IMAGE REPRODUCTION METHOD AND APPARATUS

FIELD OF THE INVENTION

The present invention relates generally to image reproduction, and, for example, to methods and image reproduction systems for reproducing an input image which take into account the producing technology of the input image.

BACKGROUND OF THE INVENTION

Current copying techniques typically involve digital-image-data processing. Users of color copiers normally expect a copy of an original to have a good picture image quality (PIQ), i.e. to be similar in color to the original and free of artifacts. Typically, color information from the scanned original is converted into color information used to control the amounts of different colorants to be applied by the printer, by means of a device-specific color conversion, also called "color mapping" or "color calibration". Some copier manufacturers allow the user to choose between different copy options, such as Laser, InkJet, Offset, Text, Mixed, Photo, etc. Each of these options may be associated with a different color conversion and may change settings of several copier-imaging pipeline algorithms, such as resolution, sharpening, contrast, compression, etc. U.S. Patent No. 6,353,675 Bl discloses a method and a copier in which the producing technology, called "marking process" (such as inkjet marking process, line-on-line xerographic marking process or rotated-screen xerographic marking process) is automatically detected by generating and analyzing an image power spectrum. For example, in a photographic image, the power continues to decrease rapidly with increasing frequency, whereas, for an inkjet image, the power falls initially and then stagnates at a relatively higher level than the photographic image, which reflects the higher halftone frequencies due to the error-diffusing halftones and stochastic screens commonly used in inkjet printers. Based on this, an automatic determination is made as to whether, for example, a scanned input image was produced by a photographic process or by an inkjet printer. A color calibration specific to the producing technology determined is then chosen depending on the result of the automatic producing-technology determination, and a copy of the input image is then produced using this color calibration.

A similar method and copier are described in U.S. Patent Application Publication

2004/0264771 Al in which the halftone analysis is carried out in selected data blocks or areas that represent constant or near constant image data to suppress variations arising from the image data, in order to better discern the variations due to the halftoning process.

SUMMARY OF THE INVENTION

According to a first aspect, a method is provided of reproducing an input image produced with a producing technology, based on predetermined producing-technology- specific color mappings. The method comprises: obtaining input-image data representing the input image; performing an edge analysis providing edge information and a halftone analysis providing halftone information of the input image; determining, on the basis of at least the edge information and the halftone information, the producing technology of the input image; selecting, from the predetermined producing-technology-specific color mappings, a color mapping associated with the producing technology of the input image determined by the analyses; transforming the input-image data to output-image data by using the selected color mapping; and reproducing the input image by producing an output image from the output- image data.

According to another aspect, a method is provided of reproducing an input image produced with a producing technology, based on predetermined producing-technology- specific color mappings. The method comprises: obtaining input-image data representing the input image; performing a zoning analysis on the input-image data, or a representation of them, identifying one or more zones by their content type information; performing an edge analysis providing edge information and a halftone analysis providing halftone information, wherein the edge analysis and the halftone analysis are performed in different identified image zones depending on the content type information; determining, on the basis of at least the edge information and the halftone information, the producing technology of the input image; selecting, from the predetermined producing-technology-specific color mappings, at least one color mapping associated with the producing technology of the input image determined by the analyses; transforming the input-image data to output-image data by using the at least one selected color mapping; and reproducing the input image by producing an output image from the output-image data.

According to another aspect, an image-reproduction system is provided for reproducing

an input image produced with a producing technology, based on predetermined producing- technology-speciflc color mappings. The system comprises an input-image-data-obtaining system arranged to obtain input-image data representing the input image; an image-analysis system arranged to perform a producing-technology analysis comprising an edge analysis and a halftone analysis and to determine, on the basis of at least the edge information and the halftone information obtained, the producing technology of the input image; a color-mapping- selection system arranged to select, from the predetermined producing-technology-specific color mappings, a color mapping associated with the producing technology of the input image determined by the producing-technology analysis; an input-image data-transforming system arranged to transform the input-image data to output-image data by using the selected color mapping; a print system arranged to reproduce the input image by printing an output image from the output-image data.

According to another aspect, a computer program product is provided which either comprises a machine-readable medium with program code stored on it, or is in the form of a propagated signal comprising a representation of program code. The program code is arranged to carry out a method, when executed in a reproduction device or on a computer coupled to a reproduction device, of reproducing an input image produced with a producing technology based on predetermined producing-technology-specific color mappings. The method comprises obtaining input-image data representing the input image; performing an edge analysis providing edge information and a halftone analysis providing halftone information of the input image; determining, on the basis of at least the edge information and the halftone information, the producing technology of the input image; selecting, from the predetermined producing-technology-specific color mappings, a color mapping associated with the producing technology of the input image determined by the analyses; transforming the input-image data to output-image data by using the selected color mapping; and reproducing the input image by causing an output image to be produced from the output-image data.

Other features of the methods and products disclosed herein will become apparent to those skilled in the art from the following detailed description of embodiments and its accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will now be described, by way of example, and with

reference to the accompanying drawings, in which:

Fig. 1 shows different halftone types and text edges produced with different producing technologies;

Fig. 2 is a high-level architecture diagram of an embodiment of an image-reproduction system in which a producing-technology analysis is performed and a producing-technology- specific color mapping is selected for the image reproduction;

Fig. 3 is a flow diagram illustrating the reproduction of an original in the embodiment of Fig. 2;

Figs. 4 and 5 illustrate an embodiment of an edge analysis; Figs. 6 and 7 illustrate an embodiment of a halftone analysis; Fig. 8 is an architecture diagram similar to Fig. 2 of another embodiment including a zoning analysis;

Fig. 9 is a flow diagram illustrating the reproduction of an original in the embodiment of Fig. 8;

Fig. 10 is an architecture diagram similar to Fig. 8 of an embodiment in which the color- mapping chosen also depends on the main-image content of the image to be reproduced; Fig. 11 is a diagram similar to Fig. 10 of another embodiment in which zones with different image content are reproduced with their associated content-specific color mapping; Fig. 12 illustrates an embodiment in which the different pages of a multi-page original are individually processed according to Figs. 2 to 11; Fig. 13 is a schematic view of an embodiment of an image reproduction device, a copier;

Fig. 14 illustrates an exemplary way to generate producing-technology-specific color mappings.

The drawings and the description of the drawings are of embodiments of the invention and not of the invention itself

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Fig. 1 shows different exemplary halftone types and edges produced by different producing technologies. Before proceeding further with the detailed description of Fig. 1, however, a few items of the embodiments will be discussed.

Imaging devices are usually expected to provide an accurate color reproduction. For a

copier (which is a combination of a scanner and a printer, either in the form of separate devices or forming parts of one and the same device) this means that the copies produced of an original are expected to be similar in color to the original.

Colors can be represented in an objective and device-independent way by numbers in a standard-observer color space, such as the perceptually uniform CIE LAB color space ("CIE" stands for Commission Internationale de l'eclairage, or the International Commission on Illumination, and "LAB" refers to the three axes of the color space, wherein "L" refers to lightness, "A" refers to a red/green axis, and "B" refers to a yellow/blue axis):

Image input and output devices (such as scanners and printers) also represent the colors they detect or produce by numbers referring to color spaces, which are, however, device- dependent, such as RGB (Red Green Blue) typically used in scanners, and CMYK (Cyan

Magenta Yellow Key=black), typically used in printers. Strictly speaking, the RGB or CMYK values used in a scanner or printer do not refer to colors, but they specify the strength of the response of the RGB sensors or the amounts of the different CMYK colorants (inks, toners, etc.) which are to be used to print. hi order to link a certain input or output device with its device-dependent representation of sensor responses or amount of colorants with colors referring to the standard observer, the devices are usually associated with color mappings, for example, an input mapping translating a scanner's measured RGB values to CIE LAB values, or an output mapping translating the CIE LAB values of an image to be printed to the printer's CMYK values. Technically, these color mappings (also called "color profiles") are in the form of data stored in files; they can be either matrix-based or look-up-table-based, and an attempt has been made by a consortium called "International Color Consortium" (ICC) to define a standard file format for device color profiles (e.g. specification ICCl. 2001-12, 2002, File Format for Color Profiles, Version 4.0.0). Color profiles according to the ICC Profile Format Specification are normally called "ICC profiles", sometimes also "ICM profiles". For a particular combination of a certain input device with a certain output device (as, for example, is given in a copier in which the scanner and the printer are parts of the same device) the input mapping associated with the scanner and the output mapping associated with the printer may be combined to one mapping which directly translates the scanner's measured sensor responses (e.g. expressed in an RGB color space) to the printer's colorant-amount values (e.g. expressed in a CMYK color space), without going through a standard-observer color space.

Color calibrations (input mappings, output mappings, or combined mappings) are

usually determined experimentally by means of a measurement device which mimics the standard observer's color vision. For example, this may be a colorimeter (a measurement device with three sensors having spectral sensitivities like the standard observer's three receptors, called long-, medium- and short-wavelength (or L, M and S) receptors), or a spectrophotometer (which is a detector for the whole visible spectrum) with suitable software to calculate the LMS receptors' response to the measured spectrum. First, an original with different colors serving as a calibration image is produced, usually a set of different color patches, which is also called a "profiling target". There are also standards for profiling targets; for example, a standard target for profiling CMYK output devices is the IT8.7/3-target. Then, a copy of the original is produced with the copier (using any existing color mapping). The colors of both the original and the copy of the target are then measured with the standard- observer-measurement device, and the color differences between corresponding colors in the original and the copy are determined (if a perceptually uniform standard-observer color space is used, such as CIE LAB, distances between points in the color space predict how different the two colors will appear to a human observer; the value of a color difference is usually called δE). The measured color differences are then "fed back" to the given color mapping which was used for this procedure, to modify it such that, when the modified color mapping is used for the same procedure, the color differences, in principle, become zero and copies very similar in color to original images are produced. Suitable profiler software to perform such a color calibration is included in some commercially available color-management software packages, such as for example Heidelberg/PrintOpen®; Delta-E/Profiler®; GretagMacbeth/ProfileMaker®; miation/Spectral Profiler®; ITEC Color Solutions/ColorBlind®; Monaco Systems/MonacoProfiler®; Scitex/Profile Wizard®, etc. However, it has been recognized that the use of one and the same color mapping for originals produced with different producing technologies does not lead to a very accurate color reproduction. There are many different producing technologies, such as traditional photo, digital photoprint, dye sublimination, dry-toner electrophotography (DEP), liquid-toner electrophotography (LEP), inkjet (continuous-flow or drop-on demand inkjet, the latter including thermal inkjet, piezoelectric inkjet and solid-ink inkjet), offset, etc. Furthermore, even within such a "family" of producing technologies (such as DEP or inkjet), usually each printer and printer software manufacturer uses its own way to produce continuous-tone images, as will be explained in more detail below. A more accurate color accuracy is achieved

in the embodiments by using color mappings which are not only specific to the producing- technology "family" of the input image, but are also specific to printers of different manufacturers, or even different types of printers of the same manufacturer, within one family. The terms "producing technology" and "printing technology" used herein are therefore not limited an indication of the producing-technology "family" (such as DEP, LEP, inkjet, offset, etc.); rather, it may also, but does not necessarily have to, indicate different producing technologies in printers of the same family of different manufacturers, or even of the same manufacturer.

To illustrate the use of producing-technology-specific color mappings, let us assume, for example, that there are ten different producing technologies. Then, the above-mentioned procedure of calibrating the copier, i.e. experimentally determining color mappings for accurate color reproduction is carried out ten times using ten profiling targets produced with the different producing technologies. As a result, ten different producing-technology-specific mappings are obtained. If then, as in the embodiments, an original is copied with that color mapping which is associated with the original's producing technology, a more accurate color reproduction is, in many cases, obtained than that obtained by using one and the same color mapping for the different originals.

Since a manual choice of copy options is inconvenient and error-prone, many users of copiers always use a standard copy option; furthermore, even skilled users will normally only be able to distinguish certain print-technology-families (such as traditional photo and inkjet), but will not be able to distinguish all the existing families, let alone different printers of different manufacturers, etc.). hi some embodiments, therefore, the producing technology of the original is automatically determined by the scanner/copier before the original is actually reproduced, without requiring user input.

Based on what has been described above, in some of the embodiments, reproducing an input image thus includes the following activities. Input-image data which represent the input image are obtained, hi some embodiments, these input-image data are bitmap data. They may be obtained by scanning the input image which is, in that case, for example, an image printed on a print media. As mentioned above, in some embodiments, the scanner is a part of a copier or, in other embodiments, a separate device. In some of the embodiments, the input-image data are not produced by scanning a printed input image, but already-existing input-image data are received, for example over a network. These input-image data may already be in the form of bitmap data, or they may be in another form (e.g. represented by vector-graphics). Such

non-bitmap data, in some embodiments, are rendered, i.e. transformed, into bitmap data.

Then, a producing-technology analysis is automatically carried out on the input data, whereby the producing technology of the input image is determined. Depending on the result of this determination, a selection from producing-technology-specific color mappings is made. In some of the embodiments, these color mappings are predetermined, for example by the color- calibration method described above. Of course, although skilled users may carry out such color calibrations on their own in order to define and update their device's color mappings in an individual manner, normally such color mappings will be produced by the reproducing device's manufacturer or a color-management service provider and will either be delivered together with the reproducing device or be downloadable via the Internet. The selection process results in the choice of the color mapping associated with the input image's producing technology determined in the producing-technology analysis, hi some of the embodiments in which the predetermined color mappings are stored in a storage (e.g. a storage within a copier), the selection includes retrieving the selected color mapping from the storage. The input-image data are then transformed to output-image data by using the selected color mapping, and the input image is then reproduced (e.g. printed) by producing (e.g. printing) an output image from the output-image data. hi principle, an automatic detection of the original's producing technology may be solely based on a halftone analysis, as, for example, described in the publications US 6,353,675 Bl and U.S. 2004/0264771 Al mentioned at the outset. It has been recognized, however, that the discernibility of the different producing technologies may be enhanced in many instances by performing an edge analysis, in addition to the halftone analysis. Thus, in some of the embodiments, the producing-technology analysis performed on the image data includes an edge analysis providing information about edges found in the input image, and a halftone analysis, providing halftone information about the input image; and the input image's producing technology is determined on the basis of the edge information and the halftone information. Of course, the use of edge information and halftone information is not exclusive, it may be supplemented by additional information from further types of analyses.

Halftoning is a way to produce (apparent) continuous-tone images, such as photographs. There are two basic ways to produce such continuous-tone images, "contone imaging" and "halftoning" (so-called "dithering is sometimes considered as a third basic way (see Johnson, page 45) but is handled herein as a particular sort of halftoning, and is therefore encompassed by the term "halftoning"). Li digital contone imaging, e.g. used in digital photoprinting and

dye-sublimation devices, the printed image is made up of overlapping colorants of each primary color with no spaces between them, so that color transitions are produced which are very smooth, and the resulting images are very photorealistic (see Johnson, p. 46).

By contrast, in halftoning techniques, the image is broken down into little dots or spots of the primary colorants. In some halftoning techniques, the darker portions of the image have larger spots, and the lighter areas have smaller spots, hi other halftoning techniques, a halftone dot (sometimes called pixel) is made up of clusters of printer dots which usually have the same size. Different tones are then achieved by varying the number of printer dots filled with color. For example, if a pixel has four cells, five different tones can be produced (no dots: all you see is the paper; one dot: 25% tone; two dots: 50% tone; three dots: 75% tone; four dots: 100% tone). At a sufficient viewing distance, the viewer's brain merges all the spots to give him/her the impression that what he/she is seeing is one, smooth image. Halftones are therefore optical illusions tricking the viewers into thinking that they are seeing continuous tones. hi some of the halftoning techniques, the dots are regularly shaped and are placed onto the print media in a regular arrangement (these are also called "amplitude-modulated screening halftones). Certain branches of digital printing, specifically inkjet and electrophotography, now often use another halftone technique, called frequency-modulated or stochastic halftoning. Instead of regularly shaped and regularly arranged halftone dots, irregularly shaped halftone dots are arranged in a stochastic, or random arrangement (as mentioned above, this is also called "dithering", see Johnson, pp. 45-48). hi practice, nearly each printer and copier manufacturer uses its own halftoning/dithering method.

The right column of Fig. 1 illustrates, in an exemplary manner, how a uniform-color patch is printed with different printing technologies. The first three rows of Fig. 1 refer to laser-electrophotography printers with dry toner (Laser-DEP) made by three different manufacturers, Canon, HP and Xerox. As can be seen in Fig. 1, the DEP printer by Canon uses a random-halftoning technique, the one by Xerox uses regular-pattern, or cluster halftoning, whereas the DEP printer by HP uses a combination of regular-pattern and random halftoning. The 4 ^th and 5 ^th rows of Fig. 1 pertain to a laser-electrophotography printer with liquid toner (Laser-LEP) by Indigo and an offset printer by Heidelberg; both use a regular- pattern halftoning. The 6 ^th to 8 ^th rows of Fig. 1 refer to different inkjet printers by HP which all use random-halftoning techniques. Finally, the 9 ^th row of Fig. 1 shows a traditional photo which is a continuous-tone image without halftone structures. As can be taken from Fig. 1, in

some of the cases the halftone patterns of printers belonging to the same family, using the same type of halftoning (e.g. random halftoning), and even being made by the same manufacturer, are distinguishable (as, for example, in the case of the three inkjet-produced patches in Fig. 1). On the other hand, in some of the cases, printers of different families produce very similar halftoning patterns (such as the color patches printed by LEP and Offset in Fig. 1).

In the following table, the halftoning shapes typically used in printers of the different producing-technology families are summarized:

In some of the embodiments, the input image, or a certain part of it, is analyzed by a halftone analysis to discern the different halftoning shapes, some of which are illustrated in Fig. 1. In some of the embodiments, the halftone analysis includes generation of a power spectrum of the spatial frequencies found in the input image. Such a spatial- frequency power spectrum is, in some embodiments, generated by a Fourier transform, e.g. by means of an FFT software. In other embodiments a histogram of the inter-maxima/minima distance between minority pixels is generated (as, for example, described in US 2004/0264771 Al in connection with Fig. 8) which is, to some extent, similar to a Fourier spectrum of spatial frequencies. The resulting Fourier spectrum (or minority-distance histogram, see Fig. 8 of US 2004/0264771 Al) is then analyzed with respect to the frequency peaks (or distance peaks). Halftones with clusters, such as the 3 ^τd to 5 ^th rows in Fig. 1, have clear frequency peaks (or distance peaks, as in Fig. 8 of US 2004/0264771 Al). Random halftones have no, or small peaks, but their power spectrum is different from a continuous-tone power spectrum since it shows a higher level at higher frequencies (see, for example, Fig. 5 of US 6,353,675 Bl).

In other embodiments, the halftone analysis includes a spatial statistical analysis, i.e. an analysis which measures and evaluates spatial characteristics of the halftone structure obtained

in a statistical manner. For example, fluctuations of the optical density of the input image are measured, for example, at a scale of about 0.04 mm according to the "graininess" definition in the International Standard ISO 13660, first edition of 1 September 2001, chapter 5.2.3. Cluster halftones tend to exhibit a higher graininess than random halftones; furthermore, different instances of the same type of halftoning (e.g. cluster or random) may exhibit different graininess. For example, the random halftoning of the inkjet printer "DesignJet 800PS" (7 ^th row in Fig. 1) is grainier than that of the inkjet printer "Deskjet 990" (6 ^th row in Fig. 1). The spatial statistical analysis may, for example, be made on the basis of commercially available picture-quality analysis software using ISO- 13660 methods, for example by ImageXpert Inc. (www.imagexpert.com) or QEA Inc. (www.qea.com). In some of the embodiments, both a spectral analysis and a spatial statistical analysis of the sort described are carried out, and their results are combined.

As already mentioned above, in some of the embodiments, an edge analysis is carried out to determine the producing technology of the input image, in addition to the halftone analysis,. While the halftone analysis focuses on variations in the image that do not arise from the image content, but occur in halftone processes (these variations are best seen in a uniform patch, as shown at the right-hand side of Fig. 1), edges, by contrast, represent the most pronounced variations arising from the image content itself. Edges are mainly found in characters (text) and lines. The different producing technologies differ in the manner in which they reproduce edges. Therefore, the edge analysis is complementary to the halftone analysis, and combining the results of both, as it is done in some embodiments, enhances the ability to discriminate between the different producing technologies.

Three different edge characteristics maybe used in the edge analysis: (i) satellites; (ii) edge roughness (also called "raggedness" in ISO 13660); and (iii) edge sharpness (referred to as "blurriness" in ISO 13660). (i) satellites are extraneous marks that appear near an edge. They are agglomerations of colorant particles near an edge that are visible as distinct marks at standard viewing distance. According to ISO 13660, chapter 5.3.7, for example, "near an edge" means "within 500 μm" from the edge (this 500 μm region is called "character field" in ISO 13660); furthermore, the visibility requirement means, according to ISO 13660, that one dimension of a colorant particle must at least have 100 μm for the particle to be considered as a "satellite". As can be seen in the text-edge column of Fig. 1, edges printed by laser-DEP have satellites on both sides of horizontal and vertical edges; inkjets usually only produce satellites in the printing

direction, at one side, or sometimes both sides of an edge, whereas laser-LEP, Offset and photo-producing technologies usually produce no satellites. As can also be seen from Fig. 1, there are further differences within one and the same producing-technology family; for example, the inkjet printer "Deskjet 990" produces more satellites than the inkjet printer "DesignJet 5000PS". Therefore, the satellite analysis does provides some discrimination not only between families, but also between different types of the same producing-technology family.

(ii) edge roughness (or "raggedness" in ISO 13660) refers to a geometric distortion of an edge along its longitudinal direction. A ragged edge appears rough or wavy rather than smooth or straight. An exemplary measure of raggedness, according to ISO 13660 (chapter 5.3.2), is the standard deviation of residuals from a line fitted to the edge, calculated perpendicular to the fitted line (excluding satellites). As can be seen in Fig. 1, edges produced by laser-LEP, Offset and photo are excellent with respect to roughness, those produced with laser-DEP are usually good, and those produced with inkjet are usually medium to pure. Again, the edge roughness may vary between printer types within the same producing-technology family (as, for example, between laser-DEP "HP Color LaserJet 8500" and laser-DEP "Xerox" (see Fig. 1), so that the measure of edge roughness also provides information which helps to discriminate between different printer types within the same producing-technology family.

(iii) edge sharpness is referred to as "blurriness" in ISO 13660 (actually, sharpness and blurriness are complementary terms). Edge sharpness refers to the density transition from background to line/character ("white-black transition" in the case of a white substrate and a black character or line), seen in a direction perpendicular to the edge. For example, according to ISO 13660, chapter 5.3.1, a measure of the blurriness is the average distance between the edge's inner and outer boundaries (wherein these boundaries correspond to 10% and 90% of the difference between the line/character- and background-reflectance factors). In other words, an edge is blurry if there is a relatively smooth density transition from background to line/character on a path perpendicular to the edge. As can be seen in the text edge column of Fig. 1, laser-DEP, laser LEP and inkjet produce sharp edges, offset produces edges a little bit less sharp, and photo has smooth edges.

In the following table, the edge characteristics typically obtained in printers/copiers of the different producing-technology families are summarized:

In some embodiments the edge analysis includes all the three analysis methods mentioned, a satellite analysis, an edge roughness analysis and an edge sharpness analysis. In other embodiments only two of the these three analysis methods are used, e.g. the satellite and edge roughness analyses, the satellite and edge sharpness analyses, or the edge roughness and edge sharpness analyses. In still further embodiments only one of the three analysis methods is used. In some of the embodiments, the different analysis methods are implemented on the basis of commercially available picture-quality analysis software, e.g. using the ISO- 13660 methods described above, for example the analysis software by ImageXpert Inc. (www.imagexpert.com) or QEA Inc. (www.qea.com). In some of the embodiments, the location of satellites in the satellite analysis is based on the bwmorph tool within the Matlab Mathematics Package program marketed by The MathWorks, Inc. (www.mathworks.com). The normal bwmorph (image, 'clean') function of Matlab is modified so as to extend to more than one-pixel detection so that bwmorph finds larger colorant particles.

Each of the five analysis methods described (spectral analysis, spatial statistical analysis, satellite analysis, edge roughness analysis, edge sharpness analysis) may provide its result in the form of probability values indicating, for the different producing technologies, the probabilities that the input image analyzed has been produced with the respective producing technology. In some embodiments, the results of the different analysis methods carried are combined, for example, by means of a probability matrix. The overall result are combined probability values indicating, for the different producing technologies, the probabilities that the input image analyzed has been produced with the respective producing technology. The most probable producing technology is the best "estimate"; it is considered as the producing technology with which the original image was produced, and a color mapping associated with

this producing technology is chosen and used in the subsequent image reproduction. In other embodiments, the results of the different analyses are combined in a cascaded manner. For example, the results of the halftone analyses and the edge analyses are combined separately so that a combined halftone-analysis result and a combined edge-analysis result are obtained, and the combined halftone- and edge-analysis results are then combined to eventually determine the best estimate of the producing technology.

In some embodiments, a zoning analysis is performed before the producing-technology analysis (halftone analysis and edge analysis) is carried out. The aim of the zoning analysis is to distinguish zones with text/line content (i.e. zones having mainly edges) from zones with graphic content and photo image content (i.e. zones which are more uniform). As will be explained in more detail below, the "content segregation" thus obtained is later used in the producing-technology analysis such that the edge analysis is selectively performed in one or more text/line zones found, and, correspondingly, the halftone analysis is carried out in one or more graphics/photo zones.

Zoning analysis algorithms are known to the skilled person, e.g. from US 5,767,978. For example, a zoning analysis used in some of the embodiments identifies high-contrast regions (called "strong edges" in US 5,767,978), typical for text/line content, and low-contrast regions ("weak edges"), typical for continuous-tone zones, such as graphics or photo content. In some embodiments, the zoning analysis calculates the ratio of strong and weak edges within a pixel region; if the ratio is above a predefined threshold, the pixel region is considered as a text/line region which may be combined with other text/line regions to form a text/line zone; low ratio pixels are treated analogously to define graphics/photo zones. Other zoning analyses count the dark pixels or analyze the pattern of dark and bright pixels within a pixel region in order to identify text/line elements. The different types of indication for text/line and graphics/photo such as the indicator based on strong/weak edge recognition and the one based on background recognition, may be combined in the zoning analysis. As a result of the zoning analysis, text/line zones and graphics/photo zones are found and identified in the bitmap-input image, e.g. as illustrated in Fig. 3 of US 5,767,978.

Often, the zoning analysis, on the one hand, does not require the image information to be analyzed with a very high resolution. On the other hand, some zoning algorithms require a considerable computational effort. Therefore, in some embodiments, the zoning analysis is not carried out on the bitmap-input image, but on an input-image representation having a lower resolution than the (original) input image. Finding zones in the lower-resolution version of the

image requires less computations and thus enables the zoning analysis to be performed without considerable delays (i.e. in a "real-time manner"), using a given computer/microprocessor with limited performance. The zoning information obtained (e.g. an indication which pixel belongs to which zone) is then easily upscaled to the full-resolution input image on which the subsequent producing-technology analysis is then performed. hi some embodiments, the content segregation achieved by the zoning analysis is then used in the subsequent producing-technology analysis in the following manner: the edge analysis is performed on one or more text/line zones, whereas the halftone analysis is performed on one or more graphics/photo zones. This content-selective application of the edge and halftone analyses is not mandatory since the edge- and halftone-analysis methods described also work on mixed contents; therefore, in alternative embodiments no zoning analysis is carried out, and the edge and halftone analyses are carried out in a content-non- selective manner (e.g. on a whole input page, irrespective of its content). However, the content segregation and content-selective application of the edge and halftone analyses described enhances the analyses' ability to discriminate between the different producing technologies and is therefore used in some embodiments.

Depending on the input image which is to be reproduced in a particular case, there may be no text/line zone or no graphics/photo zone, hi some of the embodiments, if no text/line zone is found, edges are edge-analyzed within graphics/photo zones to extract edge information in an at least approximated manner. Conversely, if there is no graphics/photo zone found in the input image, but big-font colored text is present in a text zone, this big- font colored text is analyzed with the halftone analysis to obtain halftone information in at least an approximated manner. The more information (text/lines and uniform graphics/photo) is analyzed, the better is the estimate, i.e. the higher is the probability of a correct determination of the input image's producing technology. The information obtained in the producing-technology analysis, i.e. the best estimate of what producing technology was used to make the input image, is then passed, for example, to a "print pipeline engine" controlling the print pipeline of the printer or copier which is to produce the output image. The print pipeline engine selects the color mapping which is associated with the producing technology determined, and retrieves it from the printer's or copier's memory to perform the best input/output color conversion for the output image, hi exceptional cases, depending on the input-image content and/or if the input image was produced with an unknown producing technology, the producing-technology analysis may fail.

In some embodiments, an indicator for such an analysis failure is the probability distance between the greatest and the second-greatest probability determined. If these two probability values are too close to clearly discriminate between the two, no reliable estimate of the producing technology can be made. In such cases, in some embodiments, a default average color mapping is chosen which is a good compromise for all common producing technologies. The producing technology is not the only parameter for which the provision of different color mappings specific to this parameter is useful. For example, the perceived colors often depend on the media on which an image is printed (e.g. normal paper, photo paper, transparency, etc.), and the picture image quality of a copy may be further improved if the color mappings used in the reproduction process are also specific to the print media on which the original is printed and/or the print media on which a copy is to be printed. Therefore, in some embodiments, not only color mappings specific to the producing technology of the original are stored, but rather one or more different color mappings for each printing technology specific to the print media are stored and appropriately selected, depending on the producing technology detected and the print media used. In some embodiments, this is only done with regard to the original's print media, in other embodiments this is only done with regard to the print media of the copy to be produced, but in still other embodiments this is done with regard to both the original's and the copy's print media (in the latter case, assuming three different print-media types, there are nine different print-media-type combinations, and, correspondingly, nine print-media specific color mappings for each producing technology). Some embodiments are equipped with a functionality that automatically classifies the input- and/or output-print media, for example on the basis of the print-media classification technique described in U.S. patent No. 6,725,207. In other embodiments, the user can manually specify the input- and/or output-print media as a "copy option".

Another parameter influencing the color reproduction, besides the producing technology (and, if applicable, the print media) is the image content. Therefore, in some embodiments, the choice of the color mapping to be used does not only depend on the producing technology determined, but also on the image content, as determined by means of the zoning analysis. For example, in some embodiments, different color mappings are stored for text/line content and graphic/photo content, for each producing technology. When an input image is to be reproduced, the right color mapping is automatically selected, and used in the reproduction process, based on the automatic detection of the producing technology and the image content. Besides the color mapping, different settings in the copy pipeline, such as resolution,

sharpening, contrast, compression, etc. are automatically chosen depending on the image content determined in the zoning analysis.

If the input image (which is typically what is printed on one page) only has one or more zones of one content type (text/line content or graphic/photo content) the color mapping (and copier settings, if applicable) associated with this content type is chosen. However, input images often have mixed contents, e.g. text/line zones and graphics/photo zones. In some of the embodiments, in such a case, the main-content type determines which one of the different content-type-specific color mappings is selected and used for the reproduction. What is the "main-content type" can be defined in different ways. In one embodiment, the main-content type of the input image is that content type with the largest zone in the input page. In another embodiment, the main-content type is the content type which has the largest total area (= sum of the areas of all zones belonging to the same content type), in another embodiment, the main-content type is the content type with the greatest number of zones, etc.

Whereas in the embodiments mentioned above each input page is always reproduced in a uniform manner with the same color mapping (which, as explained above, may be optimized with regard to the main-content type), in alternative embodiments the individual zones with different content types within one and the same page are reproduced with different color mappings; each zone is reproduced with that color mapping (and copier setting, if applicable) which is associated with the content type of the respective zone. For example, if the upper half of an input page is text, and the lower half is photo image, the upper half is then printed with the color mapping associated with text, and the lower half is printed with that associated with photo (as explained above, in the embodiments the choice of the color mapping also depends on the producing technology determined and, if applicable, on the print media). There are different alternative ways in which such zone-individual color mappings are applied to the different zones in the reproduced image. In some of the embodiments, the original color values in the bitmap-input image of the pixels belonging to a particular zone are replaced (i.e. over-written) by other color values, according to the color mapping associated with the particular zone considered. This modified bitmap-input image is then processed through the print pipeline and printed in a usual manner. In other embodiments, the content-type information, as determined in the zoning analysis, is added to the bitmap-input image. The information added to a pixel is also called "tag", and the process of adding these data to the bitmap image is called "tagging". In the print pipeline, when the bitmap-input image is transformed, for each pixel that color mapping is used which is indicated by the pixel's

content-type tag.

As mentioned above, the producing-technology analysis (as well as the print-media analysis and/or the zoning analysis) and the selection of a color mapping based on the analysis results are carried out without considerable delay (i.e. in a "real-time manner"). Therefore, in some of the embodiments, when a multi-page document is to be reproduced, the automatic analysis to determine the producing technology (and to determine the print media and/or content types, if applicable) and the automatic selection of the color mapping associated with it are carried out individually for each page of the multi-page document, so that pages of the document produced with different producing technologies are individually reproduced with their associated (and thus different) color mappings. hi some embodiments the image-reproduction system is a copier, i.e. an integrated scanner/printer system; other embodiments are systems with a separate scanner and printer, wherein a multi-purpose computer (and, in some cases, a network connection) may be between the scanner and the printer. In other embodiments the "input-image-obtaining system" is not a scanner, but a different sort of image-recording device, such as a video camera. In still other embodiments, the "input-image-obtaining system" is for example, a computer or an embedded processing unit of a printer/copier programmed to receive images which are already in the form of digital data (e.g. vector-graphics data) and "render" them, i.e. transform them into pixelized image data (i.e. bitmap data). Some embodiments have an image-analysis system arranged to perform the combined producing-technology analysis described above (including edge and halftone analyses) to determine the producing technology of the input image. A color-mapping-selection system is arranged to select, from the predetermined producing-technology-specific color mappings, a color mapping associated with the producing technology of the input image determined by the combined producing- technology analysis. An input-image data transforming system is arranged to transform the input-image data to output-image data by using the selected color mapping, which are finally printed out by the printer.

Different embodiments use different digital-printing technologies to print the output- image data. Some of the embodiments use electrophotographic printers (laser printers) with solid or liquid toner; other embodiments use inkjet printers, for example thermal or piezo drop-on-demand printers, continuous-flow or solid-ink ink-jet printers.

The embodiments of the computer program product bestow the functionalities described herein on a multi-purpose computer or a processing unit of a printer/copier. The "machine-

readable medium" on which the program code may be stored is any medium that is capable of storing or encoding data representing a computer program. The term "machine-readable medium" shall accordingly be taken to include, for example, solid state memories and, removable and non-removable, optical and magnetic storage media. Another representation of the program code is in the form of a propagated signal which enables the program code to be distributed over a network.

Turning now to Fig. 2, a high-level architecture diagram of an embodiment of an image- reproduction system 1 is shown. Different color mappings 2 specific to different producing technologies (denoted by "X", "Y" and "Z" in Fig. 2) are stored in a storage 3. An input image, or original 4 is to be reproduced, the output image, or copy, is denoted by 14. In the example of Fig. 2 the original is a physical print media (such as a sheet of paper) with the image printed on it. First, the original 4 is "rendered"; e.g. a scanner 5 scans the original 4 and produces digital image data 6 representing the input image. The input-image data 6 are, for example, in the form of a bitmap in which, as usual, the color of each pixel is represented by digital density values for each of the scanner's primary colors. In other embodiments, the input image is already available in a form represented by data, e.g. by vector-graphics data, hi such embodiments, another rendering component (replacing the scanner 5) converts this data representation into a bitmap, as the one shown at 6 in Fig. 2. Then, a producing-technology (PT) analyzer 7 is applied to the input-image data 6. It includes an edge analyzer 8 and a halftone analyzer 9. Let us assume in the example of Fig. 2 that the result of the combined edge and halftone analyses is that the producing technology (PT) of the original 4 is "X". This result is communicated to a mapping selector 10 which, in turn, selects the mapping 2 associated with the producing technology "X" and retrieves it from the storage 3. The mapping selector 10 delivers the selected mapping 2 to an image transformer 11 which transforms the input-image data 6 into output-image data 12 using the selected mapping 2. The mapping 2 translates a representation of a color in the color space of the scanner 5 (which is, for example, the scanner's device-dependent RGB color space) into a representation of (ideally) the same perceived color in the color space of the printer 13 (for example, the printer's color space is a CMYK color space). The printer 13 then prints the copy 14. Due to the fact that the color mapping 2 chosen is specifically adapted to the original's producing technology a better picture image quality will typically be achieved than that achieved with color mappings non-specific to the producing technology.

Fig. 3 is a flow diagram of the activities carried out in the embodiment of Fig. 2. First,

the input-image data are rendered. For example, at 20, input-image data 6 (Fig. 2) in the form of bitmap data are produced by scanning the input image 4 (Fig. 2). Alternatively, at 30, an already-existing data representation of the input image (e.g. a vector-graphic representation) is transformed into bitmap data. The input-image data in bitmap form are then analyzed at 40 with respect to the input image's producing technology, here by performing two analyses, an edge analysis 41 and a halftone analysis 42. At 50, the producing technology of the input image is determined on the basis of the analysis results, e.g. by combining probability values for the producing technologies considered obtained by the different analyses. At 60, the producing-technology-specific color mapping which is associated with the producing technology determined is selected. A set of predetermined producing-technology-specific color mappings, from which said color mapping is selected, is stored in a memory in the color-reproduction device (e.g. in the copier), but may, in other embodiments be loaded, e.g. from a color-mapping server, via a network. At 70, the output-image data 12 (Fig. 2) are produced using the selected color mapping. At 80, the output image 13 (Fig. 2) is reproduced. In some embodiments, the reproduction is made by a printer (which may be a part of a copier) on a print media, such as paper; in other embodiments the reproduction is, for example, a display on a monitor.

Figs. 4 and 5 illustrate an embodiment of the edge analysis 41. It is composed of three sub-analyses 411, 412, 413 which are sensitive to different characteristics of printed edges. The first sub-analysis 411 detects and evaluates satellites near printed edges. As mentioned above, satellites are agglomerations of colorant particles near an edge that are visible as distinct marks at standard viewing distance. According to the ISO 13660 definition (chapter 5.3.7) which is used in this present embodiment and is illustrated in Fig. 5a, "near an edge", for example, means within a line field which is an area of 500 μm from the edge (more precisely, from the edge threshold which is defined by a line fitted to the edge points at which the reflectance of the background is reduced by a certain factor, e.g. by 60%). Furthermore, as illustrated in Fig 5a, the visibility requirement means, according to ISO 13660, that one dimension of a colorant particle must at least have 100 μm for the particle to be considered as a "satellite". The measure to characterize the occurrence of satellites according to ISO 13660 which is calculated in this embodiment is the ratio of the area of satellites to the total area of the line field.

The second sub-analysis 412 analyzes edge roughness, or "raggedness", which is a geometric distortion of an edge along its longitudinal direction. A measure of raggedness

according to ISO 13660 (chapter 5.3.2), which is used in this embodiment and is illustrated in Fig. 5b is the standard deviation of "residuals" (see Fig. 5b) from the edge threshold, calculated perpendicular to the edge threshold (excluding satellites).

The third sub-analysis 413 analyzes edge sharpness. Edge sharpness refers to the density transition from background to line/character ("white-black transition" in the case of a white substrate and a black character or line), seen in a direction perpendicular to the edge.

According to a measure defined in ISO 13660 (chapter 5.3.1), a measure of the blurriness (which is complementary to the edge sharpness) which used in this embodiment and is illustrated in Fig. 5c is the average distance (seen in a direction perpendicular to the edge) between the edge's inner and outer boundaries, wherein e.g. these boundaries correspond to 10% and 90% of the density difference between background and line/character (or about 10% and 90% difference between the line/character- and background-reflectance factors).

Fig. 6 illustrates an embodiment of the halftone analysis 42. It is composed of two sub- analyses 421, 422 which are sensitive to different halftone characteristics. The first sub- analysis 421 is illustrated in Fig. 7a. It produces a power spectrum of the spatial frequencies in the input image, e.g. by applying an FFT (Fast Fourier Transformation) to the entire image or part of it (a "region of interest"). An schematic exemplary power spectrum of spatial frequencies is shown in Fig. 7a. Halftoning techniques with regular halftone patterns exhibit distinct frequency peaks at characteristic positions and distances which are therefore a sort of "fingerprint" for certain producing technologies. The power spectrum is analyzed with respect to the occurrence of frequency peaks and, if applicable, their positions and distances, as shown in Fig. 7a, to obtain the "fingerprint" and infer, or at least narrow down, the producing technology.

The second sub-analysis 422 analyzes the spatial structure of halftone patterns in a statistical manner. It is illustrated in Fig. 7b. According to the "graininess" definition in ISO 13660 (chapter 5.2.3), on the basis of which the second sub-analysis 422 of this embodiment is carried out, fluctuations of the density of the input image are measured at a scale of about 0.04 mm. To this end, a "region of interest", for example a square of 12.7mm x 12.7 mm is divided into 100 uniform, non-overlapping square tiles, as shown in Fig. 7b. Li each tile (with a dimension 1.27mm x 1.27mm) 900 evenly spaced, non-overlapping measurements of the density are made (which corresponds to the measurement scale of about 0.04 mm mentioned above), and for each tile i, the standard deviation s; of these measurements from the average for this tile is calculated. Based on this, a normalized square root of quadratic sum of the

standard deviations Sj for all tiles is calculated as the graininess measure.

Each of the measures provided by the five sub-analysis methods is sensitive to different characteristics of the producing technology. Each analysis method delivers, as a result, a list of probability values indicating, for the different producing technologies in question, the probabilities that the input image analyzed has been produced with the respective producing technology. These probability values are combined, for example, by means of a probability matrix. The thus combined probability values indicate, for the different producing technologies, the probability that the input image analyzed has been produced with the respective producing technology. The most probable producing technology is then chosen as the best "estimate" of the producing technology. While the edge and halftone analyses are applied to the entire input image, in the embodiment of Fig. 2, an alternative embodiment of an image-reproduction system 1' is shown in Fig. 8 in which a content segregation and, subsequently, a content-specific application of the edge and halftone analyses are carried out. As in Fig. 2, a scanner 5 scans the original 4 and produces digital-image data 6, in the form of a bitmap, representing the input image. In addition to what has been explained in connection with Fig. 2, a zoning analysis is carried out to segregate different content (text/line from graphics/photo). In some embodiments, the zoning analysis is performed on the (full-resolution) input-image data 6. In the embodiment shown in Fig. 8, however, the zoning analysis is carried out on a version of the bitmap data 6 a lower spatial resolution. To this end, a lower-resolution-image (LRI) generator 15 produces lower-resolution image data 16 from the input image data 6, for example by combining 3x3 adjacent pixels into one pixel. A content analyzer 17 performs the zoning analysis on the lower-resolution image data 16. As a result, the image represented by the lower-resolution image data 16 is segmented into zones with different contents. In the example of Fig. 8, the content analyzer 17 has found a "text zone" and a "photo zone"; the corresponding zoning data are denoted by "18" (the terms "text zone" and "photo zone" are used here instead of "text/line zone" and "graphics/photo zone" for the purpose of brevity). The zoning data 18 indicate which pixels of the lower-resolution image data 16 and - after an upscaling process to the full-resolution input image data 6 - which pixels of the input image data 6 belong to which zone. The combination of the input image data 6 and the "upscaled" zoning information is denoted by 6' in Fig. 8. A producing-technology (PT) analyzer T is applied to the input-image data/zoning information 6' such that its edge analyzer 8 performs the edge analyses on the text zone, and its halftone analyzer 9 performs the halftone analyses

on the photo zone (or zones, if there are more than one zones of the respective type). As in Fig. 2, it is assumed here that the result of the combined edge and halftone analyses is that the producing technology (PT) is "X". As in Fig. 2, this is communicated to a mapping selector which, in turn, selects the mapping 2 associated with the producing technology "X" and retrieves it from the storage 3. The mapping selector 10 delivers the selected mapping 2 to an image transformer 11 which transforms the input-image data 6 into output-image data 12 using the selected mapping. Also the further components and activities (a printer 13 prints the copy 14) are as in Fig. 2.

Fig. 9 is a flow diagram of the activities carried out in the embodiment of Fig. 8. First, as in Fig. 3, the input-image data are rendered, either by scanning the input image at 20 or by transforming an already-existing data representation of it into bitmap data, at 30. At 90, the lower-resolution-image data 16 (Fig. 8) are produced. At 91, a zoning analysis is performed and content type information (e.g. text/line or graphics/photo) is provided for each identified zone. At 40', the producing-technology analysis methods are applied to the zones in a content- selective manner. This is, at 41', the edge analysis is performed in the text-line zone(s), and at 42' the halftone analysis is performed in the graphics/photo zone(s). As in Fig. 3, at 50, the producing technology is determined on the basis of the analysis results; at 60, the producing- technology-specific color mapping which is associated with the producing technology determined is selected; at 70, the output-image data are produced using the selected color mapping; and at 80, the output image is reproduced. Fig. 10 illustrates an embodiment in which the color mappings not only depend on the producing technology, but also on the image content (such as text or photo). In the exemplary Fig. 10, 2 x 3 color mappings 2' (here called "maps") are shown, namely maps for two different contents, "text" and "photo", and for three different producing technologies, "X", "Y" and "Z". The maps 2' are stored in a storage 3'. As in Figs. 7 and 8, the zoning analysis results in the input image data 6 being segmented in zones with different image content. In the example of Fig. 10, the segmented input-image data 6' have a smaller text zone and a larger photo zone. As in Fig. 8, the producing-technology (PT) analyzer T includes an edge analyzer 8 applied to the text zone and a halftone analyzer 9 applied to the photo zone. The exemplary result of the combined edges and halftone analysis is that the producing technology (PT) is "X". This result is input into a mapping selector 10'. In addition to what has been described in connection with Fig. 8, a main-content determiner 19 is provided. It compares the different content zones found in the

zoning analysis, and determines what is called the "main-content type". In the example of Fig. 10, the main-content type is that content type with the largest total area which, in the present instance, is the content type "photo". The result of the main-content determiner 19 is also input into the mapping selector 10' which, in turn, selects that mapping 2' which is associated both with the producing technology and the content type determined. In the example of Fig. 10, this is the mapping "MAP X PHOTO". As in Fig. 8, the mapping selector 10' delivers the selected mapping 2', "MAP X PHOTO" to the image transformer 11, which uses it to transform the input-image data 6' to output-image data 12. In other words, an input image is reproduced with that color mapping which is adapted to its main content (here "photo"), so that only a minor part of the image (here the text part) is reproduced with a color mapping not adapted to it.

By contrast, in the embodiment of Fig. 11 the output image is not reproduced with a uniform color mapping, such as the main-content color mapping in Fig. 10, but zones with different image content are reproduced individually with their associated content-specific color mappings. Accordingly, in Fig. 11, the content types of the zones found in the zoning analysis are input into the mapping selector, here denoted by 10". In the example of Fig. 11, a text zone and a photo zone have been found (as in Fig. 10); consequently, the content types "text" and "photo" are input into the mapping selector 10', in addition to the result of the producing-technology analysis, which is again "X". On the basis of this input information, the mapping selector 10" selects all the mappings associated with the content zones found, and the producing technology determined, here the "MAP X TEXT" and "MAP X PHOTO", and delivers them to the image transformer 11'. The image transformer 11' uses the "MAP X TEXT" to transform the colors in the text zone, and the "MAP X PHOTO" to transform the colors in the photo zone. Colors of pixels which do not belong to any zone maybe transformed with a default average mapping. As a consequence, different parts of one and the same output image (which is normally what is printed on one page of a print media, e.g. one side of a paper sheet) are reproduced with different color settings which are adapted to the contents of the different image parts.

Fig. 12 illustrates an embodiment in which the different pages of a multi-page document to be reproduced are processed individually. The exemplary multi-page document shown consists of three pages 4, denoted by "PA ¹ GE 1", "PAGE 2", and "PAGE 3". The pages 4 are individually rendered, and the resulting data representations 6 (input-image data) are individually subjected to the producing-technology analysis according to one of the

embodiments described in connection with Figs. 2 to 10 above. Let us assume, in the example of Fig. 12, that the producing technology of "PAGE 1" is "Y", that of "PAGE 2" is "Z" and that of "PAGE 3" is "X". According to what has been described above in connection with Figs. 2 to 10, one or more color mappings are then selected for each page individually, depending on the page's producing technology (and, if applicable, on the print-media types and/or zone contents). In the example of Fig. 12, the "MAP Y" is selected for "PAGE 1", the "MAP Z" is selected for "PAGE 2" and the "MAP X" is selected for "PAGE 3". Each page is then color-transformed with its selected map, or maps. The copies 14 of the pages which are finally printed out are therefore individually reproduced with color mappings associated with their producing technology, in order to obtain a good picture image quality in multi-page copy jobs in the case of originals with pages produced using different producing technologies. The processing of each single page is carried out in "real-time manner" during the copy job such that no delay noticeable to the user occurs.

In Fig. 13, as an example of an image-reproduction device, a copier 100 is depicted schematically. The illustrated segmentation of the copier 100 in different components is functional and does not necessarily imply a corresponding structural segmentation. For example, some of the functional components shown can be merged with other functional components, or can be composed of several functional sub-components.

The copier 100 has an input-image data-obtaining component in the form of a scanner 102 with a scan bed 101 which can be covered by a scan lid 103. The scanner 102 is able to scan any type of image printed on a printing media, e.g. a paper sheet, and to generate a digital representation thereof, e.g. a bitmap image represented in an RGB color space. The copier 100 also has a memory 107 for storing digital image data. The image data need not necessarily originate from the scanner 102, but may likewise be data from an external computer, a digital camera, etc., transmitted to the copier 800 via a network connection. The memory 107 also stores the different producing-technology-specific and content-type-specific color mappings, e.g. in the form of look-up tables according to the ICC color-profile standard.

An image processor 105 provides the functionalities of a producing-technology analyzer, a color-mapping selector and an image transformer, as described above. The image processor 105 is arranged to receive input-image data, here bitmap-input images, from the scanning part 102 or, if applicable, the memory 107 or the network connection. The image processor 105 produces output-image data, based on the color mapping(s) selected. A print processor 109 is arranged to transform the output-image data into print data and instructions

for a printer unit 111, which prints the image on an output media, e.g. on an output image, paper sheet, hi some embodiments, the actual color transformation is not carried out by the image processor 105, but by the print processor 109. To this end, the image processor 105 delivers the image data and the selected color mapping to be used to transform the image's color values or a reference to this color mapping. Fig. 14 illustrates an exemplary way to generate the producing-technology-specific color mappings used in Figs. 2 to 13. At 200, a target is produced with a certain producing technology, which is called "producing technology X" in Fig. 14. As mentioned above, a target is usually a calibration image with different patches of representative colors. On the one hand, at 201, the colors in the original target, as perceived by the standard observer, are measured, for example, by means of a colorimeter or a spectrophotometer. On the other hand, at 203, the target is reproduced with the reproducing device and any already-available color mapping (which is, e.g. in the form of an ICC profile). Furthermore, at 203, the colors in the copy of the target are measured, in the same manner as the colors in the original were measured at 202. Finally, at 204, the measured colors in the original and the copied target are compared, and a corrected color mapping (e.g. ICC profile) is generated on the basis of the measured colored differences between the copied and original target. The correction is such that the color differences are minimized, if this target were copied again with the corrected color mapping. The resulting corrected color mapping is specific to the producing technology X and enables input images produced with this producing technology to be reproduced with a good picture image quality. The same procedure is carried out using targets produced with the other producing technologies of interest, in order to obtain a set of producing-technology- specific color mappings for all the producing technologies of interest. In order to make the set of color mappings also specific to different print media and/or different contents, the procedure of Fig. 14 may also be carried out for targets (or copies) printed on different print media and/or targets with different image contents.

Thus, a general purpose of the disclosed embodiments is to provide methods and products which enable copies of color images to be made with a good picture-image quality. AU publications and existing systems and methods mentioned in this specification are herein incorporated by reference. Although certain methods and products constructed in accordance with the teachings of the invention have been described herein, the scope of coverage of this patent is not limited thereto. On the contrary, this patent covers all embodiments of the teachings of the invention

fairly falling within the scope of the appended claims either literally or under the doctrine of equivalents.