EP1958774A2 - Method for controlling printing processes - Google Patents

Abstract

In a process to regulate a technical printing process e.g. crop mark, colour registration, colour density regulation point registration, or colour location regulation, a sensor measures a current value reproduced on the print medium. The value registered is compared with a pre-printed and measured target value. The difference between the two signals is used as a correction value signal for the print process. In the correction process, the actual value and the target value correction is first determined at reduced resolution and the correction then calculated by correlation. In the vicinity of the maximum, the correlated sub-pixel approximation is then effected by interpolation.

Description

The invention relates to a method for the regulation of printing processes according to the preamble of claim 1.

The regulation of printing processes, such. Example, a cut-off register control, a color register control or a color density control, usually using printed on the printed product pressure control marks, which are measured with the aid of sensors, in which case determined actual values of the print control marks with Druckvorstufebasierten target values for the print control marks are compared to on the basis of this comparison To determine actuating signals for the printing process to be controlled. Such print control marks, which are printed on the printed product outside of the actual print image or subject matter, are particularly disturbing in newspaper printing, since there is no trimming of the printed product in the newspaper print and, accordingly, the print control marks are visible in the end product.

The DE 42 37 004 C2 discloses a non-marking control method for controlling the color density of a printed product, wherein for this purpose the printed image of the printed product is measured directly by means of sensors. Actual values determined here are compared with reference values which are based on image data or prepress data.

The DE 101 49 158 A1 discloses a likewise without a brand-working method for cutting register control on a printing machine, wherein also in this method for the control pressure pre-stage-based setpoints are used.

In non-marking methods for controlling printing processes, the printing material is moved past a sensor fixed in the running direction or transport direction of the printing material to measure a printed image printed on the substrate, wherein the sensor detects a measuring location within the printed image by measuring to measure the printed image. Likewise, in the case of printing control processes based on print control marks, the printing material is moved past a fixed sensor, wherein the sensor detects a measuring location on the printing material for measuring the print control marks.

Either a one-dimensional sensor or a two-dimensional sensor can be used as a sensor for measuring a measuring location on the printed substrate, wherein a one-dimensional sensor provides a one-dimensional measuring vector as the measured value, whereas a two-dimensional sensor provides a two-dimensional measuring matrix as the measured value. When using a camera as a sensor, a measurement accuracy between 10 microns and 50 microns can be easily achieved, but such a high resolution on the one hand results in a relatively small site and on the other hand requires very short exposure times of less than 1 microseconds. If such short exposure times are to be realized, then robust LED flashes can not be used, but then relatively expensive lamp flashes must be used, which exhaust relatively quickly and therefore must be maintained very frequently. Furthermore, the measurement of the measuring location with such a high resolution requires large calculation times.

There is a need for a method for controlling printing processes, in which even when working with low resolution, a highly accurate control can be realized.

On this basis, the present invention has the object to provide a novel method for the regulation of printing processes.

This object is achieved by a method for controlling printing processes according to claim 1. According to the invention, the or each system deviation is determined in such a way that the actual value and the desired value are first offset at reduced resolution in a correlation method, and then an interpolation is performed in the region of the maximum of the correlation to the subpixel approximation.

For the purposes of the present invention, to determine the or each control deviation, the respective actual value is first calculated with reduced resolution in a correlation method, then an interpolation for sub-pixel approximation is performed in the region of the maximum of the correlation. With the invention, it is possible to calculate actual values and setpoints with relatively low resolution in a correlation method, whereby computing time can be saved.

However, since each value of a correlation vector resulting in the correlation method or of a resulting correlation matrix represents a very precise floating-point number, the position of the maximum of the correlation and thus the highly accurate control deviation can nevertheless be determined in a very high resolution by means of the downstream subpixel approximation. In this case, even over the prior art, the resolution can be exceeded.

Fig. 1:

ein Diagramm zur Verdeutlichung des erfindungsgemäßen Verfahrens.

Preferred embodiments of the invention will become apparent from the dependent claims and the description below. An embodiment of the invention will be described, without being limited thereto, with reference to the drawing. Showing:

Fig. 1:

a diagram to illustrate the method according to the invention.

The present invention relates to a method for controlling printing processes, in particular for cutting register control or color register control or for measuring location in the color density control or color location control or any other printing process in which a measuring location on a substrate is measured by means of a sensor for controlling. The measuring location on the printing material is measured by means of a stationary in the running direction or transport direction of the substrate pressure sensor, for which purpose the printing material is moved past the stationary sensor.

When measuring the defined measuring location at least one actual value is detected, wherein the or each measured actual value is compared with each one pre-pressure based setpoint. On the basis of at least one control deviation determined from the or each actual value and the respective nominal value, at least one actuating signal for the printing-technical process to be controlled is determined.

For the purposes of the present invention, to determine the or each control deviation, the respective actual value is first offset with the respective desired value with reduced resolution in a correlation method, wherein an interpolation is then carried out in the region of the maximum of the correlation to the subpixel approximation. The correlation is preferably performed as normalized cross-correlation.

wobei gilt: $$\mathrm{cov}\left(AB\right)=\frac{1}{n}{\displaystyle \sum _i}\left(A_i-{\mu }_A\right)(B_i-{\mu }_B);$$

$${\sigma }_A=\sqrt{\frac{1}{n}{\displaystyle \sum _i}(A_i-{\mu }_A{)}^2}\mathrm{und}{\sigma }_H=\sqrt{\frac{1}{n}{\displaystyle \sum _i}(B_i-{\mu }_B{)}^2};$$

und wobei weiterhin gilt: $${\mu }_A=\frac{1}{n}{\displaystyle \sum _i}{A}_i\mathrm{und}{\mu }_B=\frac{1}{n}{\displaystyle \sum _i}{B}_i\mathrm{.}$$

Normalized cross-correlation is performed using the following equations, where cor is the normalized cross-correlation, and where A is the actual value and B is the setpoint. $$\mathit{cor}=\frac{\mathrm{cov}\left(AB\right)}{{\sigma }_A{\sigma }_B};$$

where: $$\mathrm{cov}\left(AB\right)=\frac{1}{n}{\displaystyle \underset{i}{\Sigma }}\left(A_i-{\mu }_A\right)(B_i-{\mu }_B);$$

$${\sigma }_A=\sqrt{\frac{1}{n}{\displaystyle \underset{i}{\Sigma }}(A_i-{\mu }_A{)}^2}\mathrm{and}{\sigma }_H=\sqrt{\frac{1}{n}{\displaystyle \underset{i}{\Sigma }}(B_i-{\mu }_B{)}^2};$$

and where furthermore: $${\mu }_A=\frac{1}{n}{\displaystyle \underset{i}{\Sigma }}{A}_i\mathrm{and}{\mu }_B=\frac{1}{n}{\displaystyle \underset{i}{\Sigma }}{B}_i\mathrm{,}$$

The quantity n corresponds to the number of measuring points in the actual value as well as the setpoint points in the nominal value and thus the resolution thereof, the resolution of the actual value corresponding to the resolution of the nominal value.

Then, when a one-dimensional sensor is used to perform the method, the sizes A and B are vectors with n elements. Then, when a two-dimensional sensor, for. As a camera, is used, the sizes A and B are matrices with n elements, wherein the n elements of the matrices are structured in columns and rows. It is also possible to use a plurality of one-dimensional sensors in order to provide as actual value a two-dimensional measuring field or a measuring matrix which comprises a plurality of one-dimensional measuring tracks or measuring vectors.

As a result of the normalized cross correlation, depending on whether a one-dimensional sensor or a two-dimensional sensor is used, there is a one-dimensional correlation vector or a two-dimensional correlation matrix.

As already stated, the correlation between the actual value and the desired value takes place at a reduced or reduced resolution, preferably with a reduction by a factor of 2 ^nm .

For a two-dimensional sensor, this leads to a reduction of the data amount to 1/64 with a reduction of the resolution to 1/8 at m = 3. Such a reduction is also referred to as a reduced resolution in the sense of a third level of the image pyramid.

When a correlation is performed with such a greatly reduced resolution or data quantity, the position of the maximum in the correlation vector or the correlation matrix is highly quantized, but each value of the correlation vector or of the correlation matrix is a very precise floating-point number, so that the position can be determined by interpolation of the maximum in a very high resolution or accuracy can be determined.

Fig. 1 illustrates the inventive method for the case in which a one-dimensional sensor is used, wherein in Fig. 1 10 are plotted on the horizontal axis 10 values 11, 12 and 13 of a correlation vector resulting from the correlation performed at the reduced resolution. In the region around the maximum of the correlation vector, namely in the region of the value 12, there is a subpixel approximation by interpolation, wherein the curve 14 was obtained by interpolation. The maximum of the correlation determined in this way by interpolation is in Fig. 1 denoted by the reference numeral 15. In Fig. 1 is the interpolation realized via a parabolic fit.

With the method according to the invention, it is possible to use an LED flash with a relatively long exposure time of 30 μsec for detecting the or each actual value in conjunction with a sensor designed as a camera. This allows a relatively large site with an area of a few hundred square millimeters, z. B. a measuring location with the dimension 32 mm x 24 mm, to be measured. In this respect, relatively robust and inexpensive hardware can be used to carry out the method according to the invention.

The method according to the invention can be used for cut-off register control, for color register control as well as for measuring location determination in color density control or chromaticity control. The method according to the invention is suitable both for non-marking control methods and for control methods using pressure control marks.

Thus, the inventive method z. Example, in a markless color register control using a separate location for the ink black and a common location for the inks cyan, magenta and yellow are used. As a sensor, a two-dimensional color camera is then preferably used.

It is also possible to use the method according to the invention for a markless color register control using a multicolored measuring location with a distinctive structure in a printing ink.

LIST OF REFERENCE NUMBERS

10

axis

11

correlation value

12

correlation value

13

correlation value

14

interpolation

15

interpolated correlation maximum

Claims (12)

Method for controlling printing processes, in particular for Schnittregister- or color register control or for measuring location in the color density control or color control on a printing machine, wherein for determining at least one actual value, a measuring location on a substrate by means of at least one sensor is measured, the or each measured actual value is compared with in each case a pressure-bias-based setpoint value, and wherein at least one control signal for the pressure-technical process to be controlled is determined based on at least one of the or each actual value and the respective setpoint, characterized in that the or each control deviation is determined such that First, the actual value and the setpoint are calculated at a reduced resolution in a correlation method, and then an interpolation is performed in the region of the maximum of the correlation to the subpixel approximation. A method according to claim 1, characterized in that at least one one-dimensional sensor is used as sensor, which provides a one-dimensional measuring track or a measuring vector as an actual value, wherein the actual value is calculated with a setpoint formed as a one-dimensional setpoint track or desired value. A method according to claim 2, characterized in that a plurality of one-dimensional sensors are used to provide as actual value a two-dimensional measuring field or a measuring matrix comprising a plurality of one-dimensional measuring tracks or measuring vectors. A method according to claim 1, characterized in that a two-dimensional sensor is used as sensor, which provides a two-dimensional measuring field or a measuring matrix as an actual value, wherein the actual value is calculated with a setpoint formed as a two-dimensional setpoint or setpoint. Method according to one or more of claims 1 to 4, characterized in that the actual value and the setpoint are calculated using a normalized cross-correlation. Method according to one or more of claims 1 to 5, characterized in that the offsetting of the actual value and setpoint value at a reduced resolution in the sense of an m-th plane of the image pyramid is carried out with a resolution reduced by 2 ^rn . A method according to claim 6, characterized in that the offsetting of actual value and setpoint at a reduced resolution in the sense of a third level of the image pyramid is performed with a resolution reduced by a factor of 8. Method according to one or more of claims 1 to 7, characterized in that when detecting the or each actual value together with a sensor designed as a camera, a flash is used for exposure of the printing material, wherein the camera has a relatively large measuring location with an area of some One hundred square millimeters detected, and wherein the flash has a relatively long exposure time of more than 10 microseconds. A method according to claim 8, characterized in that an LED flash is used as the flash. Use of the method according to one or more of claims 1 to 9, for cutting register control on a printing machine. Use of the method according to one or more of claims 1 to 9, for color register control on a printing machine. Use of the method according to one or more of claims 1 to 9, for measuring location determination in the color density control or color location control.

Images