EP0741026A2 - Image inspection and ink supply method in printed products of a printing machine - Google Patents

Abstract

An automated process is used in which the image data obtained from camera scanning of the printed output of a printing process is fed to a computer system for comparison with a set of reference data. The difference data is processed and the location of the fault on the printed sheet is indicated on a display screen. There are a large number of printing faults that can be due to a large number of different factors. The relationship between the faults and the factors causing them are defined within an information matrix and this is used for rapid fault identification analysis.

Description

The invention relates to a method for image inspection and color guidance on print products of a printing press, the actual image data of the print images of the print products preferably being determined in online operation and compared with target image data for error detection.

A method of the type mentioned at the outset is apparent from US Pat. No. 5,187,376. By means of the method, the print images of the continuously produced print products are examined for errors in the ongoing production operation by means of an optical detection device. This is done by obtaining actual image data from the print images of the print products and comparing them with target image data. The target image data originate from an error-free image original, so that an error can be inferred from the comparison mentioned when a deviation occurs between the actual and target data. A deviation that occurs indicates that there is an error, but it is fundamentally not clear what type of error it is. If various available measures were taken to correct the error without the exact type of error being known, this can lead to problems in the production run.

It is also known that if a fault occurs, sheets of a sheet-fed printing machine which are faulty are separated, which can be done, for example, by means of a switch or by shooting in strips. Although these known measures lead to error identification, they do not include an error analysis.

Based on this basic problem, the invention has for its object to provide a method for image inspection and color guidance on print products of a printing press, in which a purposeful and reliable in terms of the printing process error detection or error correction takes place.

This object is achieved in that if an error occurs before a change in the color guide is checked whether a cause other than the color guide is possible for the deviation due to the type of error. If a deviation of the image data that indicates an error is determined by the target / actual comparison, then a corrective measure is carried out in a targeted manner in such a way that there is a high probability of being excluded before a change in the color guidance, in particular the implementation of a color control can that the determined deviation has a cause other than the color guide. All known printing errors, such as slugs, registers, colors, dampness, stenciling, ions, sliding, duplicating, transfer registers, etc. come into consideration as other causes of error. Furthermore, operating errors such as changing the paper type, too little color in the ink fountain and device errors such as dirty optics, fiber breakage of the optical fiber, lamp defects or voltage drops in electrical equipment can also be responsible for the determined deviation between the target and actual values. In all of these cases, the color must not be regulated, for example, an error message can be sent to the printer or the corresponding error is automatically eliminated by suitable measures. If a color deviation is detected in the target-actual comparison, this can of course be due to incorrect color guidance (wrong setting of the zone opening of the inking unit), but the other errors mentioned above are not caused by this Adjustment of the zone opening or zone openings can be eliminated. If the color control tries to compensate for this pressure disturbance by changing the zone opening, the effects cannot be foreseen; they can lead to instability in the control process, but without the elimination of errors. In the method according to the invention, in the case of online image inspection, it is ensured before any color guide correction that may be necessary that all possible other causes of the fault can be excluded. Only when it has been ensured that other causes of the problem are out of question, is the color control activated, for example by adjusting the color zone.

According to a further development of the invention, it is provided that a type of fault analysis is carried out, in which different, predetermined criteria are used in succession for fault type detection, preferably systematically. Certain criteria are therefore specified for each type of error, which are used for systematic troubleshooting, so that the type of the present error can be predicted with a high degree of probability. If the type of error has been determined, a corrective action can definitely be taken.

It is also advantageous that an error correction measure is carried out in accordance with the determined type of error. If there is a high likelihood of a certain error, only the intended, specific measure is taken to remedy the situation.

If the type of error is ambiguous, the type of error that has the greatest probability is assumed to be present. So if several of the above-mentioned criteria are met when determining the type of fault, this may result in a clear determination of the type of fault difficult combinations of criteria must be used. Certain combinations of the criteria provide information about the type of error, whereby additional aids such as the probability calculation can be used to achieve the greatest possible accuracy in determining the type of error.

Finally, it is advantageous if a change in color guidance, in particular activation of the color control, is carried out last in the case of a plurality of possible types of error for troubleshooting. An intervention in the color control is therefore not only the last measure in the event of an unambiguous determination of the error, but also the last action if an ambiguity cannot be ruled out based on the error type analysis.

Fig. 1

ein Flußdiagramm der dem Verfahren zugrundeliegenden Verfahrensschritte und

Fig. 2

eine Fehlerermittlungsmatrix.

The drawings illustrate the invention with reference to the figures, which show:

Fig. 1

a flow chart of the process steps underlying the method and

Fig. 2

an error determination matrix.

In the course of the progressive automation of the printing process in printing machines and the introduction of quality standards, automatic image inspection in the printing machine is becoming increasingly important in the production process. By means of an optical detection device, which can be arranged, for example, in the delivery or in the area of the impression cylinder of the last printing unit of the printing press, the continuously freshly printed sheets are detected with regard to the printed images of the printed products, actual image data being determined and being fed to electronics, in particular a computer . Aim this The procedure is to identify defective sheets. For this purpose, the determined actual image data are compared with target image data. The target image data are stored in the computer. They represent an ok sheet, i.e. an error-free printed copy.

If a deviation is found in the target / actual comparison, it can be determined on the one hand which sheet or which sheet has this deviation and on the other hand the location of the deviation in the printed image can also be traced, i.e. the defective sheets and also the locations of the defects can, for example, be revealed to the printer be shown on a display. Possible causes of errors can be: slugs, lint, registers, sliding, duplicating, paper type, paper color, lighting, glass fiber defects, distance between sheet and measuring bar of the optical detection device, moisture, no or too little color, paper transport, no paper, dirty optics, stripes, temporary Color changes, for example due to changes in speed, etc.

The invention is based on the fact that certain measures can be uncritical when eliminating errors and others can be very critical for the printing process, so that the critical troubleshooting measures can only be carried out if it is ensured that another type of error is not present. Color management is critical in the printing process, so that color management changes are the last measure to be taken when troubleshooting.

A systematic procedure is required for the error type analysis.

The causes of errors and their typical appearance are given below.

Register deviation

• on the whole sheet
• solid color on solid edges
• multicolored on solid edges
• in a multi-color grid
• not in full tone
• not in the single color grid
• in the differential image as a gradient image of the target image

Humidity error

• For example, in a 50% to 80% grid due to dot gain (smear)
• under coloring on solid edges (water noses)
• reinforced at the start of printing
• globally even on non-printed surfaces
• preferred places, usually at the beginning of the print and here mostly to the side
• no rhythmic appearance

Slugs

• very limited error impact (no expansion)
• is a one-off event or a rare event (lint on sheet)
• occurs suddenly, the front sheets are not affected
• Fault location first shows under-coloring and then over-coloring of a color (lint on rubber blanket)

Optics dirty or defective optical path (the optical detection device)

• errors can be found in non-image areas and in the image
• Continuous errors in the printing direction (line systems) limited locally to this

paper

• in non-image areas
• across the entire width
• in a grid (small area covers)

Slide / duplicate

• in the grid
• not in full tone
• over the entire arch
• global
• not on unprinted areas
• no preferred positions
• depending on the area coverage combination
• Color value
• rhythmic appearance in time

Color guide errors

• in the solid field
• in the grid
• as a zonal change in color value
• in the mixed field (for example gray field) as a difference vector in the direction of the corresponding solid color
• build up slowly
• usually not across the entire sheet width

Stripes

• certain distance
• certain period
• Expression depends on the printing speed.

In the case of the type of fault analysis, a systematic processing of criteria and an associated elimination process are preferably carried out in order to reduce the number of possible sources of error until ideally only one error remains.

Another possibility is to link the type of error and the measuring location or further variables or auxiliary variables in a matrix (FIG. 2). Once the matrix has been created, the probability of a particular type of error can be determined and in this way the error that has the greatest probability of occurring can be determined.

As an alternative or in addition to the above, it is also possible to include a temporal component in the error type analysis, for example by capturing the image data or the difference values over several sheets, so that their chronological history flows into the error assessment.

The following shows options for identifying the error with the greatest possible probability by means of mutual error analysis or plausibility checks.

The flow chart shown in FIG. 1 illustrates the procedure for the type of fault analysis. It can be seen that whenever a criterion is met, the next type of error is passed on until finally - as the last measure - a color control measure is taken. It is assumed that the printed images of the printed products are mentally divided into zones according to the zonal division of the inking unit, with mental separations taking place transversely to these zones, so that individual fields arise. Measurement locations of the optically working detection device. If a color field reports a color deviation during production due to a deviation in the mentioned target-actual comparison, then - before the color control is activated - it must be excluded with a high degree of probability that the change in the measured value has a cause other than the color guide. The other causes of the error have already been explained above. The other operating errors such as changing the paper type, not enough color, etc. have already been dealt with and device errors have also been mentioned. In all of these cases, which do not concern the color guidance, the color must not be regulated in the event of an error, but an error message must be sent to the printer and / or an error remedy must be carried out manually or automatically. Since the measured values cannot be seen a priori as to the cause of a change in the measured value, an attempt must be made to determine the cause by evaluating auxiliary measured values and auxiliary color measurement fields.

a. Einzelmeßwerte:

grad > 3 dE von Bogen zu Bogen

b. Einzelmeßwerte:

grad > 5 dE nach 8 Bogen

c. Mittelwerte :

grad > 2,0 dE nach 5 Bogen.
(aus 16 Bogen gleitendes Mittel)

According to FIG. 1, it is determined in step 1 by means of the optical detection device that there is a color deviation in one or more specific color measuring fields. In step 2, a check is carried out for a paper white field. In order to detect deviations due to changes in the paper surface, fluctuations in the lighting, glass fiber breakage or changes in distance, a color measuring field must be defined on the print image that monitors an unprinted paper spot across the entire width of the zone. If there is a significant change in the mean value or the standard deviation of the image data in this measuring field, the further color measuring fields in the same zone will also show these changes, which, however, are generally not due to a change in the color guide. If it is found in step 3 that there is no error, then the process continues to step 4. If an error is found, it is displayed according to step 5. In step 4 there is a check for scatter in the color measurement field. If the scatter changes by more than 20%, in particular 50%, in the current color measurement field to be controlled, a color control measure, which of course also takes place in color-free operation, must be excluded. Changes in the color routing generally have little or no influence on the scattering of the pixel values within the color measuring field. If the scatter changes significantly, this can be attributed to error causes such as slugs, lint or the like. The permissible change in the scatter depends on the size of the measuring field. The smaller the measuring field, the greater the permissible spread. The test is carried out in step 6. If there is no error, go to step 7. In the event of an error, a message is issued according to step 8. In step 7 the time gradient in the color measurement field is checked. The color scheme is sluggish and color changes are slow. The maximum measured color gradient in the event of a brief sharp increase (dead beat) when driving in the color is on average 0.4 dE per sheet (dE = delta-E). This means that if there are color changes greater than 0.5 dE per sheet, the color guide must be excluded as the cause. However, since statistical process and measurement fluctuations of the measured values are of the order of magnitude from sheet to sheet, no direct comparison sheet to sheet can be made here, but several sheets must be evaluated and different methods combined:

a. Single measurements:

degrees> 3 dE from bow to bow

b. Single measurements:

degree> 5 dE after 8 sheets

c. Mean values:

degree> 2.0 dE after 5 sheets.
(from 16 sheets sliding average)

In step 9, the evaluation is carried out, in the event of an error according to step 10, a message is given or, if there are no errors, a transition is made to step 11. In the course of step 9, it is also possible to check whether the measured image data periodically scatter in the time range, for example that every second or third sheet has an error, this can be caused by duplication. As a result, the mean values remain relatively constant, but the color control could be called up due to single value tolerance violations. Since such effects are noticeable in a significant increase in the scatter, for example the last 16 or 64 measured values, such a cause of error can be excluded or identified by checking this scatter. In step 11, the field of a zone is compared with similar fields of the same zone. If fields are located in the same zone that have a similar or identical color composition, they must also show the same tendency for the control deviation. If, for example, a solid cyan field reports too little color in a zone, the corresponding gray field in the same zone must also report a cyan deviation - apart from a factor. In the simplest case, when searching for the measuring field, a quasi-identical auxiliary field is defined for each control field, which has a certain distance from the control field in the printing direction. With gray field control in standard printing, the full-tone auxiliary fields can serve this purpose. If an error is determined in accordance with step 12, this is reported in step 13. If there are no errors, go to step 14. Step 14 relates to the comparison of fields of one zone with fields of neighboring zones. Since measurement / control fields are selected for each zone and for each color, the equivalent fields of one or more neighboring zones can be used for checking in the event of an error message in one zone. A color deviation in one zone must be observed in a weakened form in the neighboring zone.

If there is an error of this type, this is determined in step 15 and displayed in step 16. If there are no errors, go to step 17. In step 17, the color change vector is checked (logic of the adjustment). For the control, the rough composition of the control field must be known, i.e. the area coverage of the colors involved in the printing must be determined. Before the color control is called, it should be checked whether the change in color values can be caused by the colors in the field. If an error is determined in step 18, the message is given in step 19. Otherwise, the process proceeds to step 20. Step 20 relates to checking the register. A register adjustment causes short-term duplication with large changes in the measured value in the grid, whereby the output color value is reached again after a few sheets. Because of the large color gradient, this error is already intercepted by the temporal color gradient, but rough register monitoring should nevertheless take place. This can be done in particular by means of an auxiliary color measuring field on a solid edge (per color). This register check takes place in step 21. If there is an error, this is indicated in accordance with FIG. 22.

On the other hand, a transition is made to step 23, which activates the color control if there is no error to date, that is to say if no error has been found in the preceding steps (step 24).

In summary, it should be noted that the cause of the error is automatically analyzed and output or corrected for a quick rectification of sources of error and to avoid waste. The knowledge available from the process engineering know-how of the printing process is implemented in search strategies and evaluation algorithms, their sequential application or parallel linkage to allow a conclusion on the cause of the error if an error occurs.

FIG. 2 shows a matrix as has already been defined above. Possible errors are shown on one axis and phenomena occurring on the other axis during the measurement. In order to analyze an error with a sufficiently high probability, the ascertained phenomena are preferably automatically evaluated. A cross in the matrix means that this phenomenon has occurred. A dash means that the phenomenon is not present. The greater the number of phenomena to be assigned in the case of an error, or the more reliably a phenomenon can be linked to an error, the greater the likelihood that the correct error will be determined. Furthermore, it can be said that with an increasing number of relevant measured variables in the matrix, an error assignment can be made more clearly and reliably.

The line "hardware error" is only to be understood as a placeholder for various hardware components that occur in the evaluation electronics and whose defect causes changes in the measured values. Since not all hardware components can be listed, the principle will be shown here using two examples:

Example 1: (defective CCD element)

The measurement is carried out in three color channels (X, Y, Z), each with a CCD element; if one of them fails, this is noticeable for all measured values, ie a cross should be entered in the "Hardware" line in each box in the matrix, equivalent to the "no paper" line. The matrix must therefore be expanded to include the columns: X values, Y values, Z values, because a CCD defect is only noticeable in one of these columns.

Example 2: (Preprocessing unit (VVE) defective)

In a measuring system, depending on the maximum printing speed, between two and eight preprocessing units are used, which process the measured values zonally, ie the first one is responsible for zones 1 to 4 and the eighth for zones 29 to 32. Another column In the matrix with which measured values are checked according to this organization, conclusions can be drawn about defective preprocessing units.

Reference list

1

step

2nd

step

3rd

step

4th

step

5

step

6

step

7

step

8th

step

9

step

10th

step

11

step

14

step

15

step

16

step

17th

step

18th

step

19th

step

20th

step

21

step

22

step

23

step

24th

step

X, Y, Z

Color channels

Claims (5)

Method for image inspection and color guidance on print products of a printing press, wherein actual image data of the print images of the print products are preferably determined in online operation and compared with target image data for error detection,
characterized by
that if an error occurs before a change in the color control, it is checked whether a cause other than the color control is possible for the deviation due to the type of error. Method according to claim 1,
characterized by
that an error type analysis is carried out, in which different, predetermined criteria are successively used for error type detection, preferably systematically. Method according to one of the preceding claims,
characterized by
that an error correction measure is carried out in accordance with the determined error type. Method according to one of the preceding claims,
characterized by
that if the type of error is ambiguous, the type of error which has the highest probability is assumed to be present. Method according to one of the preceding claims,
characterized by
that in the case of several types of error in question for correcting a color change, in particular activation of the color control, is carried out last.

Images