EP1291179B1 - Test means and method for controlling offset and digital printing - Google Patents

Abstract

Test target for determination of technical print parameters comprises a number of point clusters, that are formed from a number of adjacent image points and are printed as an identifiable print pattern on a print specimen. An Independent claim is made for a method for determining technical print parameters in which the test target is automatically recognized, measured and the required technical print parameters are calculated.

Description

Conventional print control strips are usually designed for densitometric or visual inspection. Print control strips are subdivided into test elements that allow to check the individual functions of the printing process. This results in control strips of considerable size, typically 12mm x 150mm. This type of control strip is not generally applicable to process controls, especially when it lacks positioning space. The use of conventional control agents is equally limited if the cutting away after printing is not possible and thus the control agent interferes with the viewer.

• Messungen im Bild
• Miniaturisierung des Kontrollstreifens

In principle, there are two promising methods to deal with this problem:

• Measurements in the picture
• Miniaturization of the control strip

Measurements in the image relating to the original image data are preferably used for quality evaluation of the printing result. In contrast, controls based on test elements provide information about the printing process. An overview of the current state of the art is given in the IFRA report (Ifra, 2001).

The MiniTarget measuring system described by Künzli et al pursues the second method (Künzli, 1998). It represents a big step towards miniaturization of quality control devices. With its dimensions of 7mm x 10mm, it is possible to capture the entire image information of the MiniTarget by single measurement with a CCD camera and to derive therefrom the printing parameters mathematically. In the essential characteristics, the MiniTarget concept is understood as a continuation of an idea that is already used in conventional print control strips. In particular, it is analogous to how print control strips are constructed on test elements which are arranged on a small area. With the existing dimensions, the MiniTarget still appears disturbing to the reader. In order to be invisible, or more appropriately "imperceptible", a size is required which is significantly less than 1mm ² .

DE 197 38 992 A1 describes a pad block for acquiring quality data in multicolor pad printing according to the preamble of claim 1.

EP-A 0 847 859 describes a control strip for media that are accessible to image capture.

Object of the present invention is to further miniaturization of the test target to an extent that this is perceived at least not obvious. This requires a new methodological approach.

This object is achieved by the test target according to claim 1 and the method according to claim 10. Preferred embodiments of the present invention are defined in the dependent claims.

• • Mustererkennung: Schafft die Voraussetzungen für die Ermittlung von Punktcluster Positionen und für die Adaption der Messblende an das Punktmuster der Messprobe
• • Charakterisierung von Punktclustern: Damit ist es möglich, die drucktechnischen Parameter aus Punktclustern anstatt aus Rastertonflächen abzuleiten

The image analysis has been used in particular for integral area measurements of grids. However, this application does not exhaust the potential of image analysis for quality and process control. Spatial image analysis combined with sophisticated data analysis are the key to two processes that are fundamentally different from integral measurement techniques such as densitometry and spectrophotometry:

• • Pattern recognition: Creates the prerequisites for the determination of point cluster positions and for the adaptation of the measuring orifice to the dot pattern of the test sample
• • Characterization of point clusters: This makes it possible to derive the printing parameters from point clusters instead of raster tone areas

A new test equipment and methods for the determination of printing parameters have been developed. The invention relates to a pattern of recording dots, such as lines and areas with polygonal or circular boundaries, which is reduced in size so that it is not or almost no longer visible as part of the printed image for the observer.

This test target is magnified microscopically and preferably recorded via CCD camera. The signal to noise ratio of the measurement is advantageously increased, which is averaged over several successive print patterns. Using sophisticated image analysis, it is possible to determine the printing parameters directly from the miniaturized patterns. The measures to reduce the size of the test target are described. The application to the type of process control is then outlined by way of example, which is essentially based, for example, on the geometric features of a specific dot pattern.

The dot pattern or test target is so small that it is virtually imperceptible to the naked eye by the observer. The dot pattern is formed by a plurality of dot clusters in a typical arrangement, preferably without point closure of the dot clusters with each other. The arrangement of the point clusters is typical in the sense that the dot pattern is identifiable as such in the print copy. The arrangement of the point clusters in the test target, ie the dot pattern, is preferably periodic. The test target has a total area, ie area of the point clusters + area of the spaces, of at most one square millimeter and is preferably square. Preferably, its area is less than 0.5 mm ² . The size specifications with respect to the area advantageously also apply if a virtual frame is drawn around the entirety of the point clusters of the test target, ie the size specification applies to the area within the frame. Each of the dot clusters is preferably printed in exactly one printing ink and is preferably formed from at least two recording dots in the X direction and two recording dots in the Y direction, ie, preferably at least four recording dots immediately adjacent in the X and Y directions. In rotary printing, the X and Y directions correspond to the circumferential and lateral directions. In terms of the greatest possible miniaturization, the printed pixel can advantageously be understood as a recording point.

By means of a sophisticated image analysis, the test target can advantageously contain both the density, the color locus and the area coverage of each of the individual colors Point clusters, as well as being determined in multicolor printing of the passer. In addition, a reliable diagnosis in terms of pushing and doubling is also possible. In multicolor printing, the dot pattern can also be used to control the gray balance.

The point clusters are at least so large that they can be analyzed planimetrically. The evaluation is advantageously carried out in the printing press during printing.

To determine color densities and color locations in the overprinting of several colors, the TestTarget (Dot Pattern) according to the invention can be extended with a supplementary TestTarget (Trapping Pattern). The supplemental target is preferably the same size and shape as the dot-pattern based test target. For example, in full-color printing, the supplemental TestTarget may contain the following four color patches, depending on the color order: C / M, C / Y, M / Y, and C / M / Y. The Trapping Pattern can be placed either side by side to the Dot Pattern or independently on the page of the printed copy.

The invention can be used advantageously in offset printing, in particular in wet offset printing. A particularly preferred field of application is newspaper printing on web-fed rotary printing presses. A test target according to the invention can advantageously also be imprinted in the image or on image-free areas, in particular on a side edge, of a printed copy of the newspaper edition. The greatest advantages of the invention in multi-color printing, although it can also be used with advantage in monochrome printing.

Design of dot patterns

Hereinafter, a preferred design of a dot pattern using the example of four-color printing is explained and explained with reference to the above-mentioned aspects.

Spezifikationen des Punktmusters im "invisible TestTarget"

1. a) Anordnung der Punktcluster C, M, Y und K
2. b) Darstellung von a) für p=q=2 im quadratischen Raster
3. c) Druckbeispiel des Musters gemäß b)

The test target is based on a specific dot pattern as shown in Figure 1.

Specifications of the dot pattern in the "invisible TestTarget"

1. a) Arrangement of the point clusters C, M, Y and K.
2. b) Representation of a) for p = q = 2 in the square grid
3. c) Printing example of the sample according to b)

C, M, Y, and K in Figure 1a symbolize clusters of recording dots in the corresponding colors, n is the total number of dot clusters in the horizontal and vertical directions, respectively. The clusters are preferably equal in number to the x and y directions, i. 1x1, 2x2, 3x3, ..., p x p adjacent recording points. The dot size corresponds to the addressability of the output device, which is indicated in dpi (dots per inch). The unprinted space in the TestTarget is q pixels in both the x and y directions.

Figure 1b shows an example of a realization of 1a for the parameters n = 4, p = 2 and q = 2. The point clusters are labeled in matrix notation by elements with the indices (i, j).

The dot pattern shown in the case of four-color printing is transferable to multi-color printing. The arrangement of the colors is preferably such that the point clusters running diagonally from top left to bottom right have the same color and all the primary colors are arranged in the top line in succession.

For dot-pattern based test targets, a fictitious frame is preferably defined in such a way that a cell results in a periodic regularity. This measure is taken with regard to the evaluation, which should not depend on the positioning of the metering orifice. This topic will be discussed below.

The control method is characterized by the fact that from the test patterns described in an example in Figure 1 both the register deviations and the printing parameters can be taken. It is specifically mentioned that not only sliding, doubling, area coverage can be determined by means of spatially resolving image analysis evaluation, but point cluster-based parameters can be determined, which are not accessible in conventional densitometry or spectrophotometry.

The presented type of point clusters is characterized by another notable feature. The dot patterns correspond to small area halftone dots containing characteristic information of the printing process. Small-scale point clusters, as shown in Figure 1, respond to process-related variations in pressure significantly more sensitively than halftone dots in conventional control strips or in the MiniTarget. The most important influencing factor is the tone value increase, which manifests itself in the fact that dots are usually rendered too large in comparison to the theoretical area occupancy. The effect is visualized in Figure 1b / c. The dot gain increases the dot by a fixed amount, which does not depend on the dot diameter. As a result, the effect increases in inverse proportion to the dot radius. As a result, the type of dot patterns shown in Figure 1 are particularly sensitive to process variations.

The page length of the test target in Figure 1b is calculated as follows: $$\mathrm{page\; length}\left(\mathit{in\; \mu m}\right)=\frac{25400\mathit{microns}}{\mathit{dpi}}\left(n*p+\left(n-1\right)q\right)$$

The percentage dot area A is $$\mathit{A}\left(\mathit{in}\%\right)=\frac{p^2}{{\left(p+q\right)}^2}*100\%$$

Typical values for A are 25% with p = q, 16% with p = 2 and q = 3, 11.1% with p = 2 and q = 4.

For the dot pattern in Figure 1b, the page length at 635 dpi output is 0.56 mm. The area is 0.31 mm2. This is more than a hundred times smaller than the MiniTarget after Künzli, 1998.

A supplementary test target was developed especially for the density and color measurements. This test target is preferably the same size and shape as the dot-pattern based target. In the case of four-color printing, it is subdivided into four quadrants containing printed color patches. These are in the case of four-color printing, depending on the color order, for example, C / M, C / Y, M / Y, and C / M / Y. The target may be placed side by side with the dot pattern based target, or independently. It is specially used to determine the color densities in the overprint and the ink acceptance. For this purpose, a method is used, which is presented in (Künzli, 2000).

When designing the "Invisible TestTarget", attention must be paid to the choice of the parameters in equation (1a / b), n should be chosen small in view of the requirement for a small-area test target. On the other hand, ensure that the printed dot clusters are representative of the printing process. In determining p and q is to pay particular attention to the Tonwertzunahme. It should be avoided that point clusters of the same color on the sample touch, since then the image analysis assessment can no longer be made point cluster specific or at least considerably more difficult. For this reason, a test pattern with 16% or 11.1% area coverage instead of 25% area coverage may be beneficial.

Production of the test samples

Four samples were prepared for the investigations. The dot patterns were programmed in the programming language PostScript. The specifications are listed in Table 1. Table 1: Specifications of the samples at a glance Sample No .: Production substratum Dot pattern # 1: electrophotographic printer Coated paper p = 2, q = 2 (CMYK, 400 dpi) No. 2: electrophotographic recycled paper p = 1, q = 2 (A = 11.1%) Printer (monochrome, 600 dpi) (see Figure 3) No. 3: Newspaper printing newsprint p = 4, q = 8 (A = 11.1%) (monochrome, 1200 dpi) (see Figure 3) # 4: Computer simulation Dot pattern for evaluation of push / duplicate (see Figure 4)

measuring device

The components of the image analysis system preferably include a 3-CCD camera, a microscope, a frame grabber card, and image analysis software. In particular, correct adjustment of the microscope, the illumination and the camera must ensure that a stable and high-contrast image is obtained during the measurement. Too high amplification of the measurement signal results in a strong noise of the image.

To determine color values, the image analysis system can be equipped with a spectrophotometer operating according to the spectral method instead of a camera operating according to the tristimulus method.

In addition, the size of the measuring aperture as well as its positioning on the sample to be measured is critical. If the orifice plate is set small, only a few point clusters are detected. It must therefore be risked that the sample is not representative, ie that the different-color point clusters are not taken into account according to their proportionate occurrence in the test sample. This results in a systematic measurement error (Romano, 1999). This is particularly large if the ratio of the orifice size and the dot cluster period has a low value and is between two consecutive integers, eg 1.5. For this reason, the acquired image section is adapted by software to the fictitious image frame, which represents an integer multiple of a period. As a result, parameters are obtained which are independent of the Meßblendeneinstellung. Basically, a single point cluster per ink is sufficient to characterize the printing process. However, for reasons of sample preparation and metrology, it is recommended that at least four point clusters be provided per ink for the measurement.

Despite the extremely small dimensions, the two test fields presented are suitable for carrying out color and density measurements. The measurement of the solid density and the color values of the individual colors is carried out on point clusters. For measurements of overprinted colors, the color field based supplementary test field is used, which is described earlier. For this it is necessary that the measuring ranges are calculated by software. For the measurements, a calibration of the 3-chip CCD camera is required (Künzli, 2000).

methods

The assessment of the digitized images takes place by means of image analysis (Demant Ch., 1998, Jähne B., 1997). The image analysis is based on the RGB image of the test pattern, which is displayed in sufficiently high resolution. The diameter of a point cluster should have been taken by at least 30 pixels. Most measurements are made separately for each ink. In this case, the channel is evaluated with the color which is complementary to the observed ink. Setting the threshold is a basic image analysis process to distinguish the point clusters from the paper white (Barratte Ch., 1995). In the greyscale histogram of the gray value image, the signals of paper white and of the printing ink are determined. The threshold value is the arithmetic mean of the modal values of paper white and the ink. Artifacts are eliminated in the measurement by image analysis removed small-scale structures outside the point cluster positions.

• Berechnung der Mittelpunkte für alle Punktcluster als arithmetisches Mittel aus den x-und y-Koordinaten der zu den Punktclustern gehörenden Pixel
• Unterteilung des Punktmusters in Punktcluster mit periodischer Regelmässigkeit wie z.B. in Abbildung 1 und anschliessende Berrechnung der Passerabweichungen
• Ermittlung der Parameter der Punktcluster, insbesondere Fläche, Gestalt (elliptisch oder rund), die Gleichmässigkeit des Randes
• Berechnung des Schiebens und Dublierens

A preferred evaluation is as follows:

• Calculation of the centers for all point clusters as an arithmetic mean of the x and y coordinates of the pixels belonging to the point clusters
• Subdivision of the dot pattern into point clusters with periodic regularity such as in Figure 1 and subsequent calculation of the register deviations
• Determination of the parameters of the point clusters, in particular surface, shape (elliptical or round), the uniformity of the edge
• Calculation of pushing and doubling

Averaging the signals

Conventional print control strips are characterized by the fact that the printing technology parameters can be taken from the test elements with sufficient accuracy. In contrast, miniaturized control agents are limited in this respect since the sample surface contains only a small number of point clusters. These point clusters are subject to random fluctuations, which are due to pressure and material technology. To a lesser extent, random fluctuations result from the optical measuring process. By means of signal averaging these difficulties can be largely eliminated.

(Bovik, 2000). Im Zusammenhang mit der vorliegenden Untersuchung werden zwei unterschiedliche Arten betrachtet, welche verschiedene Zwecke verfolgen:

• Bestimmung der Eigenschaften eines individuellen Punktclusters in hoher Genauigkeit
• Bestimmung der mittleren Geometrie von Punktclustern

Averaging over N signals increases the signal to noise ratio by the factor $\sqrt{N}$

(Bovik, 2000). In the context of the present study two different species are considered, which have different purposes:

• Determination of the properties of an individual point cluster with high accuracy
• Determination of the average geometry of point clusters

In order to determine the properties of an individual point cluster with high accuracy, the measurement process is repeated with the settings kept constant. It is assumed that the noise components of the signals are distributed in a Gaussian manner (Al Bovik, 2000). The averaging is done according to equation 2a. It turned out that this type of averaging is not relevant because the signal to noise ratio is sufficient. $$\begin{array}{cc}I\left(x,y\right)=\frac{1}{N}{\displaystyle \underset{i=1}{\overset{N}{\Sigma }}}{I}_i\left(x,y\right)& \mathrm{i}\mathrm{=}\mathrm{i}\mathrm{-}\mathrm{te\; measurement}\end{array}$$

In the second case, signals from different point clusters are used to calculate the mean geometry S (x, y) of a point cluster (Equation 2b). For this purpose, the light intensities I of N different point clusters are detected from test patterns of the same or successive prints. Then the signals are logarithmized. This takes into account that the optical density correlates with the ink layer thickness. Finally, the images of the N point clusters with a common center are superimposed and divided by N. This results in an image representing the average geometry of point clusters (Figure 2). $$\begin{array}{ll}S\left(x,y\right)\alpha \frac{1}{N}{\displaystyle \underset{i=1}{\overset{N}{\Sigma }}\log \left(I\left(x_i+x.y_i+y\right)\right)}& \mathrm{i}\mathrm{=}\mathrm{i}\mathrm{-}\mathrm{ter\; point\; cluster}\\ & \left(x_i,y_i\right):\mathrm{Center\; of\; the\; i}-\mathrm{ten\; point\; clusters}\\ & \left(x,y\right)\mathit{\epsilon \; Area\; of\; the\; point\; cluster}\end{array}$$

The physical meaning of the first averaging process is obvious, whereas the physical meaning of the second averaging process must be commented on. In the real world, a mean geometry of a point cluster calculated in this way does not exist. However, the characteristics of the average geometry contain printing process-specific features which are not apparent at the individual point cluster due to irregularities in the materials used and the reproduction. In an analogous way, process parameters can be obtained by statistical treatment of the numerical data of all N point clusters (Table 2). A correspondence between the parameters measured at the averaged point and the mean values of the parameters measured at the individual point clusters is to be expected in the case of area, shape, density and color values.

register deviations

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{Magenta}\mathrm{=}\mathrm{0.25}\left(\mathrm{X}\mathrm{-}\mathrm{Magenta}\left(\mathrm{1},2\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Magenta}\left(\mathrm{2},\mathrm{3}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Magenta}\left(\mathrm{3},4\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Magenta}\left(\mathrm{4},1\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{Magenta}\mathrm{=}\mathrm{0.25}\left(\mathrm{Y}\mathrm{-}\mathrm{Magenta}\left(\mathrm{1},\mathrm{2}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Magenta}\left(\mathrm{2},3\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Magenta}\left(\mathrm{3},4\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Magenta}\left(\mathrm{4},1\right)\right)\end{array}$$

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{Yellow}\mathrm{=}\mathrm{0.25}\left(\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{1},\mathrm{3}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{2},\mathrm{4}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{3},\mathrm{1}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{4},\mathrm{2}\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{Yellow}\mathrm{=}\mathrm{0.25}\left(\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{1},\mathrm{3}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{2},\mathrm{4}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{3},\mathrm{1}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{4},\mathrm{2}\right)\right)\end{array}$$

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{Black}\mathrm{=}\mathrm{0.25}\left(\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{1},\mathrm{4}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{2},\mathrm{1}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{3},\mathrm{2}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{4},\mathrm{3}\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{Black}\mathrm{=}\mathrm{0.25}\left(\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{1},\mathrm{4}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{2},\mathrm{1}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{3},\mathrm{2}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{4},\mathrm{3}\right)\right)\end{array}$$

For the test target in Figure 1b, the centers of each dot pattern for C, M, Y and K are calculated according to Equation (3a-d). (i, j) denote the center point of the point cluster in matrix notation with respect to Figure 1b. $$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{cyan}\mathrm{=}\mathrm{12:25}\left(\mathrm{X}\mathrm{-}\mathrm{cyan}\left(\mathrm{1},1\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{cyan}\left(\mathrm{2},2\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{cyan}\left(\mathrm{3},3\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{cyan}\left(\mathrm{4},4\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{cyan}\mathrm{=}\mathrm{12:25}\left(\mathrm{Y}\mathrm{-}\mathrm{cyan}\left(\mathrm{1},1\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{cyan}\left(\mathrm{2},2\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{cyan}\left(\mathrm{3},3\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{cyan}\left(\mathrm{4},4\right)\right)\end{array}$$

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{magenta}\mathrm{=}\mathrm{12:25}\left(\mathrm{X}\mathrm{-}\mathrm{magenta}\left(\mathrm{1},2\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{magenta}\left(\mathrm{2},\mathrm{3}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{magenta}\left(\mathrm{3},4\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{magenta}\left(\mathrm{4},1\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{magenta}\mathrm{=}\mathrm{12:25}\left(\mathrm{Y}\mathrm{-}\mathrm{magenta}\left(\mathrm{1},\mathrm{2}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{magenta}\left(\mathrm{2},3\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{magenta}\left(\mathrm{3},4\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{magenta}\left(\mathrm{4},1\right)\right)\end{array}$$

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{Yellow}\mathrm{=}\mathrm{12:25}\left(\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{1},\mathrm{3}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{2},\mathrm{4}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{3},\mathrm{1}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Yellow}\left(\mathrm{4},\mathrm{2}\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{Yellow}\mathrm{=}\mathrm{12:25}\left(\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{1},\mathrm{3}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{2},\mathrm{4}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{3},\mathrm{1}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\left(\mathrm{4},\mathrm{2}\right)\right)\end{array}$$

$$\begin{array}{r}\mathrm{X}\mathrm{-}\mathrm{Black}\mathrm{=}\mathrm{12:25}\left(\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{1},\mathrm{4}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{2},\mathrm{1}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{3},\mathrm{2}\right)\mathrm{+}\mathrm{X}\mathrm{-}\mathrm{Black}\left(\mathrm{4},\mathrm{3}\right)\right)\\ \mathrm{Y}\mathrm{-}\mathrm{Black}\mathrm{=}\mathrm{12:25}\left(\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{1},\mathrm{4}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{2},\mathrm{1}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{3},\mathrm{2}\right)\mathrm{+}\mathrm{Y}\mathrm{-}\mathrm{Black}\left(\mathrm{4},\mathrm{3}\right)\right)\end{array}$$

$$\begin{array}{c}\mathrm{DX}\left(\mathrm{Magenta}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{X}\mathrm{-}\mathrm{Magenta}\mathrm{-}\mathrm{X}\mathrm{-}\mathrm{Black}\\ \mathrm{DY}\left(\mathrm{Magenta}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{Y}\mathrm{-}\mathrm{Magenta}\mathrm{-}\mathrm{Y}\mathrm{-}\mathrm{Black}\end{array}$$

$$\begin{array}{c}\mathrm{DX}\left(\mathrm{Yellow}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{X}\mathrm{-}\mathrm{Yellow}\mathrm{-}\mathrm{X}\mathrm{-}\mathrm{Black}\\ \mathrm{DY}\left(\mathrm{Yellow}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\mathrm{-}\mathrm{Y}\mathrm{-}\mathrm{Black}\end{array}$$

Finally, the register deviations DX and DY are calculated according to equation (3e-g), using black as reference. $$\begin{array}{c}\mathrm{DX}\left(\mathrm{cyan}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{X}\mathrm{-}\mathrm{cyan}\mathrm{-}\mathrm{X}\mathrm{-}\mathrm{Black}\\ \mathrm{DY}\left(\mathrm{cyan}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{Y}\mathrm{-}\mathrm{cyan}\mathrm{-}\mathrm{Y}\mathrm{-}\mathrm{Black}\end{array}$$

$$\begin{array}{c}\mathrm{DX}\left(\mathrm{magenta}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{X}\mathrm{-}\mathrm{magenta}\mathrm{-}\mathrm{X}\mathrm{-}\mathrm{Black}\\ \mathrm{DY}\left(\mathrm{magenta}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{Y}\mathrm{-}\mathrm{magenta}\mathrm{-}\mathrm{Y}\mathrm{-}\mathrm{Black}\end{array}$$

$$\begin{array}{c}\mathrm{DX}\left(\mathrm{Yellow}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{X}\mathrm{-}\mathrm{Yellow}\mathrm{-}\mathrm{X}\mathrm{-}\mathrm{Black}\\ \mathrm{DY}\left(\mathrm{Yellow}\mathrm{/}\mathrm{Black}\right)\mathrm{=}\mathrm{Y}\mathrm{-}\mathrm{Yellow}\mathrm{-}\mathrm{Y}\mathrm{-}\mathrm{Black}\end{array}$$

The absolute values of the registration deviations are obtained from the spacings in the dot cluster patterns of the same colors, which are defined in dpi units of the output device.

The determination of characteristic quantities of the point clusters

Some parameters can be determined directly from the gray value images of the dot patterns in FIG. 3a. In this case, the threshold value method converts the gray value image into a binary representation in order to separate the point clusters from the paper white.

The percentage area coverage is a determining factor in the printing process. In contrast to conventional densitometry, image-analytical surface measurement is based on the principle of planimetry. First, the areas of the point clusters are determined individually. Then the dot cluster areas are summed up for all colors and divided by the area of the fictitious metering panel. The resulting value corresponds to the percentage area coverage.

• Umfang des Punktclusters
• grösster/kleinster Durchmesser des Punktclusters
• Lagewinkel (α)

The following parameters are preferably determined from the geometric areas of the point clusters:

• Scope of the point cluster
• largest / smallest diameter of the point cluster
• Position angle (α)

For the point clusters, further parameters are calculated from the above characteristic values.

The parameter E describes the geometry of the point cluster with respect to an ellipse shape. This parameter is used below to determine push and duplicate. $$\mathrm{e}\left(\mathrm{in}\%\right)=\left(\frac{\mathrm{smallest\; diameter}}{\mathrm{largest\; diameter}}\right)*100\%$$

Depending on the resulting value of E, the point cluster tends to have a circular shape (E = 1), an ellipse shape (E within 0% and 100%), or a straight line (E = 0%).

The so-called factor R is a measure of the smoothness of the edge course of point clusters (Haberäcker, 1995), which is a characteristic variable of the printing process. R is the ratio of the point cluster area to the square of the circumference (Figure 5): $$\mathrm{R}\mathrm{=}\frac{\mathrm{4}\mathrm{\pi }\mathrm{*}\mathrm{measured\; area}}{{\mathrm{Scope\; point\; clusters}}^{\mathrm{2}}}$$

Depending on the resulting value of R, the boundary is classified as smooth (R = 1), fringed (R within 0 and 1) or fractal (R = 0).

Process parameters with diagnostic function

Pushing and doubling are two typical effects in the printing process, which indicate a disturbed process control (Romano F., 1998). Pushing can be caused by different rotational speeds of the two cylinders and manifests itself in broadened lines transversely to the printing direction. The effect manifests itself visually in vertical lines, which are widened and therefore appear darker.

Duplication is caused by register problems between different printing units of multicolor printing machines and manifests itself in the fact that the point appears laterally offset and attenuated again. Visually, one recognizes the effect that line fields in one direction appear darker due to the broadening. Unlike pushing, duplication can occur in any orientation.

Both types of deviations are determined visually or metrologically by means of sliding or doubling fields. Figure 4 shows a computer simulation of the effect.

By means of image analysis methods, both effects can be derived from the dot pattern which has already been used for the measurement of the register and the color.

A presence of pushes is derived according to equation (4) from the factor E in conjunction with the positional angle of the point clusters. Statistical treatment will reveal trends in this regard. A preferred direction is proven when the standard deviation of the angle is sufficiently small.

• Transformation des originalen Grauwertbildes mittels Schwellenwertverfahren in ein binäres Bild. Der Schwellenwert wird in der Weise gesetzt, dass eine durch dublieren bedingte Punktverbreiterung eingeschlossen wird. Sodann werden die Koordinaten (x,y)-2 der Mittelpunkte der Punktcluster berechnet
• Transformation des originalen Grauwertbildes mittels Schwellenwertverfahren in ein binäres Bild. Der Schwellenwert wird in der Weise gesetzt, dass eine durch dublieren bedingte Punktverbreiterung nicht eingeschlossen ist. Sodann werden die Koordinaten (x,y)-1 der Mittelpunkte der Punktcluster berechnet

Duplication manifests itself in the histogram of the gray values. This essentially shows two signals originating from the paper white and the printed dot clusters. When doubling, the signal of the point clusters is slightly broadened or even shows a page maximum. The extent of doubling is determined according to the following steps:

• Transformation of the original gray value image into a binary image by means of thresholding. The threshold is set to include doubling point broadening. Then the coordinates (x, y) -2 of the centers of the point clusters are calculated
• Transformation of the original gray value image into a binary image by means of thresholding. The threshold is set so that doubling point broadening is not included. Then, the coordinates (x, y) -1 of the centers of the point clusters are calculated

The doubling effect D results from the difference between the two vectors according to equation 6: $$\mathrm{D}\mathrm{=}\left(\mathrm{x},\mathrm{y}\right)\mathrm{-}\mathrm{2}\mathrm{-}\left(\mathrm{x},\mathrm{y}\right)\mathrm{-}\mathrm{1}$$

Results and discussion

The methods described in the previous chapter have been experimentally verified. For this purpose, samples 1-4 in Table 1 were used. The test results of Samples 1 and 2 are summarized in Table 2. Table 2 Image analysis results compared to the specified values of a monochrome test pattern prepared on an electrophotographic printer and in newspaper printing. Notation in "(s) electrophotography image analysis specification Absolute area point clusters (in μ m ² ) 1416 (127) 1792 Area coverage point clusters in% 8.7 (0.8) 11.1 Scope point clusters (in μ m ) 159 (14) 169 Diameter of smallest / largest point clusters (in μ m ) 39/46 (3/2) 42.3 / 42.3 Angle in degrees 73 (59) Ellipsoid factor E in% 87 (4) Equation (4) Edge smoothness R in% 0.72 (0.04) Equation (5) Absolute area point clusters (in μ m ² ) 10769 (990) 7174 Area coverage point clusters in% 16.8 (1.5) 11.1 Scope point clusters (in μ m ) 458 (33) 339 Diameter of smallest / largest point clusters (in μ m ) 110/124 (6/7) 84.7 Angle in degrees 71 (50) Ellipsoid factor E in% 88 (2) Equation (4) Edge smoothness R in% 0.65 (0.06) Equation (5)

Representative image analysis results are shown in Figures 2 and 3. Figure 2 shows a three-dimensional plot of a point cluster before and after the averaging.

Relief representation of a point cluster: printed by electrophotography (left: 126μ m · 126μ m ) and in newspaper printing (right: 250μm · 250μm). The vertical axis represents gray values.

Relief of average dot clusters: electrophotographically printed (left: sample size 36, 126μ 126μ m · m) and newsprint (right: sample size 41, 250 .mu.m · 250 microns). The vertical axis represents gray values.

Figure 3 shows grayscale images of dot clusters before and after the threshold process.

Gray value images of four point clusters, printed by electrophotography. (left: 250 μ m × 250μ m) and in newspaper printing (right: 500μ m × 500μ m).

Binary images after thresholding the grayscale images in Figure 3a.

The area coverage of the test sample of sample 1 is 8.7%. The sample size is 36 point clusters. The standard deviation of 0.8% results from random process and material variations, which result in point clusters of different sizes. Orientative measurements show that densitometric area measurements generally give larger values. This is justified by the optical light capture, which is not taken into account in the image analysis. In addition, it has been found that the accuracy of the measurements depends significantly on the lighting as well as the choice of the threshold. It is therefore important to pay particular attention to reproducible measurement conditions. The following parameters refer to geometric point clusters and parameters obtained numerically from the surfaces of the point clusters. The dot circumference is 159 μ m , which is slightly smaller than the specified value of 169 μ m due to the smaller area coverage. The ellipsoid factor E is 87%, from which a roundish geometry is concluded. This finding is confirmed from the averaged structure in Figure 2. This also explains that the angle scatters strongly and thus can not recognize a preferred direction. The mean is 73 degrees and has a standard deviation of 59 degrees. The factor R of 0.72 indicates an uneven edge structure. This finding is visually consistent with the relief representation in Figure 2a and the gray scale images in Figure 3. The results of the newspaper print patterns show a similar behavior as Sample 1.

Density and color measurements can be performed equally on the test pattern in Figure 1. Previously, the image analysis system is calibrated by a method described in the literature (Künzli, 1998). It allows the conversion of RGB values into colorimetric and density values. Apertures are calculated for the measurements, which selectively capture the point clusters.

The averaging of signals was examined on the electrophotographic (Sample 2) and the newspaper print (Sample 3) printed samples. Figure 2 shows the point clusters before and after the averaging. It is obvious that the averaged Structures have smoother gradients compared to the structures of individual points. The averaged structures in Figure 2 indicate that the electrophotographically produced point clusters tend to have a square structure. In contrast, the dot clusters in newspaper printing have a rounded shape. The comparison of the results in Table 2 shows that the parameters of the point clusters calculated from the individual values agree well with the parameters which were determined from the averaged structures. This applies to the absolute areas, the percentage area coverage, the ellipsoid factor E and the largest and smallest diameters. In the case of the edge smoothness R, 0.92 in newspaper printing and 0.91 in electrophotography are obtained for the averaged structures. These values are, as expected, significantly higher in comparison to the corresponding averages of 0.65 and 0.72 in Table 2. This difference stems from the smoothing effect which results from the averaging.

The registration differences were determined according to equation (3a-g) from the test pattern of the color printer (sample 1). Deviations were obtained which are between 15 and 55 μm. This value is below the specified diameter of the dot cluster which amounts to 63 μ m. The values determined lie within the tolerances of offset printing, which are given in ISO 12647-2 as 83.3 μ m for 60 l / cm. Overall, the results show that the analysis of the registration deviations can be made on a test pattern which does not appear conspicuously.

Figure 4 shows the computer simulation of pushing and doubling.

1. a) Muster von runden Punktclustern
2. b) Schieben
3. c) Dublieren

Assessment of pushing and doubling (computer simulation)

1. a) Pattern of round point clusters
2. b) pushing
3. c) duplication

Slide analysis was performed on a sample prepared by computer simulation (see Figure 4b). The evaluation is based on the ellipsoid factor E in Equation 4. The smallest and largest diameters were found to be 1.72 and 1.77 relative units, and the angle to 172 degrees. These values are in good agreement with the input values of the computer simulation. This is a pushing proven quantitatively. The duplication analysis in Figure 4c is done according to Equation 6. The resulting vector difference is calculated as the distance and angle with respect to the center point coordinates of the point cluster without duplication. A distance of 0.18 relative units and an angle of 145 degrees are obtained. Slight variations in these values are attributed to digital noise introduced during modeling.

Conclusions and outlook

The present study shows that the newly developed test pattern or TestTarget in combination with the described evaluation methodology for the control of the Multi-color printing can be used. The process is especially popular where process control is required, but for reasons of space, no control strips can be used. With a test pattern area that is well below 1mm ² , an almost unlimited use is possible. In particular, a targeted positioning can now be made close to selected image locations. With the described method, the limits of what was possible were rigorously explored. They culminate in the statement that the parameters relevant to printing technology can ultimately be taken from a test pattern which is made up of one point cluster per printing ink. However, an averaging over a plurality of successively printed print patterns is advantageous. All of the described measures, which in their entirety allow such a miniaturization of the test target, were tested experimentally or by simulation.

With the application of prediction methods, as currently developed at EMPA, St. Gallen, Switzerland, extrapolation of the results over the entire range of the tone curve is promising (Mourad, 2001).

references

1. 1. Barratte Ch., Dalphond JE, Mangin PJ and Valade JL, (1995), An Automatic Determination of Threshold for the Image Analysis of Prints. Proceedings of the 23rd Research Conference of IARIGAI, Paris, 451-469
2. Second Bovik Alan C., (2000), Handbook of Image and Video Processing, Academic Press, New York
3. Third Demant Ch., Streicher-Abel B. and Waszkewitz P., (1998), Industrial Image Processing, Springer Verlag, Heidelberg
4. 4th Haberäcker P., (1995), Praxis der digital Bildverabeitung. 4th edition, Hanser, Münch s
5. 5th Ifra, (2001), Closed loop printing quality control for coldset web offset (to be published), Darmstadt
6. 6th Jähne B., (1997), Digital Image Processing, Springer Verlag, Heidelberg
7. 7th Künzli H., et al, (1998), Mini Targets: A new dimension in print quality control, SPIE Conf. Proc., Vol 3300, pp 286-91
8. 8th. Künzli H., (2000), How to convert RGB signals to colorimetric and densitometric values, Proceedings of SPIE, Vol. 3963, 70-76
9. 9th Künzli H., (2000), MiniTarget for Quality Control in Newspaper Printing, TAGA Proceedings, 509-520
10. 10th Mourad S., et al. (2001), Predicting Transmittance Spectra of Electrophotographic Color Prints, Proceeding of SPIE, Vol. 4300, 50-57
11. 11th Romano D., (1999), Recorder spot size and its effect on image quality and halftone reproduction, 280-293
12. 12th Romano FJ, Romano RM, (1998), Encyclopaedia of graphic communications, GATF, Prentice Hall, London on

Claims (14)

1. A test target for determining typographic parameters, which comprises a plurality of dot clusters which are each formed by a plurality of adjacent recorded dots and printed onto a printed copy in the form of such an identifiable dot pattern,
   characterised in that
   the test target also comprises unprinted intermediate spaces between adjacent dot clusters, wherein the extent of the unprinted intermediate spaces corresponds to an equal first number of recorded dots in both the horizontal and vertical direction.
2. The test target according to claim 1, characterised in that it contains a plurality of dot clusters of one printing colour.
3. The test target according to any one of the preceding claims, characterised in that it contains at least one individual colour dot cluster for each of a plurality of individual colours, preferably all the individual colours, of the print.
4. The test target for determining typographic parameters in accordance with claim 1, wherein each of the plurality of dot clusters is formed by an equal second number of adjacent recorded dots in both the horizontal and vertical direction.
5. The test target according to any one of the preceding claims, characterised in that none of the dot clusters is in dot closure with an adjacent dot cluster of the test target.
6. The test target according to any one of the preceding claims, characterised in that the dot clusters form a matrix as the dot pattern, the rows and columns of which are fully occupied by the dot clusters as matrix elements.
7. The test target according to any one of the preceding claims, characterised in that the dot clusters form a matrix, in the rows of which the individual colours of the print are arranged in order.
8. The test target according to any one of the preceding claims, characterised in that the dot clusters form a matrix, in the columns of which the individual colours of the print are arranged in order.
9. The test target according to any one of the preceding claims, characterised in that the dot clusters are arranged in the form of a matrix in which the diagonals are each formed by dot clusters of only one individual colour of the print, wherein each individual colour of the print forms at least one diagonal with preferably at least two dot clusters.
10. Use of the test target according to any one of the preceding claims to determine typographic parameters, wherein

    a) the test target is automatically recognised,

    b) measured,

    c) and evaluated in order to determine typographic parameters such as for example area coverage, full-tone density and/or colour value and/or in order to diagnose shifting and/or doubling.
11. The use of the test target according to the preceding claim, characterised in that the test target is recognised as a test target by image analysis.
12. The use of the test target according to any one of the preceding claims, characterised in that the dot clusters of the test target are measured planimetrically.
13. The use of the test target according to any one of the preceding claims, characterised in that the diameter, size and shape (the position angle α, elliptical parameter E and factor R) of the dot clusters are determined.
14. The use of the test target according to any one of the preceding claims, characterised in that the typographic parameters determined are used to test and preferably control and/or regulate the printing process in a control and/or regulating unit.