JPH0331221B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for detecting flaws in glass bottles using changes in the intensity of scattered light.

Glass bottles used as containers for beer, soft drinks, etc. are generally collected and used many times, but as the number of times they are used increases, the top and bottom ends of the cylindrical part of the bottle tend to develop scratches. At the same time, the flaw strength (flaw depth) and flaw width increase. Furthermore, as the number of uses increases, the number of irregular flaws on the entire surface of the cylindrical portion increases. Such glass bottles have a reduced commercial value and deteriorated strength, and are at risk of breakage. In particular, when used as a container for carbonated beverages, the interior is under pressure and is easily damaged.

Therefore, in order to automate the process and save labor, bottling factories need a device that can detect flaws in glass bottles.

However, at present, when detecting glass bottles with flaws, they are sorted by human visual judgment, and accurate detection is not possible. Recently, optical flaw detection devices have been announced that detect flaws as changes in scattered light, reflected light, and transmitted light. Due to non-uniformity, etc., changes in the amount of detected light received, a shift in the timing of obtaining flaw information, etc. are likely to occur, and as will be described in detail below, there is a problem that there are many erroneous determinations.

Using a beer bottle as an example, the intensity distribution of flaws will be explained with reference to FIG. 1. The determination criteria are set as follows.

Initial state: There are few or no flaws.

2nd stage state: A weak, bead-shaped flaw A occurs at the lower end of the cylinder.

3rd stage state: Although there are visible flaws in the shape of a bead on the upper end B and lower end A of the cylindrical part, the strength of the flaws is light.

Fourth stage state: Intense flaws are formed on the upper and lower ends of the cylindrical portion, and flaws are also formed on the entire cylindrical portion.

Terminal condition: There are strong, bead-like flaws on the upper and lower ends of the cylinder, and considerable flaws also occur on the entire cylinder.

Of the above-mentioned flaws, only the third stage is acceptable as a product; however, these standards are conceptual standards, and even if the flaws are of the same degree visually, they must be examined in detail. Looking at the flaws, there are flaws that are small in strength (flaw depth) but wide, and conversely, flaws that are narrow in width but strong, and there is little difference in strength and width in the distribution around the circumference. They are not uniform, and their strength and width vary greatly depending on their circumferential position. In particular, regarding the strength and width of one round of flaws, if the bottle has uneven letters or marks that indicate the manufacturer's name C, anti-slip D, capacity E, etc., the fourth or final stage of the bottle should be marked accordingly. Differences in judgments between parts become noticeable, and in optical detection, the dispersion in judgments becomes even larger.

Figure 2 shows an example of the optical sensor output obtained by scanning the convex capacitance display surface at the bottom end of a beer bottle, where weight is generated. These are the detection results of the bins that are determined, and (b) is the detection result of the bins that are determined to be in the fourth stage state, which is close to the terminal state. From these detection results, it is possible to determine the intensity state without signal processing, such as comparing the average value with a reference value, but in order to scan the bottle once, it is necessary to make major changes to the bottle transfer device on the production line. As a result, it requires a much more complex and precise transfer device than conventional transfer devices. Furthermore, when determining the flaw intensity by detecting a single point on the sensor output curve or scanning a narrow range as shown in Figure 2, the flaw intensity varies significantly depending on the detection position, and the probability of misjudgment is extremely high. Become.

The present invention has been made in view of the above-mentioned circumstances.The projector and the receiver are placed in the same plane position and approximately perpendicular to the bottle transport direction, and the flaw intensity of about half a circle due to the bottle transport is detected. It is an object of the present invention to provide a detection method that enables reliable flaw intensity detection at a relatively low cost by determining the intensity when the peak value of a detected value does not exceed a reference value.

FIG. 3 shows the arrangement of a light projector and a light receiver in the present invention. The bottle 2 is transferred by the transfer device 1 in the direction shown by the arrow in the figure, and a projector 3 has one or more light sources to irradiate the flawed part of the bottle 2 with light, and scattering from the flawed part of the bottle 2. The light receivers 4 for detecting light are arranged with their respective optical axes at angles α on substantially the same plane and substantially perpendicular to the direction of bottle transport.

In this configuration, the bottle 2 is transported in the direction of the arrow by the transport device 1, and in the process of passing through the intersection of the optical axes of the emitter 3 and the light receiver 4, about half of its circumferential surface is optically scanned, and the intersection is moved to the lower end. By setting such a flaw location, the light receiver 4 detects a flaw intensity distribution curve corresponding to the flaw intensity.

The emitter 3, the receiver 4, and their signal processing circuits are constructed as shown in FIG. The projector 3 has a reflector 7 on the back of a lamp 6 that is powered on by a power source 5 to obtain a parallel beam of light toward the flaw detection area of the bottle 2, and this parallel light is sent to condensing lenses 8 and 9. It is configured to be focused into a spot shape up to the width of the flawed part. The light receiver 4 has the scattered light from the flawed part of the bottle 2 as its optical axis, and the scattered light is collected by lenses 10 and 11, and the detection area of the flawed part is set by the slit 12 as shown in FIG. As shown, the vertical frame is set to have a width a and a height b, and the light receiving flux passing through the slit 12 is detected by the light receiving element 13 as a voltage (or current) signal corresponding to the light intensity. The detection signal of the light-receiving element 13 is amplified to an appropriate level by the amplifier circuit 14, and the amplified signal is compared with a comparison reference value by the comparison/judgment circuit 15. A bin position signal and a comparison reference value are given to this comparison judgment.

In such a configuration, when the bottle 2 is transported and the flaw site is scanned in the axb detection area,
As shown in Figure 6, the detection area is −π/2<θ<
Move at π/2. At this time, the amount of light Iθ received by the light-receiving element 13 has the following relationship if the intensity of the flaw is uniform over the entire scanning range and the scattered light due to the flaw has no directionality.

Iθ=K・Io・a・b・cosθ/(cosθ) {L+r(1−cosθ)} ² K=・Io・a・b/{L+r(1−cosθ)} ^2However , Io depends on the projector. The luminous flux density irradiated to the bin, K is the scattered light quantity constant, L is the shortest distance between the bin and the light receiving surface, and r is the radius of the bin.

In the above equation, if L is made sufficiently larger than r, the amount of received light Iθ will be approximately constant in the scanning range. Therefore, for actual flaw detection, the flaw intensity distribution in the scanning direction can be obtained as the output of the amplifier circuit 14, and the comparison judgment circuit 15 determines whether the peak value from the flaw intensity distribution exceeds the comparison reference value. It is possible to obtain a result of determining whether the bottle is good or bad.

FIG. 7 shows experimental data for verifying that the amount of received light Iθ is approximately constant within the scanning range. Figure a
The flaw intensity distribution curves shown in center, a, b, and c are shown in a, center, and center when the emitter 3 and receiver 4 are located on the center line with respect to the sample bottle 2A, as in a, b, and c shown in b. These are the measurement results obtained by rotating the bottle 2A once in B when the bottle is positioned 20 mm from the center line, and C when the bottle is positioned 35 mm from the center line.
In the experiment, the radius r of the bin 2A was 37.5 mm, the shortest distance L to the receiver was 600 mm, and the detection area width a
is 1 mm, height b is 20 mm, and angle α between emitter and receiver is
is set to 45 degrees. As is clear from the curves in Figure 7a, even if the positions of the emitter and receiver are off the center line of the bottle, even if there is a difference in absolute level with respect to the flaw position, they become similar, and as the bottle is moved, It is possible to accurately measure the flaw intensity distribution by scanning the half circumferential surface.

Note that the difference in the absolute values of the flaw intensity distributions A, B, and C can be reduced by increasing the distance L and decreasing the angle α, and can also be automatically corrected using the bin scanning position signal. can.

Figure 8 shows sample bottles (20 bottles for each condition) that have been visually classified in advance according to the five criteria described above, and each sample bottle is scanned four times by changing its scanning position by 90 degrees with respect to the transport direction. , the highest peak value at each position is taken, and the highest and lowest values are further shown for four scans. Figure A shows the measurement results for the sample bottle in the initial state, B to E.
shows each measurement result from the second stage state to the final stage state. As is clear from this experiment, no matter which direction the bin is oriented, there is a part in the flaw intensity distribution where a relatively high output (peak) is obtained, and the flaw intensity can be determined with certainty by scanning approximately half the circumference of the bin. It is possible to determine whether the bottle is good or bad based on the measurement. For example, in Fig. 8, if the comparison standard is set to V _REF , all sample bottles that are visually determined to be in the early stage or second stage state are also determined to be good products by the method of the present invention, and are visually determined to be in the final stage. Regarding the bottles, all but one bottle is judged to be defective, and good judgment results can be obtained for the third and fourth periods as well, considering the uncertainty of visual judgment. We were able to obtain a comparable correlation with discrimination. For actual flaw detection, a reference value V _REF was given to determine the quality of the sample bottle, and then the sample bottle was visually inspected, and the results showed that the discrimination was sufficiently valid.

As described above, according to the method of the present invention, it is possible to easily and reliably obtain the flaw intensity distribution of a bottle and its quality determination using an existing bottle transfer device. In addition, by comparing the peak value of the flaw intensity distribution curve with a comparison standard to determine pass/fail, it is possible to almost eliminate misjudgments due to non-uniformity in the flaw intensity portion of the bottle and non-uniformity in the bottle feed speed. . Note that since the present invention uses scattered light, there is almost no influence on the determination due to differences in the color of the bottles. Furthermore, the projected light is made into a spot so that it is less susceptible to scattered light from bottles and the like. In particular, in the present invention, by arranging the distance between the light receiver and the glass bottle at such a distance that the amount of light received from the half circumference of the transported glass bottle is approximately constant, it is not necessary to rotate the bottle in the detection area of the light receiver. However, the degree of flaw can be detected over almost half the circumference, making it unnecessary to rotate the bottle and increasing the detection speed.

It goes without saying that the method of the present invention can be applied not only to round bottles but also to square bottles, containers made of various materials having special curved surfaces and uneven surfaces, and detection of parts.

[Brief explanation of drawings]

Fig. 1 is a diagram showing the occurrence of flaws in a glass bottle, Fig. 2 is a diagram showing an example of optical flaw detection, Fig. 3 is a diagram of the equipment layout of the method of the present invention, and Fig. 4 is the equipment configuration of the method of the present invention. Figure 5 is an explanatory diagram of the flaw detection area in the method of the present invention, Figure 6 is a diagram for explaining the principle of scanning the detection surface in the method of the present invention, and Figure 7 is a diagram for verifying the method of the present invention. FIG. 8 is a diagram showing a sample bottle discrimination result by the method of the present invention. 1... Transfer device, 2... Bin, 3... Floodlight,
4...Receiver, 5...Power supply, 12...Slit,
13... Light receiving element, 14... Amplifying circuit, 15...
Comparison judgment circuit.

Claims (1)

[Scope of Claims] 1. Light is irradiated by a projector toward a flaw detection site on a glass bottle that is being transported in a certain direction, and the optical axis of the projector is approximately in the same plane and approximately perpendicular to the direction in which the glass bottle is being transported. A glass bottle flaw detection method detects scattered light from a flawed part of a glass bottle using a light receiver arranged as shown in FIG. In the method, the light receiver is arranged at least a distance from the glass bottle such that the amount of light received from the half circumferential surface of the glass bottle is approximately constant as the glass bottle is transported. Detection method. 2. A method for detecting flaws in a glass bottle according to claim 1, characterized in that the quality of the glass bottle is determined by comparing the peak value of the detection signal of the light receiver with a comparison standard. 3 In claim 1 or 2,
A glass bottle flaw detection method characterized by adjusting the flaw detection area of the glass bottle with a slit in the photo receiver.