has higher resolution and takes less time for one scan. .

That is, as mentioned above, some one-dimensional image sensors have about 2,000 to 3,000 light receiving elements arranged, while two-dimensional image sensors have about 400 to 500 light receiving elements arranged in the horizontal direction and 300 to 400 in the vertical direction. Currently, one-dimensional image sensors are used that are arranged in a matrix, and one-dimensional image sensors have a resolution of 4.
~7 times higher. Further, since the total number of light receiving elements of a two-dimensional image sensor is about 120,000 to 200,000, the time required to scan one screen is about 40 to 100 times longer. Therefore, for inspecting bottles that are rotating at high speed, a one-dimensional image sensor is more advantageous in terms of high speed. For this reason, the embodiment shown in FIG. 7 uses a combination of a one-dimensional image sensor and a two-dimensional image sensor.

〔Effect of the invention〕

As described above, according to the present invention, a one-dimensional image sensor is used to rotate a container with its axis parallel to the longitudinal direction of the one-dimensional image sensor and fixed, and when the container is at a certain rotational position, an image is generated. The signal obtained from the sensor ° is compared with the signal obtained from the image sensor when the container is at a rotational position slightly different from the above rotational position, and based on the results of this comparison, the presence or absence of defects is determined. Therefore, scanning speed and inspection accuracy can be significantly improved.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram showing an apparatus that can be used to carry out the inspection method according to an embodiment of the present invention, FIG. b) is Fig. 2(a)
3 is a block diagram showing an example of a signal processing system, FIG. 4(a) is a diagram showing light rays passing through the bottle at different angles, and FIG. 4(b) is a diagram showing the brightness of the part shown in . Figure 4(c) and (d) are the lines So and 84 in Figure 4(b). FIG. 5 is a block diagram showing another example of the signal processing system.
Fig. 6(a) is a diagram showing a state in which line S4 in Fig. 4(b) overlaps the seam of the bottle, and Fig. 6(b) is a portion of line So, 84 in the case of Fig. 6(a). FIG. 6(C) is a diagram showing the value obtained by differentiating the brightness of FIG. 6(b) along the scanning direction, and FIG. 7 is a diagram showing the brightness of the image corresponding to FIG. 8 is a schematic diagram showing an apparatus that can be used to carry out the inspection method of the embodiment. FIG. 8 is a diagram showing an example of the image of the monitor television shown in FIG. 1... Bottle, 2... Turntable, 3... Lighting device, 6...
...Camera, 8...One-dimensional image sensor, 9...
-Signal processing system, 10...Exclusion device, 11...Amplification circuit, 12...A/D converter, 13.17...Shift register, 14.18...Subtraction unit, 15... - Comparison section, 16...setting section. Applicant's agent Kiyoshi Inomata (α) (b) 111-Brightness Figure 3 To 70 Figure 4 (α) Figure 5

Claims (1)

[Claims] 1. A container is brought to an inspection position near the one-dimensional image sensor, and with the axis of the container fixed parallel to the longitudinal direction of the image sensor, the container is centered around the axis. While rotating, the container is illuminated, an optical image of the container is formed on the image sensor, a set of signals is obtained by scanning the image sensor, and a set of signals is obtained by repeating the scanning. The sets of signals are sequentially generated, and one set of signals is compared with another set of signals obtained at a different rotational position slightly rotated from the rotational position of the container at the time when the one set of signals was obtained. and determining the presence or absence of a defect in the container based on the comparison result. 2. The method of claim 1, characterized in that the set of signals is obtained by differentiating the signals generated while scanning the image sensor from one end to the other. 3. A method according to claim 1, characterized in that a two-dimensional image sensor is used together with the one-dimensional image sensor to generate a monitor image.