JPS63147271A - Method for detecting printed matter - Google Patents

Abstract

PURPOSE:To cover a fluctuated groups within an allowable range by using a standard consisting of a success/failure deciding standard and an investigating standard strict less than said deciding standard in case the success/failure of measured points are decided based on the light reflecting quantity. CONSTITUTION:If a code 1 shows an initial standard, a full line 1a shows a success/failure deciding standard together with a broken line 1b showing an investigating standard strict less than said deciding standard. The detection value of the next print surface is set between the line 1a of a standard 1 and the line 1b and outside the success/failure deciding standard 1a and within the investigating standard line 1b. In such a case, a new standard 2 is additionally registered based on the detection value of the print surface. The standard 2 consists of a success/failure standard 2a and an investigating deciding standard 2b. Such operations are repeated to set plural standards in a fluctuated group 4. Thus the fluctuated groups can be covered within an allowable range.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is designed to prevent printing errors in a wide variety of printed materials (Honkawa 111!
The term "error" is used in a broad sense to include stains, etc.).

A common method for detecting printing errors is to convert the state of reflected light from the printing surface into an electrical signal, and compare this with the electrical signal from the paper surface to be detected. It is used to judge whether the product is good or bad.

The present applicant has spent many years researching and developing a dynamic method of moving the paper surface and extracting changes in reflected light within a certain zone of the paper surface as a signal. For example, JP-A-60
See JP-A-58534, JP-A-60-125508, etc.

The applicant has also researched and developed a static detection method in which the paper surface is stopped and changes in reflected light within a certain zone of the paper surface are extracted as a signal. In particular, improvements have been made to the stationary paper surface detection method for detecting collation errors.

In either detection method, light is applied to multiple measurement points on the printed surface, the amount of reflected light at all points is detected, and based on the detected value, plus and minus allowances are set at all measurement points as a standard. Then, light is applied to a plurality of corresponding measurement points on the subsequent printed surface, the amount of reflected light from the subsequent printed surface is detected, and the detected value of the reference and the subsequent printed surface are compared for each measurement point. to identify printing errors.

Conventionally, such standards have been of fixed type or update type.

When using a fixed standard, use the detected value of the printing surface that should be the standard and a fixed width above and below it as the standard, fix that standard, and use the printed surface of all other subsequent printed matter as the fixed standard. pass/fail was determined by

In addition, in the case of an update type standard, a new standard is updated one by one based on the detected value of the printed surface when the previous standard was determined to be good. For example, as shown in Fig. 15, when the detected value of the next printed surface is included in the original standard 1, the next standard 2 is updated as a new standard, and the next standard is added to the standard 2. When the detected value of the printed surface is received, a new standard 3 is created from the detected value of the printed surface and updated one by one.

, is trying to.

When a fixed standard is used, pass/fail judgment becomes too strict, and the probability that a printed surface that should originally be judged as "good" will be judged as defective increases especially as time passes. Therefore, there is a drawback that depending on the type of printed content, results of pass/fail judgment may be unsuitable for practical use.

In addition, the method of updating the criteria one by one has the advantage of changing the criteria over time, allowing the pass/fail judgment to be modified appropriately according to the flow of work, but the method shown in Figure 15 Even if the detected value of the printed surface falls within the variation group 4 within the allowable range indicated by reference numeral 5, there is a drawback that it may be incorrectly determined to be defective.

This invention solves the drawbacks of the prior art as described above, and provides a printed matter detection method that prepares a plurality of standards and can cover a group of dispersion within an acceptable range. The purpose is to

[In order to achieve the above-mentioned object, the present invention shines light on a plurality of measurement points on the printing surface, detects the amount of reflected light at all points, and calculates the plus and minus allowances based on the detected values. Set at all measurement points and use as a standard.
After that, the amount of reflected light from the subsequent printed surface is detected by shining light on a plurality of corresponding measurement points on the subsequent printed surface, and the detected value of the subsequent printed surface is compared for each measurement point with the above-mentioned standard. In a printed matter detection method for determining printing errors, the criteria include a pass/fail criterion and a more lenient scrutinizing criterion. The gist of this invention is a printed matter detection method characterized by additionally registering a new standard based on the detected value of the printed surface when a determination is made.

In order to overcome the drawbacks of the prior art as described above, the applicant first attempted the following solution. In other words, an attempt was made to additionally register all the detected values of the printing surface that were included in the original standards as new standards. In this way, it is theoretically possible to set a large number of standards to cover a group of variations within a certain allowable range.

However, when they actually attempted to do so, they realized that the number of criteria was too large and it would be impossible to solve the problem without using a large computer.

Therefore, further research was conducted and a printed matter detection method according to the present invention was developed. In this invention, the criteria are made up of pass/fail discrimination criteria and scrutinized discrimination criteria that are softer than the pass/fail discrimination criteria. New standards are additionally registered based on the detected values.

For example, as shown in FIG. 1, assuming that the reference numeral 1 indicates the original criterion, the solid line 1a indicates the pass/fail criterion, and the broken line 1b indicates the stricter criterion. If the detected value of the next printed surface is located between the solid line 1a and the broken line 1b of standard 1 and falls within the range of pass/fail judgment standard 1a but falls within the scrutiny judgment standard 1b, A new standard 2 is additionally registered based on the detected value of the printed surface. This new standard 2 is pass/fail judgment standard 2a and scrutiny judgment standard 2.
Consists of b. Similarly, a new criterion 3 is additionally registered to consist of pass/fail discrimination criterion 3a and scrutiny discrimination criterion 3b. By repeating this method, a plurality of standards are set in the variation group 4.

In short, it not only determines whether the printed surface is acceptable, but also examines whether it can be used as a new standard, and if it is worthy of the new standard, it is added and registered as a new standard.

For example, first, pass/fail judgment is made on the printed surface based on the pass/fail criteria, and then, for printed surfaces that are judged to be defective,
The detected value of that printed surface is examined to see if it can become a new standard, and if it is determined to be worthy, a new standard corresponding to the detected value of that printed surface is additionally registered. If the pass/fail criteria determines that the printing surface is good, the criteria may not be updated. If necessary, even if the pass/fail criteria determines that the printing surface is acceptable, a new standard that corresponds to the detected value of the printed surface may be updated. Although new standards are added and registered, when the number of new standards added exceeds a preset number, it is also possible to update the standards by deleting them in the order of oldest registration. Also,
Even if the criteria are not updated when the pass/fail judgment criteria determines that the criteria is acceptable, when the number of new criteria added exceeds a preset number, the criteria will be updated by deleting the oldest registered criteria. You can also do it.

It is desirable to arbitrarily set the above-mentioned pass/fail discrimination criteria and examination discrimination criteria in advance so that the most suitable one can be selected depending on the type of print content.

Further, the pass/fail determination criteria may be configured to include a pass/fail level and the number of defective points, and the scrutiny criteria may be configured to include a scrutiny level and the number of scrutiny points. Differently from this, it is also possible to configure the pass/fail discrimination criteria to consist of pass/fail level and defective rate, and the scrutiny discrimination criteria to consist of scrutiny level and scrutiny rate.

1-1 The printed matter detection method according to the present invention can be applied to both a static detection method in which the printed surface is detected while it is stopped, and a dynamic detection method in which the printed surface is detected while moving. Although various types of sensors can be used, it is preferable to use a type of sensor in which a large number of pairs of light emitting elements and light receiving elements are arranged, for example, the sensor described in Japanese Patent Application Laid-open No. 125508/1983. An example of this type of sensor is illustrated in FIG.

In the case of a static detection method, using one sensor as shown in Figure 9 will only detect a very limited area of the print surface, but depending on the print content, it may be sufficient. Something happens. For example, this is the case when collating errors are detected. A sensor is installed on the table on which the signatures are placed, and after the bottom signature is drawn, it is possible to determine the printing surface of the next signature.

Further, the printed matter detection method according to the present invention can also be applied to detection of rotary printing of newspaper. In this case, one or more sensors are arranged in one or more rows across the width of the web-like newspaper passage. FIG. 2 representatively shows only one sensor among the plurality of sensors arranged in this way, and one set of a light emitting element and a light receiving element. Move printing surface A in the direction of arrow B.

In reality, a plurality of sensors 10 are arranged across the entire width of the printing surface A, and each sensor is provided with a plurality of sets of light emitting elements 21 and light receiving elements 18, so when viewed as a whole, , a large number of light emitting elements 21 and a large number of corresponding light receiving elements 18 are present at predetermined intervals over the entire width of the printing surface A.

As the printing surface A is moved, a large number of measurement points (pixels) are arranged like the eyes of the base over the entire printing surface, as illustrated in FIG. 9, for example.

(In reality, due to the characteristics of the reflected light, the measurement points naturally overlap each other by a considerable width in the vertical and horizontal directions.) To simplify the diagram, one light receiving element 18 is shown in FIG.
In FIG. 9, one measurement zone consisting of a series of measurement points corresponding to one light-receiving element 18 is shown by hatching to make it easier to distinguish it from other measurement zones.

Now, as the light receiving element 18 moves toward the printing surface, the content of printing on the printing surface (that is, the amount of light received) changes.

In other words, each light-receiving element 18 exhibits a change in the amount of light received depending on the print content. The amount of light received is converted into an analog electrical signal by the light receiving element 18. This analog electrical signal is amplified by an amplifier, then enters the A/D converter E, where it is converted into a digital electrical signal. The A/D converter E performs sampling at a constant time using a signal from the encoder Q (or clock circuit). In other words, it is time-divided. Therefore, if the printing surface A is moved at a constant speed, digital electrical signals are output at measurement points at regular intervals one after another along the direction of movement toward the printing surface.

FIG. 3 shows one measurement zone (corresponding to the hatched area in FIG. 9) when measuring from the front edge to the rear edge of the printing surface A with one light-receiving element 18 while moving the inside of the print in this way. An example of an analog electrical signal and a digital electrical signal is shown.

Such a digital electrical signal is stored in the first memory F.

Then, the digital electrical signals obtained from the subsequent printing inputs are entered into the second memory C.

Also, instead of sending the digital electrical signal to memories F and C immediately, it is sent from the A/D converter E to the gradation sensitivity corrector P, where the gradation sensitivity is corrected and then stored in the first memory F and second memory 〇. You can also do that.

The shading sensitivity corrector P corrects the discrepancy between the comparison of the input value by the sensor 10 and the comparison by human vision, and is used to compensate for differences in the material of printed matter (paper, metal, etc.), screen shape, and other shading configurations. In order to accommodate this, it is preferable to be able to arbitrarily set a large number of correction characteristics depending on the intended use.

The gray scale sensitivity corrector P will be explained with reference to FIG. The vertical axis shows the black and white level (that is, the amount of received light), and the horizontal axis shows the electrical signal (voltage in this example). Line R indicates the density sensitivity characteristic of the light receiving element 18 (this is an ideal one,
(Actually, there is an error), and the black/white level and voltage are directly proportional and form a straight line. but,
Since human shading sensitivity characteristics vary depending on the type of printing and other factors, they are usually different from the shading sensitivity characteristics of the light receiving element 18. Therefore, a gradation sensitivity corrector P is provided to correct the gradation sensitivity characteristics as shown in the correction curves 81 to S4, for example.

Taking the correction curve S3 as an example, a positive correction is made to the line R of the sensitivity characteristic of the light receiving element 18 between T1 and T2, a negative correction is made between T2 and T3, and in the intermediate region between black and white levels, The difference between 8I and light is made to appear as a larger voltage difference. In other words, the correction curve S
3 is suitable for detecting printing defects with high sensitivity in intermediate density regions.

The correction surfaces 8981 and S4 are suitable for detecting printing defects with high sensitivity in dark and light areas close to black and white, respectively. The correction curve S2 is suitable for detecting printing errors with high sensitivity in gray areas close to both black and white.

Now, returning to FIG. 2 and explaining, each digital electric signal corresponding to the first memory F and second memory C is taken out and input to the arithmetic circuit H. This arithmetic circuit H extracts the absolute value of the difference between the corresponding digital electric signals of the first memory F and the second memory C as a voltage difference. This voltage difference is
The comparator circuit J compares the pass/fail level set value from the pass/fail level setter I for setting an arbitrary pass/fail level according to the print content in advance, and the digital electric signal in the first memory F and the digital electric signal in the second memory 〇 are compared. If the difference with the digital electric signal is equal to or greater than the pass/fail level setting value, it is determined that it is an error point. When it is determined that it is an error point, only the result of failure is entered into the counter K. If it is good, it is not counted.

The pass/fail levels to be set include a detection 11 pass/fail level, an ink density pass/fail level, a point limit pass/fail level, an error limit pass/fail level, and others.

The detection width pass/fail level is a pass/fail level set above and below the center of the reference, and when the set value above and below is exceeded, it is determined as an error point.

The ink density pass/fail level is a pass/fail level for changes in ink density. When updating the standards to
It is desirable to set the ink Ii1 degree pass/fail level independently of the detection width pass/fail level.

The point limit pass/fail level is the pass/fail level of the displacement of the printing surface A in the front-rear direction with respect to the sensor 10. This point limit pass/fail level is set when the accuracy of the printed matter feeding mechanism is low. If the set value of the point limit pass/fail level is large, the detection judgment will be lenient.

The error limit pass/fail level is the pass/fail level of the deviation of the printing surface A in the left-right (horizontal) direction with respect to the sensor 10. The setting value is determined particularly in relation to detection accuracy.

For example, as shown in FIG.
Detection width pass/fail levels are set above and below 0, and electrical signals 201 and 202 of the detection width pass/fail level set value are set. Set. After that, the printed surface is detected and the electrical signal 205 obtained thereby is set at the detection width pass/fail level 1i12 at each measurement point SE.
02, if the ink density pass/fail level set value 204 is exceeded at the measurement point SG, an error point is issued in either case.

For example, the difference between the pure black level and the pure white level is divided into 256 levels, and the contents of the first memory F and the contents of the second memory C are compared to find the difference.

Such operations are carried out in the first memory F and the second memory C.
Align the order of all the digital electrical signals stored in the , and compare them in order from the starting edge to the ending edge of printing surface A, the first one, the second one, the third one, and so on. point). Then, the number of error points is counted by a counter.

When the contents of the first memory F and the second memory 0 are completely compared, the data in the counter is input to the display circuit as shown in FIG. and the total number of measurement points (
m numbers) and the actual count on the counter. Furthermore, a percentage as a discrimination standard is set in advance by the defect setting device U for pass/fail discrimination, and the calculation result of the arithmetic circuit and the preset percentage (
Finally, it is determined whether the printing on the entire printing surface is "good" or "defective". At that time, only the judgment result of "good" is counted in the force rank M.

Note that the defective rate setting device U for pass/fail determination may be configured to allow all values with an error point count of 1 or more to be set, but normally the defective rate can be set arbitrarily within the range of 0.1 to 100%. to be set.

As described above, the pass/fail of the printed surface is determined based on the pass/fail level and defective rate of the pass/fail criteria.

On the other hand, an arbitrary examination level is set in advance by an examination level setting device V for examination determination. The detected value of the printed surface and its preset examination level are compared in a comparator circuit J, and the number of error points is counted. When all the detected values in a predetermined range of the printing surface are compared, the data on the counter is input to the arithmetic circuit. Then, the percentage of the total number of measurement points and the actual count on the counter is calculated. An arbitrary inspection rate is set in advance by an inspection rate setting device W, and the set value is compared with the calculation result of the arithmetic circuit to determine whether the printed surface is good or bad. At that time, only the good judgment results are counted by the counter M.

In this way, the printed surface is judged based on the examination level and examination rate of the examination judgment criteria, and when it is determined to be good, the detected value of the printed surface and a certain width around it are updated as the new standard. N is sent to the first memory C and additionally registered as a new pass/fail determination criterion.

In the example described above, the defective rate and the examination rate were used as the pass/fail determination criteria and the examination determination standard, but the number of failure points and the number of examination points can be used instead of such percentages. Particularly when the number of measurement points is relatively small, for example in the case of collation error detection, it is desirable to use the number of defective points and the number of inspection points instead of using percentages. Table 1 shows such an example.

In Table 1, symbols A to D indicate ranks that should be selected depending on the type of printed content on the printing surface. For example, the printed content of palanque includes printed content of only fine print. The 8th rank is the printed text in medium letters, the C rank is the gravure print content, and the D rank is the picture print content. The numbers at the pass/fail level and the examination level indicate the number of stages, that is, the number of levels when dividing the range between pure white and pure black into 256 stages. The number of defective points and the number of inspection points indicate the number of times a product is determined to be defective.

Taking the detection values shown in FIG. 12A as an example, the symbols P1 to Plo indicate the positions of the ten light receiving elements 18. For example, when eight ranks in Table 1 are selected in advance, the pass/fail level is 5, the number of defective points is 1, the scrutiny level is 2, and the number of scrutiny points is 2.

Therefore, the printed surface (1) shown in FIG. 12A is judged to be good. Although the printed surface (2) is determined to be defective, it becomes a new pass/fail determination criterion. The printed surface (3) is determined to be defective,
Moreover, it does not become a new pass/fail judgment standard.

Figure 12B is similar to Figure 12A, but the concept of pass/fail level and scrutiny level is slightly different.

Still another embodiment of the invention will be described with reference to FIGS. 4-8.

In FIG. 4, the vertical axis indicates the black and white level of the printing surface, that is, the amount of reflected light. Line X indicates the level of the pure white printing surface,
Line Y indicates the level of the completely black printed surface. (A), (B), (C), . . . (G) are a series of measurement points along the movement direction of the printing surface A by one light receiving element 18 by the encoder 84 (FIGS. 7 and 8). It shows the position of. Since the same measurement points are taken for each light-receiving element 18, a large number of measurement points are distributed overlappingly over the entire printing surface at regular intervals vertically and horizontally as shown in FIG.

For convenience of explanation, only one light-receiving element 18 will be described. In FIG. 4, the signal 100 represents an analog electrical signal of the preceding printing surface, and the signal 100a is an analog electrical signal obtained by adding a pass/fail level M' to the signal 100. Yes, signal 10
0b is an analog electrical signal obtained by subtracting the pass/fail level M' from the signal 100. The subsequent printed surface is measured and the above-mentioned comparison judgment is performed, and if the measurement point does not fall between Shin@100a and Shin@100b, it is determined that it is an error point.

Now, even though the subsequent printed surface is a good print, the amount of reflected light is low due to differences in the material of the printed surface! ! may be detected as an error point. In order to avoid such discrimination errors, the correction level α is automatically added and/or subtracted from E1′ (or by operator adjustment) to obtain E+ −(E+ '±α)〈±M, and then the comparison is performed. , the difference in the amount of reflected light is corrected and the judgment is made. The value of α can be determined as the difference between the reference average value and the detected value average value. If we use this value of α to shift the detected value in parallel to correspond to printing fluctuations, we get
It can reduce memory capacity and improve stability.

In addition to performing parallel translation for one set of light-emitting and light-receiving elements as described above (both in static and dynamic detection methods), it is also possible to translate multiple sets of light-emitting and light-receiving elements arranged on one or more sensors. The detection values of the light-emitting element and the light-receiving element can be moved in parallel as a whole to correspond to printing variations.

For example, FIG. 13 shows an example of such a translation method. The quality of the printed surface is determined after the detected value average value is shifted by the difference α between the reference average value and the detected value average value. In the example shown in FIG. 13, the overall average value of the detection values detected by a plurality of light receiving elements is taken, and after being moved by the difference α between the reference average value and the detection average value, The quality of the printed surface is determined by comparing the detected value with a reference.

The same thing can be said about the detection value detected by one specific light-receiving element (for example, regarding the hatched part A in Fig. 9), which is the difference α between the detection value average value and the reference average value.
You can also compare the detected value with the reference after moving by that 0 minute.

When the paper quality of the subsequent printed surface A is dark, even if the printed content is exactly the same, the signal 101 of the subsequent printed surface approaches the line Y as a whole due to the influence of the paper quality. Signal 1 as it is
If 01 is compared with signal 100, there is a possibility that the printed surface will be erroneously determined as a printing error, so a correction level α, which is the difference between the average values of the two, is added to signal 101, and signal 101 is compared with signal 100.
’ and then compare.

Such a pass/fail level mainly corresponds to paper quality, printing unevenness, and sensor drift, but the point pass/fail level described next mainly corresponds to the encoder 84 (7th to 8th points).
This is to cope with the misalignment between the printed surface A and the printed surface A.

Regarding one measurement point, the above-mentioned pass/fail level is the pass/fail level (tolerable value) in the vertical direction, and the point pass/fail level is the pass/fail level (tolerable value) in the horizontal direction.

If the pass/fail level and the point pass/fail level are set to exactly the same value, the tolerance range will be a square centered on the reference point. Of course, the pass/fail level and the point pass/fail level can be made independent of each other.

In that case, the two values will be different and will fall within the permissible range of the rectangle.

Now, in either case of pass/fail level only (Figure 4) or a combination of pass/fail level and point pass/fail level (Figure 5), if the sampled data exceeds the plus or minus pass/fail level setting value, it is determined as an error point. do.

Then, use one (or both) of the following methods to determine a printing error.

Judgment method (1): If the number of error points among all measurement points exceeds a set value (this corresponds to a set value of defective rate), it is determined that there is a printing error.

Judgment method (2): First, a location with consecutive error points is defined as an error block, and if the next calculated value exceeds the defective rate setting value, it is determined as a printing error.

L+S/P L is the peak value of the error block S is the sum of the levels of the error block P is the number of error points of the error block Next, with reference to FIG. 7, another more specific embodiment of the method of the present invention will be explained. explain.

The sensor 10 has a configuration similar to that described above and includes light receiving elements 18 of N091...No, n.

The sensor 10 reflects light from each of a large number of light emitting elements 21 onto a stationary or moving printing surface A, and receives the reflected light by each corresponding light receiving element 18 and converts it into an analog quantity of electricity.

The output of each light receiving element 18 is amplified by an amplifier 72. The amplifier 72 is preferably a buffer amplifier that performs impedance conversion in order to prevent noise and the like from entering.

The amplified analog electrical signal is sent to an AD converter 71
sent to. Although it is desirable from the viewpoint of detection speed to provide various components after the AD converter 71 for each light receiving element 18, in the case of low-speed printing, etc., a multiplexer 79 may be provided and used together.

When detecting the printing surface A while moving it, the AD converter 71 samples an analog electrical signal and converts it into a digital electrical signal in response to a measurement pulse from the measurement time controller 82. The measuring time controller 82 is activated by signal pulses from a starter 83. The starter 83 is installed so as to be linked to the power of the rotary press, the paper surface, etc., and generates a signal pulse at the same time as printing starts. Sampling is encoder 8
This is done in a time-division manner using pulses from 4. The encoder 84 can be attached to the power shaft of the rotary press or a part linked thereto.

In the AD converter 71, the analog signal is converted into a digital signal, and the digital signal is sent to the gray scale sensitivity corrector 73, where gray scale sensitivity correction is performed (as already explained with reference to FIG. 11). , which are further stored in the storage circuit 75 for each measurement point according to instructions from the memory controller 74.

The memory circuit 75 can be roughly divided into a reference memory 75a and a detection memory 75b in terms of their roles. The reference memory 75a stores reference information, and the detection memory 75b stores information on detected values of subsequent printed surfaces.

There is no functional difference between the reference memory 75a and the detection memory 75b. Therefore, if necessary, the configuration (
) may be used.

The contents of the memory 75b are compared with the contents of the memory leaf 5a by the operator 81. If the difference is greater than or equal to the pass/fail level setting value, it is determined as an error point.

That is, information on the first measurement point (hereinafter referred to as E+) on the previously printed surface of the memory 75a is derived to the computing unit 81. Subsequently, the first printing surface of the subsequent printing surface in the memory 75b
The information of the measurement points 1 to 1 (hereinafter referred to as E1') is sent to the computing unit 8.
It will lead you to 1.

Then, the arithmetic unit 81 determines whether E+' is a defective point by determining the absolute value of (E+ - E+') multiplied by M.

Among the set values of the pass/fail level M, the set value of the pass/fail level during detection can be arbitrarily set by the detection width pass/fail I novel setter 90. In this case, the pass/fail level and the point pass/fail level may be set to the same value, or they may be set to different values.

The setting value of the ink density pass/fail level can be arbitrarily set by the ink density pass/fail 1 novel setter 92.

In order to provide flexibility in the determination, a pulse of 11 il is sent from the computing unit 81 to a defective rate or defective point number detector 51 (hereinafter simply referred to as defective detector). The defect detector 51 counts the number of pulses. Next, the second printed side
Information on the second measurement point (E2) and the second measurement point on the subsequent printing surface.
The information (E 2') of the th measurement point is compared in the same way to determine whether E2' is an error point signal.

Such operations are performed for all measurement points.

When all measurement points have been compared, the defect detector 51 stores the number of error point signals.

When using the defect rate, it is determined whether printing is defective or not based on the ratio of the number of error point signals to the total number of measurement points. In other words, the determination is made based on the number of error points among all points. When using the number of defective points, compare the determined total number of defective points with the set number of defective points.

In any case, the defect detector 51 ultimately determines whether the printing on the entire printing surface is "good" or "bad." At that time, [only the judgment result of good J is counter 95]
is counted.

The failure rate or the number of failure points is arbitrarily set by the failure setting device 42 for pass/fail determination.

For example, when the failure rate setting device 42 for pass/fail determination is set to a failure rate of 0.2%, and if the total number of measurement points is 1000, the limit for the number of error points is 2 points. These two pieces of information are passed from the failure setting device 42 to the failure detector 51 for pass/fail determination.
sent to. Then, the error point signal is compared with the information sent from the failure setter 42 for pass/fail determination.

On the other hand, the scrutiny level is set to an arbitrary value using the scrutiny level setter 90a for scrutiny discrimination. Further, the ink density examination level setter 94 sets the ink 81 degree examination level to an arbitrary setting value. The arithmetic unit 81 compares the set value of the examination level with the detected value to determine whether it is good or bad. In the defect detector 51, there is a setting device for the number of examination points or examination rate (hereinafter simply referred to as the examination weight setting device) 4.
2a, the inspection rate (or the number of inspection points) arbitrarily set according to the print content is compared with the detected value to make an inspection judgment, and only a good judgment result is incremented by one at the counter 95.

As described above, the printed surface is first determined to be pass/fail using the pass/fail criteria, and then the printed surfaces that are determined to be defective based on the pass/fail criteria are further examined and judged based on the criteria. When the judgment is made, the updater 96 updates a new pass/fail judgment criterion corresponding to the detected value of the printed surface.
The information is additionally registered in the reference memory 75a of the storage circuit 75.

Further, the printing error determination result can be displayed on a display (not shown), an alarm can be issued (not shown), and printed surfaces determined to be defective can be self-ejected.

In addition, as shown in FIG. 8, the defect rate setting device 42 and the examination weight setting device 428 for pass/fail determination in FIG.
1 into one to calculate the above-mentioned value of (L+S/P).
). If the arithmetic unit 50 is excluded, the eighth
The embodiment shown in the figure has the same configuration as the embodiment shown in FIG. 7, so a description thereof will be omitted.

In any of the above-described embodiments, the present invention has a plurality of standards, and the detected value is compared with the plurality of standards to determine whether it is good or bad. Moreover, since the detected value is translated in parallel if necessary and then compared with the reference, more accurate judgment can be made.

Also, standards can be increased in various ways. For example, to give an example of how to create the first standard, a re-memory switch is provided, and when the re-memory switch is turned on, the standard based on the first detected value of the printed surface is set as the original standard. can do.

The next printing surface is detected according to the standard.

At that time, when it is determined to be defective by the pass/fail criteria, and when it is determined to be good by the scrutiny criteria, the detected value of that printed surface is updated and registered as a new criterion, and the number of new criteria to be added is set in advance. When the number exceeds the specified number, the oldest registrations can be deleted and the standards can be updated sequentially. In addition, when it is judged as good by the pass/fail criteria, the criteria corresponding to the detected value will be added and registered as new criteria, but when the number of new criteria added exceeds the preset number, the old criteria will be registered. It is also possible to update the criteria by deleting them one by one while maintaining a constant number of criteria.

Taking into consideration the storage capacity of the memory, etc., it is desirable to set the upper limit of the number of standard registrations to around 10. Of course, the present invention is not limited to such numerical values, but according to experiments by the present applicant, good results in terms of cost were obtained when the standard number of registrations was set to around 10.

The flow of the printed matter detection method of the present invention will be briefly explained with reference to FIG.

First, start the printed matter detection device.

The detected value of the first printed surface and the value within a certain range from there are registered as pass/fail determination criteria. The next printing surface is detected based on the reference. First, the printed surface is detected according to the criteria for pass/fail classification. If it is determined to be good, the standard corresponding to the detected value of the printed surface is additionally registered as a new standard.

If it is determined to be defective, parallel movement comparison is performed.

If it is determined to be good, the ship will be registered as a new standard. If it is determined to be defective, it will be judged based on the criteria for examination and discrimination. If it is determined to be good, a standard corresponding to the detected value of that printing surface is additionally registered. If it is determined to be defective, a defect signal is output. The printed surface is then visually inspected. If it is determined to be defective, remove the printed surface, release the device, and then repeat the same process again.If the visual inspection determines that it is good, press the Restart 1- button and repeat the same process as above. The process is repeated. When the number of registered standards exceeds a set number (for example, 11), the old ones are sequentially deleted and new standards are added and recorded.

1. Since the present invention is configured as described above, it is possible to detect the printed surface extremely accurately compared to the conventional single standard.

Especially when there are variations in the printed surface, the printed surface can be detected very effectively.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram showing the principle of the printed matter detection method according to the present invention. FIG. 2 is a nine-dimensional diagram for explaining an example of an apparatus for carrying out the method of the present invention. FIG. 3 is a waveform diagram showing the flow of the detection signal and the level at each detection point. FIG. 4 is a diagram illustrating the principle of an apparatus for implementing the printed matter detection method according to the present invention. FIG. 5 is a diagram showing the relationship between standards and level pass/fail novels. FIG. 6 is a diagram showing the relationship between the 1 standard li, the level pass/fail level, and the point pass/fail level. FIGS. 7 and 8 are Zorock diagrams showing other embodiments of the printed matter detection method of the present invention, respectively. FIG. 9 is a diagram showing the relationship among the printed surface, the light receiving element J5, and the measurement point. FIG. 10 is an explanatory diagram of the interrelationship between the M standard, the ink density pass/fail level, and the τf detection width pass/fail 1 novel. Figure 11 is the origin of gray scale sensitivity correction]! Figure 1 shows this. FIG. 12 shows one specific example of implementing the printed matter detection method of the present invention, and is a graph showing the relationship between pass/fail determination and inspection determination. FIG. 13 is an explanatory diagram showing a parallel shift detection method according to the present invention. FIG. 14 is a diagram showing the flow of the printed matter detection method of the present invention. FIG. 15 is an explanatory diagram showing the basic concept of a conventional printed matter detection method. 10・・・・・・・・・・・・Sensor 18...
......... Light receiving element 21...
......Light emitting element 72...
・・・・・・Amplifier 73・・・・・・・・・・・・・・・
AD converter 74・・・・・・・・・・・・・・・
Memory controller 75a...
・Reference memory 75b・・・・・・・・・Detection memory 84・・・・・・・・・・・・Ennii-
1; r-50・・・・・・・・・・・・ Arithmetic unit 9
5・・・・・・・・・・・・Counter 96...
・・・・・・・・・・・・Updator A ・・・・・・・・・
・・・・・・Printing side D ・・・・・・・・・・・・・・・
・・Amplifier E ・・・・・・・・・・・・・・-A/D
Converter F ・・・・・・・・・・・・・・・1st
Memory C・・・・・・・・・・・・Second memory H・・・・・・・・・・・・Arithmetic circuit ■ ・・
・・・・・・・・・・・・Pass/fail level setter J ・・
・・・・・・・・・・・・Comparison circuit K ・・・・・・
・・・・・・・・・Counter L ・・・・・・・・・・・・
・・・・・・Arithmetic circuit M ・・・・・・・・・・・・・・・
・Counter N ・・・・・・・・・・・・Update unit P ・・・・・・・・・・・・ Gray sensitivity corrector Q
・・・・・・・・・・・・・・・Encoder U ・
・・・・・・・・・・・・・・・Defect rate setting device■ ・・・
・・・・・・・・・・・・Scrutiny level setting device W ・・・
・・・・・・・・・Examination rate setting device agent
Patent Attorney Tanabe Withdrawal Table 1 Pass/Fail Judgment Criteria Examination Judgment Criteria Rank
Pass/fail level Number of defective points Examination level Number of examination points A 3 0
2 1B 5
1 2 2C7172 D 11 1 14
2 Fig. 1 Fig. 4 1oob Fig. 6 Fig. 9 Fig. 10 Fig. 11 Voltage - Fig. 13 PI P2 p3 p4 -----1 14
Figure d Figure 15

Claims (7)

[Claims] (1) Shine light on multiple measurement points on the printing surface, detect the amount of reflected light at all points, and set plus and minus allowances at all measurement points based on the detected values, and then use it as a reference for subsequent measurements. The amount of light reflected from the subsequent printed surface is detected by shining light on a plurality of corresponding measurement points on the printed surface, and the reference and the detected value of the subsequent printed surface are compared for each measurement point to detect printing errors. In the method for detecting printed matter, the criteria include a pass/fail criterion and a scrutinized criterion that is less strict than the pass/fail criterion. A printed matter detection method characterized by additionally registering a new standard based on a detected value of the printed surface. (2) The printed matter detection method according to claim 1, in which the pass/fail discrimination criteria and the examination discrimination criteria are arbitrarily set in advance. (3) The printed matter detection method as set forth in claim 1 or 2, wherein the pass/fail determination criteria consist of a pass/fail level and the number of defective points, and the scrutiny determination criteria consist of a scrutiny level and the number of scrutiny points. (4) The printed matter detection method as set forth in claim 1 or 2, wherein the pass/fail discrimination criteria consist of a pass/fail level and a defective rate, and the scrutiny discrimination criteria consist of a scrutiny level and a scrutiny rate. (5) When the number of new standards added exceeds a preset number, the standards are updated by deleting them in the order of oldest registration. printed matter detection method. (6) The printed matter detection method according to any one of claims 1 to 5, in which the criterion is not updated when the criterion is determined to be good by the pass/fail criterion. (7) Even when the pass/fail criteria determines the pass/fail criteria, new criteria are additionally registered based on the detected value of the printed surface, but when the number of new criteria added exceeds the preset number. The printed matter detection method according to any one of claims 1 to 4, in which the standards are updated by deleting the oldest registered items.