METHOD FOR INLINE MONITORING THE FUNCTIONALITY OF AT LEAST ONE PRINTING HEAD

Description

[0001] The invention relates to a method for inline monitoring of the functionality of at least one printhead.

[0002] In the printing industry, one of the key objectives is to produce the most consistent possible image and color impressions on the materials being printed. This so-called color consistency, even when printing different batches on different substrates, should always appear the same. This is particularly important when corporate identity is desired. The color(s) chosen by a company should always look the same, regardless of whether the printing is on glass, paper, plastic, or another substrate. The most identical appearance possible of the printed products is desirable and therefore a goal of the printing industry.

[0003] Printing methods, printers, printing inks, and especially the substrate have a significant impact on the final appearance of a printed document created from a digital file. Printing ink refers to any printing medium, such as ink or toner, used to apply color pigments to the substrate.

[0004] During industrial printing, whether in single-pass operation with a stationary printhead and moving substrate or in multi-pass operation with a moving printhead and stationary substrate, it is necessary to check the functionality of at least one printhead, and in particular all of the printheads used, in order to maintain the specified color consistency.

[0005] A local change in the printing system, where the printing behavior of an entire printhead or a section (segment) of a printhead changes, leads to a streaky print image. This streaking in the print image is also known as banding.

[0006] Changing the printer settings, and thus potentially correcting the print image, cannot be corrected by modifying the print file, whether automatically or manually. This is because these changes affect the entire image area globally, as either the output profile (ICC profile) or the input file for color management must be modified.

[0007] Therefore, it is advisable to check the printer at regular intervals to ensure that each color can be printed homogeneously across the entire print width and, if necessary, to perform a banding correction.

[0008] To correct banding, special test forms are printed and analyzed. These forms are designed so that each individual ink color has areas of varying intensity levels, extending homogeneously across the entire print width. This allows for quick and easy detection of banding, enabling appropriate countermeasures to be taken (cleaning, adjusting printhead or printhead segment tension). Once corrective measures have been implemented, the test form is printed again to verify their effectiveness. This process is repeated until the required quality is achieved.

[0009] Banding correction is therefore complex and interrupts the ongoing printing process. Consequently, the described banding correction is only performed at longer, sometimes excessively long, intervals.

[0010] DE 10 2018 201 785 B3 discloses a method for detecting and compensating defective print nozzles in an inkjet printing machine, using a test pattern in the form of a compressed print image.

[0011] US 2016 / 0031252 A1 discloses a method for monitoring a printing system in which a verification data set based on the printing data and having different spatial resolutions in a first and a second direction is used as a test pattern.

[0012] EP 3 578 939 A1 discloses a method for online quality control of decorative prints on substrates, in which a hyperspectrally measured reference image is first generated and stored as the target image in a digital format. During the ongoing printing process, digital actual images of the printed image are then generated using a conventional imaging technique and compared with the target image.

[0013] Therefore, the present invention is based on the technical problem of providing a method for inline monitoring of the functionality of at least one printhead, with which the aforementioned problems are at least partially solved.

[0014] The aforementioned technical problem is solved according to the invention by a method for inline monitoring of the functionality of at least one printhead having the features of claim 1, comprising a series of method steps which are described below including preferred embodiments.

[0015] First, a design to be printed is provided as a source image file in a source color space. This is done either by creating the source image file itself or by transferring an existing one. The source image file is typically displayed in an RGB color space (R - red, G - green, B - blue), which is used in most image processing programs. Various versions of the RGB color space exist, such as sRGB, Adobe RGB, and Adobe Wide Gamut RGB. Other color space representations are also possible, such as the YUV color model with specified luminance and chrominance, CMYK (cyan, magenta, yellow, key (black)), or color separation as used in analog printing.

[0016] The source image file is then transferred by a color management system with an input profile into an L*a*b* color space, and an L*a*b* image file is created.

[0017] Color management ensures the most accurate possible reproducibility, so that a print template can be faithfully reproduced through the printing process using any output device. The correct conversion of the data to the print-ready image file is crucial for this.

[0018] For this purpose, a color management system is generated, within which device profiles—that is, tables containing the device's color characteristics—can convert the respective device's own color space into an independent exchange color space, usually an L*a*b* color space such as CIELab. Using so-called ICC profiles (ICC - International Color Consortium), the data can be exchanged between different devices in such a way that, depending on the physical limitations of the input and output devices, a printout can be produced that is as color-matched as possible to a monitor display.

[0019] In particular, the color management system can use the PCS-CIELab exchange color space, a special implementation of the CIELab color space known as Profile Connection Space (PCS). Therefore, an ICC profile is preferably used as the input profile to transfer the RGB image file into the L*a*b* color space.

[0020] The color management system (CMS) ensures communication between the elements of the graphic design process chain by using a common language. The standard is the aforementioned L*a*b* color space, which is integrated into every ICC profile. L*a*b* is the color space that most closely approximates human vision and defines colors in absolute values ​​within a coordinate system. Therefore, at every stage of processing, the color management system transforms the L*a*b* values ​​of all peripheral devices in the chain, as integrated into the respective ICC profile, until the final print. The L*a*b* color mode is the cornerstone of the ICC architecture.

[0021] The L*a*b* color space is a color space that covers the range of perceptible colors. The L*a*b* color space is described by a three-dimensional coordinate system. The L* axis describes the brightness (luminance) of the color, with values ​​from 0 (black) to 100 (white). The a* axis describes the green or red component of a color, with negative values ​​representing green and positive values ​​representing red. The b* axis describes the blue or yellow component of a color, with negative values ​​representing blue and positive values ​​representing yellow. The scales of the a* and b* axes cover a numerical range from -128 to +127. However, other scales are also known.

[0022] When color representation is performed in the L*a*b* color space, the difference between two colors is determined by the value of ΔE00 (CIEDE2000) during difference analysis. ΔE00 is a measure of color difference, where "Δ" denotes the difference. This allows for the comparison of color print results and the quantification of color distances. To determine the distance between two color points in the L*a*b* color space, the ΔE00 (CIEDE2000) of the two color points is calculated according to the formula in ISO 13655:2008 Annex B.

[0023] In the next step, the L*a*b* image file is transferred from the L*a*b* color space to a printable color space suitable for the printer using an output profile, and a print image file is generated. Preferably, an ICC profile is again used as the ICC output profile.

[0024] The printing color space takes into account the printing inks actually used for printing, with the CMYK color space (cyan, magenta, yellow, key, or black) being the most common. However, the color space can also contain other printing inks or even more inks. If the printing color space is limited and not all color shades are included, the color management system ensures the most accurate possible reproduction of the design within the available color space.

[0025] The color management system then preferably uses a raster image processing (RIP) method, in which a pixel area of ​​the print file is created by a multitude of droplets of ink or toner dots evenly distributed across the pixel area. The RIP process can be carried out, for example, with software from Colorgate, Ergosoft, or Caldera.

[0026] The design is then printed by a printer with at least one printhead onto a substrate in a raster of pixels using at least one ink color, based on the print image file. The raster consists of rows running perpendicular to the printing direction and columns running in the printing direction. Inkjet printing is frequently used, but other printing techniques such as laser printing can also be employed.

[0027] Thus, for example, the assignment between a column of the printed image and the coordinate of the print nozzle within the corresponding printhead in the respective row can be known.

[0028] Suitable substrates include paper, foil, wood, plastic, ceramic, or minerals. Inks containing pigments are typically used as printing inks, with the use of CMYK colors being widespread.

[0029] In principle, the described procedure can also be used for printers with only one printhead with a print width of 3 cm.

[0030] The preferred application of the described method lies in the industrial sector, where relatively wide substrates are printed. For example, the method is used for printing decorative papers that are used for coating wood-based panels. Widths of up to 1 or 2.3 meters are common. For these widths of substrates to be printed, the printer has multiple printheads that are synchronously but individually controlled. In addition, the individual printheads can have individually controllable segments.

[0031] In a common application of this method, the printer or printheads are stationary, and the substrate is transported beneath the printheads, so that the printing direction corresponds to the transport direction. This printing method is called single-pass printing. Thus, the printed image or design is printed in rows perpendicular to the printing direction and in columns parallel to the printing direction. Therefore, if a malfunction occurs in one of the printheads or one of the segments of a printhead, deviations in the printed image will appear column by column.

[0032] In a single-pass printing press, the design is printed across the entire print width in one pass. For example, an image can be printed across the width using 6 printheads with 4 colors (CMYK), meaning a CMYK printer has 4 rows of 6 printheads each.

[0033] The printed image is then measured, at least in sections, but preferably across its entire width, using an optical measuring system with a spatial resolution (for example, measured in dots per inch, dpi) in the area of ​​a few pixels of the printed image, in particular with a spatial resolution of one pixel of the printed image, line by line in the printing direction, and a measurement image file consisting of rows and columns is generated. The measurement image file then contains multi-channel color information for each measured pixel.

[0034] The optical system is designed as a spectral measurement system. The spatial resolution of a spectral measurement system is typically in the range of 90 to 200 dpi, while color cameras have a higher resolution.

[0035] The preferred spectral measurement system is an inline color measurement system (ICMS - Inline Colour Measurement System) from ipac, known from the prior art, which is used for color assessment of multi-colored surfaces. The system features a spectrally arranged inline scanner that can be used for various substrate materials (paper, film, wood, plastic, ceramic, mineral) and various printing processes (gravure printing, digital printing, flexographic printing, offset printing, screen printing). The ICMS is a spatially resolved spectral scanning technology for measuring a printed image and is capable of mechanically reproducing the optical color impression perceived by a well-trained and healthy human eye and subjecting this color impression to a comprehensive, objective evaluation.

[0036] The spectral measurement system is, for example, a multispectral camera with 12 image channels per captured pixel, generating color information for each channel. Thus, a color spectrum of approximately 10 to 12 image channels is created for each captured pixel. A common sensor technology involves equipping individual pixels on a CMOS sensor with different color filters, allowing a single image capture to record multiple spectral information from a captured image area.

[0037] The spectral measurement system can also be a hyperspectral camera, in which the light is spectrally split at each pixel using an optical device, for example a prism, and individual spectral ranges are measured separately. This increases the spectral resolution compared to a multispectral camera to approximately 20 to 250 or more image channels.

[0038] In contrast to an RGB camera, this method captures not just one color per pixel, but a spectral distribution with significantly greater information depth. Standardized lighting conditions are maintained during the measurement; for example, the measurement geometry of the imaging system is 45°:0° (entry:emission).

[0039] The next step involves transferring the measurement image file into an L*a*b* measurement image file. The calculation of the L*a*b* measurement file from image information of a spectral measurement file is performed, for example, according to the CIE-15 standard, preferably using the CIE illuminant D50 and the 2° standard observer (CIE-15 ASTM 308E, ISO 13655 Annex I). The calculation of the L*a*b* measurement file from image information of an RGB measurement file is performed, for example, using a transformation specifically created for the respective measurement conditions, which is obtained through calibration using a spectrophotometer.

[0040] The L*a*b* measurement image file is then transferred to the printing color space using the output profile, and a comparison image file is generated. The same output profile is used as the one applied before printing the design when converting the L*a*b* image file from the L*a*b* color space to the printing color space. Again, the CMYK color space is the preferred printing color space.

[0041] Using the same ICC output profile ensures optimal comparability of the data to be compared subsequently.

[0042] Furthermore, the comparison image file and a predefined reference image file are registered against each other to perform a target-actual comparison. Before this comparison, a reference image file is therefore created that contains the target state of the print in at least one printing color, but preferably in all printing colors. The reference file can be generated in various ways.

[0043] Firstly, the reference image file can be generated from the print image file, especially for the CMYK color space, for at least one printing color, preferably for all printing colors. This is because the intensities of the individual printing colors contained in the print image file define the desired print result in digital form, and the printing color intensities can therefore be used as a reference image file for determining the target values.

[0044] Alternatively, the reference image file can be generated from a comparison image file for at least one printing ink, preferably for all printing inks, which is created from a previous measurement of a printed image or design and is explained above. In this case, the data of the reference image file is not generated theoretically from the original image file, but rather through a previously performed practical optical measurement.

[0045] Before the comparison between the reference image file and the comparison image file can be performed, the two image files are registered relative to each other, i.e., aligned by shifting, rotating, or distorting the pixel matrices. This achieves the most accurate possible match of the pixels or pixel groups of both image files.

[0046] The comparison step generates a difference image file from the comparison image file and the reference image file, and the difference values ​​in the difference image file are determined line by line. Theoretically, if the reference image file and the comparison image file are on the order of individual pixels or pixel groups, the difference values ​​are zero, but in practice they are noisy and fluctuate around zero.

[0047] The determined difference values ​​can be directly assigned to the corresponding printhead nozzles via their width. Since, preferably, both the print output file and the comparison image file are generated with the same output profile (ICC profile) to convert CIELAB color values ​​into the printer's color channels (CMYK), the cause of banding in the print image can be clearly attributed to a specific physical printhead.

[0048] The comparison can, in principle, be performed line by line; however, it is advantageous to average the difference values ​​column by column over a predetermined number of lines, preferably continuously during the printing process. Since the printing speed is generally relatively high and the expected fluctuations occur with a small time constant, such averaging can lead to better results.

[0049] If the values ​​in the difference image file for each of the printing inks fluctuate essentially around zero, this can serve as a measure that the print result has no or only a negligible deviation in that printing ink.

[0050] However, if the magnitude of the difference values ​​exceeds a predefined threshold, either above or below it, a malfunction of the printhead associated with the deviating pixels can be detected. This threshold can be set differently or the same for each ink color. For each section of lines in the difference image file where the threshold is exceeded, a malfunction of the associated printhead or a segment thereof can be identified. Because the optical measurement can be performed with pixel-level accuracy, even the malfunction of individual ink nozzles can be detected.

[0051] Using the method described above, color deviations of less than ΔE00= 1 in the L*a*b* color space can be measured.

[0052] When decorated surfaces are created using digital printing, optical color measurement technology can be used to check the color accuracy of the printed image. If color deviations occur during color printing to reproduce a design, these can be measured using the described method. Therefore, at least one of the printer's printheads can be adjusted by changing the control parameters within the printer control system to minimize the difference values ​​in the differential image file.

[0053] The printer can be corrected during continuous printing without interrupting the printing process, since the changes that occur and can be detected by the described method are so small that they are almost imperceptible to the human eye in the printed image.

[0054] Preferably, the printer's at least one printhead can be adjusted by changing the control parameters within the printer control system so that the difference values ​​in the differential image file are minimized. This allows the control system to correct errors generated by the at least one printhead without interrupting the printing process.

[0055] Thus, automatic correction can be achieved by combining the color management system with the printer control.

[0056] Further improved color correction can be achieved using the previously described method if the entire printing system is characterized, i.e., calibrated, in particular linearized and profiled, with the same optical color measurement system.

[0057] The invention will now be explained using exemplary embodiments with reference to the drawing. The drawing shows... Fig. 1 a first apparatus for carrying out a method according to the invention, Fig. 2 a second apparatus for carrying out a method according to the invention, Fig. 3 a third apparatus for carrying out a method according to the invention, Fig. 4 a flowchart describing a method according to the invention, Fig. 5 a printed design whose color components are analyzed, Fig. 6a and / or combined representations for one printing ink to illustrate the method according to the invention, Fig. 7a-d representation of the difference image file for four printing inks for a correctly printed design, and Fig. 8a-d representations of the difference image file for four printing inks for a design printed incorrectly in the magenta M color channel.

[0058] In the following description of the various embodiments according to the invention, components and elements with the same function and mode of operation are provided with the same reference numerals, even if the components and elements may differ in their dimensions or shape in the various embodiments.

[0059] Fig. 1 Figure 2 shows a first system for carrying out the procedure for the automated characterization of a continuous digital printing system. First, the carrier material 4 is unwound from a first roll 6, guided under the digital printer 8 and the optical measuring system 10, and then rewound onto a roll 12. In this respect, one can speak of an endless carrier material 4 with which the continuous printing and characterization are carried out. It goes without saying that the endless strip has a finite, but considerable, length.

[0060] The results of the optical measurements by the optical measuring system are transferred to a control and characterization unit 14, which in turn transfers not only the print files or other control files, but also the characterization files to the digital printer 8.

[0061] Fig. 2 A second system 2 is shown for carrying out the procedure for the automated characterization of a continuous digital printing system. In comparison to the system according to Fig. 2 The carrier material 4 is not unwound from a roll, but produced by an extrusion process. A schematic extrusion die 16 is shown, from which a strand is extruded to produce, for example, an edge banding material for use in furniture panels. For the sake of simplicity, the calenders and cooling stations that are generally necessary are not shown here. Instead of the extrusion die 16, a continuous casting device can also be used to produce a continuous strand of carrier material.

[0062] Fig. 3 A third system, Annex 2, demonstrates the implementation of the procedure for the automated characterization of a continuous digital printing system. In comparison to the systems described above, Fig. 2 and 3The substrate 4 is not designed as a continuous material, but consists of a multitude of abutting elements 18, for example, plates or sheets. The substrate 4 thus consists of individual elements 18 that are separate before and after printing. The continuous printing and measurement of color charts then takes place on the substrate 4 composed of individual elements 18.

[0063] Fig. 4 shows an embodiment of a method according to the invention for inline monitoring of the functionality of at least one printhead.

[0064] After starting in symbol 100, in step 102 a design 104 to be printed is provided as a source image file 106 in an RGB output color space. The file can be created manually or a previously created and saved source image file can be used.

[0065] The source file 106 is then further processed using a color management system 110. In step 112, the source image file 106 is linked to an ICC input profile 114 and in step 116 transferred into an L*a*b* color space in the form of a PCS-CIELAB color space. This results in an L*a*b* image file 118, which represents the design to be printed in an independent color space.

[0066] In step 120, the L*a*b* image file from the PCS-CIELAB color space is linked with an ICC output profile 122, transferred to a CMYK printing color space applicable to the printer, and a print image file 124 is created from it.

[0067] In step 126, the print image file 124 for the inkjet printer 8 is then created using a Raster Image Processor (RIP) (see Fig. 1 bis 3 ) converted into a control file 128. The RIP 128 calculates a distribution of numerous drops of each ink for every pixel to be printed and for each ink color.

[0068] In step 129, the design is printed by the inkjet printer 8 with at least one printhead based on the print image file 124 or the control file 128 on a paper web as substrate 4 (see Fig. 1 bis 3 ) in a grid of pixels consisting of lines running perpendicular to the printing direction and columns running in the printing direction, using at least one CMYK color. Due to the width of the paper web 4, the printer 8 has several printheads with individual segments of nozzles for each color.

[0069] The printed design is then measured and evaluated using a measuring system 130 and an evaluation system 143.

[0070] First, a spectral measurement system, which is located in the Fig. 1 bis 3 In step 132, the printed image, marked with reference numeral 10, is measured line by line in the printing direction with a spatial resolution of one pixel, and a spectral measurement image file 134 consisting of rows and columns is generated. The ICMS measurement system from ipac is used for this purpose. The lighting conditions during the measurement process are standardized, and the measurement geometry of the imaging system is 45°:0° (entry:egress).

[0071] In step 136, the spectral measurement image file 134 is then transferred into an L*a*b* measurement image file 138. The conversion is preferably carried out according to the CIE-15 standard, preferably in the CIE illuminant D50 and the 2° standard observer (CIE-15 ASTM 308E, ISO 13655 Annex I).

[0072] In the next step 140, the L*a*b* measurement image file 138 is transferred to the CMYK printing color space using the output profile 122, which was already applied in step 120, and a comparison image file 142 is generated. Thus, the same ICC output profile 122 is applied that was also used to create the CMYK printing image file 124.

[0073] In the evaluation system 143, a reference image file 146 is first generated in a separate step 144 for use in the subsequent process, either once or repeatedly at specified intervals. The reference image file 146 can be generated directly from the print image file 124 from step 120 for the CMYK printing inks. Alternatively, the reference image file 146 can be generated from a previous comparison image file 142' for the CMYK printing inks. A previous comparison image file 142' is a comparison image file 142 from a prior measurement cycle that took place before the current measurement process. This can be an initial measurement process or a repeated measurement process, if necessary.

[0074] The reference image file 146 is then fed to the evaluation system 143, whereby in the following process step 148 the comparison image file 142 and the previously defined reference image file 146 are registered to each other, i.e. by moving, rotating or distorting the pixel matrices to align them.

[0075] The process then continues in step 150, where a difference image file 152 is generated from the comparison image file 142 and the reference image file 146, thus creating a pixel-accurate difference value matrix. The difference values ​​of the difference image file 152 can be determined by averaging the row values ​​of a column over a predefined number of rows.

[0076] In step 154, the difference image file 152 is evaluated. If the magnitudes of the difference values ​​exceed a predefined threshold 156, a malfunction of the printhead of the inkjet printer 8 associated with the deviating pixels is detected. This can involve a change in a section of pixels within a row or an averaged row of the difference image file 152, thus identifying a malfunction of a printhead or a segment of the printhead, i.e., only a part of the printhead. This can then be subsequently addressed and corrected. A large difference in only a few pixels or just one pixel is more likely to indicate contamination or damage to a nozzle or only a few nozzles of the printhead. Such an impairment cannot be rectified by targeted control of the printhead, but can be output as an error message with printhead identification.

[0077] In decision step 158, it is determined whether the process should end with step 200 or continue. If so, the process returns (represented by arrow 160) to the printing system for continuous printing of the design. If the threshold has not been exceeded in any of the difference values, or if no error that can be corrected by the color management system has been identified, the process continues without any changes to the printing system.

[0078] However, if the differential measurement reveals a fault in an entire printhead or a segment of one of the printheads that can be mitigated or corrected by the color management system, for example, by changing a specific control voltage, then the return signal (arrow 160) transmits corresponding information to the printing system. This information is used in the printing system to, for example, modify the control voltage for one of the printheads or individual segments of one of the printheads so that the difference between the target and actual values ​​is minimized. After a change in the printing system, the procedure described in steps 129, 132, 140, 148, 150, and 154 is then continued.

[0079] This allows for the continuous correction of errors that can be corrected by the color management system without interrupting the printing process, thereby improving print quality. This error correction can also be performed for color deviations in the L*a*b* color space that are imperceptible to the human eye and less than ΔE00 = 1. This means that errors only occur in the range imperceptible to the human eye and can be corrected accordingly.

[0080] In the Fig. 5 bis 7 These are example images from a test conducted using the [tool / method - context needed] Fig. 4 The flowchart shown was performed using a spectral measurement system. Fig. 5 The image, presented in grayscale, shows the design of a reproduction of a wood-effect surface with a grain pattern. The predominant color of the design is a light brown with a grain pattern in darker brown tones.

[0081] In this test, a defect in the magenta (M) color channel was intentionally induced, starting with a well-calibrated printing system. First, a print was produced using a well-calibrated printing press (sample 1 M), and then the voltage on one of the magenta printheads was slightly altered (sample 2 M). This voltage change, applied across the width of this printhead, resulted in a color difference of ΔE00 = 1, as measured.

[0082] The Fig. 6a und 6b The square sections of the printed images of samples 1M and 2M, along with the corresponding reference images, also for magenta (M), are shown. Printed images 1M and 2M are each registered pixel by pixel with the reference image M. The respective grayscale images for the printing ink are shown. The mathematical symbols for minus and equal to represent the equation used to subtract the images from each other with pixel-level precision.

[0083] The calculated differences can take on positive and negative numerical values. Therefore, the differences are each added to a predefined standard pixel value, for example, a mean value of 50% within the representable ink intensities between 0% and 100%. Thus, the gray values ​​of the difference image are in approximately the same shade as the original images, even though the difference values ​​are smaller than the absolute pixel values ​​of the original images.

[0084] The Fig. 6a und 6b Each graph also shows a pixel-by-pixel difference value along the rows. The difference values ​​are averaged for each pixel in the row across the entire height of the columns in the displayed images. The x-axis represents the number of pixels in a row, and the y-axis shows the grayscale values ​​of the averaged difference values. A line fluctuating around the zero line (noise) is generally visible. Only at a pixel value of approximately 550 is a spike in the difference values ​​visible in both graphs, caused by a defect in a single nozzle. Such a defect cannot be corrected using the described method. However, this defect can be displayed by the 130 measurement and evaluation system as an error message.

[0085] The graph in Fig. 6a Furthermore, it shows no noticeable fluctuations, so the underlying printing process appears to be fine and no correction of the printing system is necessary.

[0086] The graph in Fig. 6b In contrast, the pixel count between approximately 710 and 980 shows a section with noise but otherwise a consistently lower curve than the zero line. This curve is caused by the printhead malfunction deliberately induced in this test. This section therefore corresponds to the incorrectly controlled printhead, meaning the cause of the error can be corrected by modifying the printhead control.

[0087] As mentioned above, the deliberate change in the printhead control results in a color difference of approximately ΔE00 = 1, which is not visually perceptible in the grayscale values ​​of the displayed print image of sample 2 M. Thus, the method can identify errors even below the visibility threshold. In contrast, a dark vertical stripe is visible in the difference image for sample 2, which also stands out visually from the adjacent areas of the graph. This is primarily because the difference image contains averaged values ​​over a large number of columns, and therefore smaller fluctuations are more visible in the more uniform image.

[0088] Fig. 7a bis 7d Figures 8a to 8d show square sections of the difference images for samples 1 and 2 for the four color channels M (Magenta), C (Cyan, blue), Y (Yellow) and K (Key, black), as well as the corresponding graphs.

[0089] For sample 1, with the exception of an error in individual pixels for samples 1 M and 1 Y, the result is a curve that fluctuates around the zero line, as already described above.

[0090] For sample 2, the following results are obtained initially according to Fig. 8a the behavior accordingly Fig. 6b The drop in the difference curve corresponds to the error induced at the printhead during the test.

[0091] The trends of the difference values ​​in the Fig. 8b bis 8d the other printing inks also show a change in the line segment that is in, even though their assigned printheads had no changed setting. Fig. 8a is clearly visible. While in Fig. 8b only a slight increase in the difference values ​​and in Fig. 8c A slight decrease in the difference values ​​can be observed, shows Fig. 8d for the K-channel a significant, although not as strong, drop as in the M-channel.

[0092] These changes in the difference values ​​arise because the spectral measurement of sample 2, due to the altered proportion of magenta M, produces a different color effect than sample 1. The spectral measurements are converted—as described above—via the CIELab color space and then into the CMYK color space using the ICC output profile. This computational mapping is not always unambiguous and can lead to the described changes in the difference values ​​in color channels that are otherwise unchanged.

[0093] However, since the greatest fluctuation in the difference values ​​occurs in the color channel where the error is present—in this case, the magenta M channel—the error can be attributed to that channel. Correcting this channel will then reduce all difference values. Another rule is that if fluctuations occur in all color channels, the key K (black) channel should not be corrected first.

Claims

1. Method for inline monitoring of the functionality of at least one printhead, - in which a design to be printed is provided as an original image file in an original color space, - in which the original image file is transferred by a color management system with an input profile into an L*a*b* color space and an L*a*b* image file is generated, - in which the L*a*b* image file, from the L*a*b* color space, is transferred with an output profile into a print color space applicable for the printer and a print image file is generated, - in which the design is printed by a printer with at least one printhead on a substrate based on the print image file in a raster of pixels of at least one printing color, the raster consisting of rows extending transversely to the printing direction and columns extending in the printing direction, - in which the printed image, at least in sections, is measured row by row in the printing direction by an optical spectral measurement system with a spatial resolution in the range of a few pixels, in particular with a spatial resolution of one pixel, and a measurement image file consisting of rows and columns is generated, - in which the measurement image file is transferred into an L*a*b* measurement image file, - in which the L*a*b* measurement image file is transferred with the output profile into the print color space and a comparison image file is generated, - in which the comparison image file and a predetermined reference image file are registered with each other, - in which a difference image file is generated from the comparison image file and the reference image file, - in which the difference values of the difference image file are determined row by row, and - in which, if the absolute values of the difference values exceed a predetermined threshold value, a malfunction of the printhead assigned to the deviating pixels is determined, and - in which the method steps are carried out in inline operation during continuous printing operation without interruption of printing.

2. Method according to claim 1, - in which a reference image file is generated from the print image file for at least one printing color.

3. Method according to claim 1, - in which a reference image file is generated from a previous comparison image file for at least one printing color.

4. Method according to any one of claims 1 to 3, - in which the difference values are averaged column by column over a predetermined number of rows, preferably continuously during the printing process.

5. Method according to any one of claims 1 to 4, - in which color deviations in the L*a*b* color space of less than ΔE00 = 1 are detectable.

6. Method according to any one of claims 1 to 5, - in which the at least one printhead of the printer is controlled by the color management system such that the difference values in the difference image file are minimized.