JPS6355651B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and circuit device for automatically and continuously detecting different patterns of printed blanks for detecting the mixing of printed blanks such as reversed cylinders, back cylinders, and different types of cans in a body maker of a can manufacturing line.

In general, if a defective printing blank (for example, a printing blank from an inverted cylinder, back cylinder, different type of can, etc.) is mixed in with a printing blank that has a printed surface with drawn characters or drawings, when feeding the printing blank to a processing machine, etc. It is necessary to detect and process defective printed blanks during processing. The conventional method for detecting defective printed blanks was to place a quality judgment monitor on standby at the processing machine and perform visual confirmation. Proposed. This method converts the amount of light reflected from the printing surface of a moving printing blank into an electrical signal, and synchronously compares it with a reference electrical signal of a constant voltage level of a model sampled and stored in advance to determine whether the signs of the mutual electrical signals are good or not. This is a process that determines the

However, this method has the disadvantage that it tends to malfunction for can types that are not uniformly bright and dark, and that adjustments must be made frequently.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above problems and provide a method and circuit device for automatically and continuously detecting different patterns of printed blanks, which reliably detects and automatically removes defective printed blanks, and generates an alarm.

According to this method, standard light is projected onto a moving printing blank, the amount of reflected light that changes sequentially according to various patterns drawn on the printing surface of this printing blank is converted into a digital signal, and the digital signal is sampled. Write to memory for each pulse, do this for multiple printed blanks, write the maximum and minimum values for each sample pulse to memory, and provide clearance to absorb mechanical heading, vibration, and tonal errors. is set as the maximum value and the minimum value to set standard data, and then this standard data is compared with the detected value of each printed blank, and if the detected value is outside the range of the standard data, it is regarded as a defective printed blank. An error signal is generated and the defective printed blank is automatically removed from the body maker's supply path or processing relay path.

Hereinafter, the present invention will be described in detail with reference to an embodiment of the accompanying drawings.

FIG. 1 shows a schematic diagram of a circuit device for automatically detecting different types of patterns on printed blanks according to the present invention. This device includes a discrimination device 1, a reflective photoelectric detection head 2, a pulse generator 3, a proximity switch 4,
5, 6 and suction 7.
The reflective photoelectric detection head 2 is of a diffuse reflection component receiving type capable of sensing color differences when the printing surface β of the printing blank α is drawn in multiple colors as shown in FIGS. 3a, b, and c. A main body 11 containing a light receiver 9 and an IC 10 including a photoelectric element 8 having a peak in the visible region such as a transistor, a light trapping path 12 having a transparent glass 15 at the lower end, a lamp mounting having a lens 16 and a light source lamp 17 at the lower end. Hole 1
3 and a hood 14 arranged vertically in parallel. As shown in FIG. 3c, the standard projection light _L1 emitted from the light source lamp 17 is focused by the lens 16 onto the printing surface β of the printing blank α, and the diffusely reflected light L1 is generated due to the color difference of the printing surface β. ₂ passes through the transparent glass 15 and the light trapping path 12 and is converted into an electrical signal proportional to the amount of light received by the photoelectric element 8. The pulse generator 3 generates 360 pulses (Fig. 4e) per one rotation of the machine (from station to station), and is used for dividing the can body pattern. Proximity switch 4 is used for machine timing and generates one pulse per machine revolution (FIG. 4d) to set the reference starting position of the can body pattern. Proximity switches 5 and 6 are used to operate the deflector for defective cans. Proximity switch 5 is attached to the machine side and detects the can bodies that are being made and flowing, and proximity switch 6 is attached to the deflector side. , the defective cans detected by the proximity switch 5 are removed. The suction 7 generates a suction on signal when the lowermost printing blanks with printed surfaces on the outer surface stacked in the body maker's hopper are suctioned and fed one by one by the vacuum cup, and the error Er is immediately activated when a defective can is detected. This is the deflector operating signal contact output. These signals are processed by a discrimination device 1 which includes an input/output signal interface and a discrimination processing microprocessor (hereinafter referred to as MPU).

FIG. 2 shows an internal schematic diagram of the discrimination device 1, in which the MPU 25 includes the pulse generator 3, the proximity switches 3, 4, 5, the suction 7, and the MPU 25.
Signals from the manual start switch 20 for starting the sensor 5 and the lamp cut-off switch 21 for the light source lamp 17 of the reflective photoelectric detection head 2 are inputted via the input interface 22 and the data bus 31, and signals for capturing data are inputted via the input interface 22 and the data bus 31. The data is output to the A/D converter via the address bus 30 and taken in from the data bus 31. Note that the A/D converter 24 has a signal detected by the reflective photoelectric detection head 2,
The electrical signal amplified via the amplifier 23 is input as an 8-bit digital signal. MPU2
5 processes these input signals and connects them to the data bus 3 as appropriate.
1 via output interfaces 32, 33, 3
Send a signal to 4. Output interface 32
For example, the output interface 33 outputs a signal each time a can body inspection is completed, and displays the number of inspected cans on the display 27, and the output interface 33 outputs an output when a defective can is detected, and displays the number of defective cans, etc. on the display 28. Activate the error relay 35. Also, output interface 3
4 displays the display 29 when removing defective cans;
Deflector operation relay 36 for removing defective cans
energize. In addition, the MPU 25 receives power from the power source unit 26 (+5V3A, +12V0.2A, -
12V0.1A) is supplied.

Next, an outline of the present invention will be explained with reference to FIG.

FIG. 4a shows the printing surface β of the printing blank α, and the shaded areas indicate areas where reflection is particularly strong. As this printing blank α moves and passes, the reflective photoelectric detection head 2 detects a line a on the printing surface β.
, and outputs an electrical signal (Fig. 4b) proportional to the amount of reflected light. This electrical signal is A/D converted, and its digital value is determined by the sampling pulse (Fig. 4 e) output from the pulse generator 3 after the mechanical timing pulse (Fig. 4 d) is output from the proximity switch 4. It is captured. In this way, data is acquired from 10 cans, the maximum and minimum values of the 10 data are memorized for each sampling pulse, and a clearance is provided for each maximum and minimum value to create standard data (Figure 4c). Set. This standard data is compared with each data from the 11th can, and if each data is within the standard data, it is determined to be a normal can, and if it is outside the standard data, it is determined to be a defective can. Depending on the type of can, the detection head 2
In some cases, the printed surface β on the irradiation line is plain, there is no change in color, or there are fine letters or patterns.In such cases, the detection head 2 is moved on the same circumference to detect changes in the printed surface β. It should be installed where there is.

As described above, one feature of the present invention is that standard data having a certain band is set in order to recognize the pattern of a printing blank.

Next, Section 5 describes how to set standard data for printed blanks and how to detect defective printed blanks.
This will be explained in detail according to the flowcharts shown in FIGS.

The flowchart in Figure 5 starts from 10 cans to 112×
This shows the procedure for inputting 10 pieces of data (112 pieces for each can) and writing the maximum and minimum values of the 10 pieces of data corresponding to the same sample point of each can into memory addresses.

Turn on the power switch and wait for stabilization time (1 minute or more)
After the time has elapsed, turn on the start switch. When the start switch is turned on, the contents of memory address 8200 (hexadecimal number) are set to 0 during the initial set, and the next
Set the contents of address 8201 to 255 (maximum value for 8-bit input). When this operation is repeated 112 times, the content at address 82DE becomes 0, the content at address 82DF becomes 255, and the process ends. That is, 0 is entered at the address where the maximum value data is stored, and 255 is entered at the address where the minimum value data is stored. Therefore, 112 each of 0 and 255 are alternately stored in 244 addresses from address 8200 to address 82DF. Further, in the initial set, the content m of an address used for calculation (for example, address 82F0) is set to 0. When the initial set is completed, other addresses used for calculations (for example,
82F1 address) is set to 0. When the mechanical timing pulse rises from the proximity switch 4, data (8-bit digital value of the output from the reflective photoelectric detection head 2) is read in synchronization with the rise of the sample pulse of the pulse generator 3. This data is compared with the contents of address 8200 (MAX ₀ ), and if the data is large, it is written to the contents of address 8200 (MAX ₀ ). Similarly, this data is compared with the contents of address 8201 (MIN ₀ ), and if the data is smaller, it is written to the contents of address 8201 (MIN ₀ ). Next, add 1 to n, check whether sampling has been performed 112 times, and if n is not 112, return along the flow line, wait for the rise of the next sample pulse, and read the next data. After repeating the same procedure as above 112 times, you will get out of this loop. As a result, at the end of this operation in the first can, 82DF from address 8200
All contents of memory up to the address are rewritten to input data. This is because, in the initial set, the contents of the memory where the maximum value is written are set to 0, and the contents of the memory where the minimum value is written are set to 255.
(maximum value). That is,
At the end of the first can, the contents of the memory from address 8200 to address 82DF are the values shown by the symbol (–ⓓ) in the graph shown in FIG.

Next, add 1 to m, check whether 10 cans have been input as sample cans, and if m is not 10, return along the flow line, set n to 0 again, and enter 2 cans.
Start sampling cans. If you repeat this 10 times (sampling of 10 cans is completed), you will break out of this loop. As a result, the maximum and minimum values at the same sample position of the 10 cans are written at the addresses where the maximum and minimum values are written. The symbols (〓〓) in the graph of FIG. 8 indicate the maximum and minimum values.

The flowchart shown in FIG. 6 shows the procedure for providing clearances at the maximum and minimum values to absorb errors in mechanical direction of movement, vibration, and color tone.

First, set n to 0 and m to 1. next
Compare the contents of address 8200 (MAX ₀ ) and the contents of address 8202 (MAX ₁ ), and when MAX ₀ < MAX ₁
Rewrite MA ₀ to MAX ₁ , then compare the contents of address 8201 (MIN ₀ ) and the contents of address 8203 (MIN ₁ ), and if MIN ₀ > MIN ₁ , rewrite MIN ₀ to MIN ₁ . Next, add 1 to m, and if m is not 7, return along the flow line. In other words, in this loop, the contents of address 8200 are compared with the contents of the six following maximum values (contents of addresses 8202, 8204, 8206, 8208, 820A, 820C), and the maximum value is set to 8200.
Similarly, the contents of address 8201 and the following 6
The contents of the minimum value (8203, 8025, 8027, 8029,
The contents of addresses 802B and 802D) are compared, and the minimum value is set as the contents of address 8201. When exiting this loop, 1 is added to n, and when n is not 106, the process returns along the flow line and executes this loop again. After completing this loop 106 times, the comparison based on all addresses is completed, and the maximum value of the contents of that address and the six following even addresses is written to the even numbered address, and the maximum value of the contents of that address and the following six even addresses is written to the odd numbered address. The minimum value among the contents of the six odd addresses is written. The reason why the number of exits from this loop is set to 106 is that there are no addresses to compare for the last six. The symbols and circles in the graph shown in FIG. 8 indicate the maximum and minimum values after setting the clearance in the traveling direction.

Next, a clearance is provided to absorb mechanical vibrations and color tone errors. First, set n to 0, add 10 to the contents of address 8200 (MAX ₀ ),
Subtract 10 from the contents of address 8201 (MIN ₀ ). Then n
Add 1 to , and if nn is not 106, return along the flow line and return to the contents of address 8202 (MAX ₁ ).
Add 10 and subtract 10 from the contents of address 8203 (MIN ₁ ). If you repeat this 106 times, you will break out of this loop. i.e. 10 for this loop and all max values
and subtract 10 from all minimum values. The symbols ▲ and ● in the graph shown in FIG. 8 indicate the maximum and minimum values after mechanical vibration and color tone clearance settings, respectively.

The flowchart shown so far shows the steps to set the standard data, and the flowchart shown in Figure 7 is based on this standard data.
This shows the procedure for performing an inspection starting from the can's grain.

First, m and n are each set to 0, and when the mechanical timing pulse rises from the proximity switch 4, 1 is added to n in synchronization with the rise of the sample pulse of the pulse generator 3. When n becomes 3, n is set to 0 again. here,
The data of the first three sample points are skipped. The reason for this is to make effective the clearance for absorbing the error in the traveling direction mentioned above. In other words, this is to ensure that the data input is in the center of the standard data. Next, when the sample pulse of pulse generator 3 rises, input the data, compare the data with the contents of address 8200 (MAX ₀ ) and the contents of address 8201 (MIN ₀ ), and if the data is greater than MAX ₀ . , or if it is smaller than MIN ₀ (deviating from standard data), add 1 to m. Next, add 1 to n, check whether the comparison of 106 sample points has been completed, and if n is not 106, return according to the flow line. Then, the data at the next sample point is compared with the standard data. Repeat this procedure 106 times to get out of this loop. After 106 comparisons are completed, if two or less data fall outside the standard data range, the can is determined to be normal, and the next can is inspected. Furthermore, if three or more pieces of data are out of order, it is determined that the can is abnormal, an error is output, and then the inspection moves on to the next can. The set value for the number of errors is preferably 0, but considering sudden noises, vibrations, etc., only when there are 3 or more defective data out of 106 data comparisons per can is considered an abnormal can.

The error output is a signal for deflector operation to remove defective cans. In addition, detection of breakage of the light source lamp 17 of the reflective photoelectric detection head 2 and measurement of the time since the suction 7 is turned off are performed, and when a breakage of the light source lamp 17 is detected, an error is output and the suction 7 is configured to set the standard data again after a certain period of time has passed from when it is turned off until it is turned on again.
It has been omitted in this flowchart.

In the present invention, the standard data is set based on the data of the first 10 cans, but it goes without saying that this number of cans can be changed as appropriate depending on the vibration condition of the can, the type of can, etc. In addition, in the present invention, a clearance is provided in the front and rear directions by comparing with the following six samples, and by adding 10 and -10 to the maximum and minimum values at each sample point, the vertical direction (level direction)
Although a clearance for error in color tone is provided, the method is not limited to the method according to this embodiment as long as an appropriate clearance can be set.

As explained above, according to the present invention, a pattern recognition method that performs comparison and discrimination based on standard data with a certain band is adopted, so defective cans such as reversed cans, reversed cans, and different types of cans can be reliably detected. In addition, defective cans can be automatically removed by a deflector that works with the device of the present invention. Furthermore, since the standard data storage operation is an automatic reset method using a single switch, adjustment and handling are easy, and it is possible to immediately respond to dirt or malfunction of the light emitter/receiver, or changes over time in the body manufacturer.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram of a circuit device for automatically detecting different types of patterns on printed blanks according to the present invention, Fig. 2 is an internal schematic diagram of a discriminating device according to the present invention, and Figs. 3 a and b are front views of a reflective photoelectric detection head. Figure and side view, 3rd
Figure c is a diagram of the same operating principle, Figure 4 is a timing chart of signals generated by each part of the device of the present invention, Figures 5 to 7 are flowcharts showing an example of the processing method according to the present invention, and Figure 8 is a diagram of the present invention. It is a graph showing a standard data setting procedure according to the invention. 1... Discrimination device, 2... Reflective photoelectric detection head, 3... Pulse generator, 4, 5, 6...
Proximity switch, 7...Suction, 20...Start switch, 21...Lamp out switch, 2
2...Input interface, 23...Amplifier,
24...A/D converter, 25...MPU, 32,
33, 34... Output interface.

Claims (1)

[Claims] 1. A process of irradiating and scanning the surface of a printing blank with light and converting the amount of reflected light into digital values corresponding to a plurality of predetermined points is performed on each of a predetermined number of print blanks as samples, Standard data each having a certain value range is formed for each of the plurality of predetermined points, and the digital values obtained corresponding to the plurality of predetermined points from the print blank to be subsequently inspected are A method for automatically and continuously detecting different patterns of printed blanks, which detects defective printed blanks depending on whether or not the value falls within the value range of the standard data at a point where the printed blanks are detected. 2. The standard data is set by detecting the maximum and minimum values of the digital values at each point for the predetermined number of sample print blanks, and giving a clearance to these detected values. A method for automatically and continuously detecting and processing different types of patterns on printed blanks according to claim 1. 3 The above-mentioned clearance rewrites the digital value at each point to the maximum and minimum values of the digital values of the points in the vicinity including that point, and then increases the rewritten values by a predetermined value for the maximum value and increases the minimum value by a predetermined value. 3. The automatic continuous detection and processing method for dissimilar patterns of printed blanks according to claim 2, wherein: is set to be decreased. 4. A reflective photoelectric detection head that generates an electric signal in proportion to the reflectance of the printed surface of the printing blank that moves and passes; a proximity switch that generates a signal at a timing every time the printing blank moves and passes; A pulse generator that oscillates a sampling pulse, and a signal obtained from the reflective photoelectric detection head is read in synchronization with the sampling pulse from the pane generator on the condition that the signal from the proximity switch rises. , detect the maximum and minimum values of the signal obtained from the reflective photoelectric detection head for each of the sampling pulses for a predetermined number of printed blanks as samples, and set a certain value range with a clearance to the detected values. When the value of the signal obtained from the reflective photoelectric detection head for the printed blank to be inspected deviates from the value range of the standard data, the deflector for removing defective printed blanks is set. A processing circuit device for automatically and continuously detecting different types of patterns on printed blanks, comprising: an arithmetic processing device for outputting driving signals;