TWI846481B - High-fidelity printing reproduction system and method thereof - Google Patents

Abstract

The present invention relates to a high-fidelity printing reproduction system and method thereof. The method comprises the following steps: analyzing each pixel of an image file by a server to obtain a to-be-converted color data with XYZ/La*b* value; converting the to-be-converted color data into a to-be-converted spectral data by the server; finding out one or more predicted color matching ink-set data corresponding to the to-be-converted spectral data; and printing the image file by using the predicted color matching ink-set data and a printing equipment.

Description

High-fidelity printing and copying system and method thereof

The present invention relates to a high-fidelity printing copying system and method thereof, and in particular to a copying system and method thereof for conversion technology applied to high-fidelity printing.

In recent years, in order to effectively save costs (Cost Down) and meet market demand, printing companies have introduced supporting processes such as combined printing, digital printing (Digital Printing), and ink-saving software. It is foreseeable that the era of winning by quantity in the past has ended, and the small-volume and diversified printing market is taking shape. "Quality" will become an important indicator of future competitiveness.

With the rapid development of digital media and the widespread use of handheld digital products, general consumers have become accustomed to the color rendering mode of digital media, such as Adobe RGB or sRGB. The color gamut of the four primary colors of CMYK in traditional printing has gradually failed to meet the needs of consumers. Therefore, high-saturation printing and high-fidelity printing, which are more expensive than traditional four-color printing, have the opportunity to become another way out for the printing industry. Therefore, how to guide the printing industry to move forward in the direction of "high saturation, high fidelity, and high quality" will be the goal of our efforts.

In order to improve the quality of digital images on print output devices, in the process of implementing the Color Management System (CMS), in order to make the colors of various media operate accurately, the color characteristic interpretation model is one of the indispensable stages. Chroma and concentration are the two most commonly used measurement units in CMS. Through the quantification technology of chroma and concentration, it has gradually been able to meet the needs of customers in a certain light source environment.

But in fact, the ambient light source for customers to calibrate printed products cannot always be at a fixed color temperature, and the variable light source causes the problem of "metamerism". Therefore, in order to reduce the probability impact of "metamerism" phenomenon, the consistency of spectral mapping must be considered while modeling color characteristics.

In addition, compared with the color source of the input image, the color gamut that can be expressed in traditional four-color printing has a lower ductility, so it is also easy to cause the impact of the probability of metamerism. Therefore, if the color gamut that can be expressed can be increased during printing production, it is also an important means to reduce the impact caused by the metamerism problem. Therefore, based on the above research background and motivation, the development of spectral wide-gamut printing color characteristic modeling technology has become the most significant direction for improving printing color quality.

In addition, both CMYK and RGB belong to a device-dependent color space. In other words, different devices detect and reproduce specific CMYK or RGB values differently. Even if the same CMYK or RGB color material combination is used, the response level of each individual color channel will vary depending on the carrier provided by the manufacturer or the display device. Even the same device may be different in different observation environments or time.

In the printing field, different printing equipment has different ink characteristics. For example, if you want to convert CMYKR to RGB using different equipment in different printing plants, you cannot do so directly. This is because the ink characteristics of CMYKR are different, so it is not possible to directly convert them using a single formula, and the converted results will be very inaccurate.

Therefore, this case uses deep learning to analyze each pixel of the image file and convert it into an XYZ/La*b* color data value under a specific light source, and then converts the color data value into a conversion spectrum data, and then finds one or more predicted color matching ink set data corresponding to the conversion spectrum data. The predicted color matching ink set data has at least one color matching color data set, and the color matching color data set includes a key color material usage dot value. In this way, when converted to six or seven-color printing output, the final obtained six or seven-color high-fidelity complex image reproduction print can present more dazzling, more powerful, and more shocking colors. Therefore, the present invention should be an optimal solution.

A high-fidelity printing and copying system is applied to a printing device. The high-fidelity printing and copying system includes: an input device for inputting at least one image file; and a server device for receiving the image file. The server device has a built-in color conversion application program. The color conversion application program is used to analyze each pixel of the image file to obtain a color data value. The color data value is At least each pixel has an XYZ/La*b* value and a converted hue angle value, and the color conversion application converts the color data value into a converted spectral data through a spectral imaging model, and then finds one or more predicted color matching ink set data corresponding to the converted spectral data through a color matching conversion model, and then uses the predicted color matching ink set data to print a copy print through the printing device.

More specifically, the color conversion application has a first depth training unit and a color tracing unit. The server device stores a first color data set. The first depth training unit performs model training on the first color data set to establish the spectral imaging model. The color tracing unit uses the spectral imaging model to convert the color data value into the converted spectral data, wherein the first color data set at least includes spectral data of each pixel of multiple ink set samples and default color data values with XYZ/La*b* values under a specific light source.

More specifically, the color data value and the default color data value are in RGB color space format.

More specifically, the first color data set is obtained by detecting the spectral data of each pixel of a plurality of ink set samples by a spectrum detection device, wherein the ink set samples include at least four color inks, and the ink set samples are CMYK or CMYK plus one or more special color inks.

More specifically, the color conversion application has a second depth training unit and a color matching unit. The server device stores a second color data set. The depth training unit performs model training on the second color data set to establish the color matching conversion model. The color matching unit finds one or more predicted color matching ink set data corresponding to the conversion spectral data through the color matching conversion model, wherein the second color data set at least includes spectral data of each pixel point of multiple ink set samples, color matching color data set and hue angle data.

More specifically, the color conversion application further includes a color gamut determination unit, which selects a color matching color data set that meets the conversion hue angle value from a plurality of predicted color matching ink set data by corresponding the hue angle data with the color matching color data set.

A high-fidelity printing copy method is applied to a printing device. The steps of the high-fidelity printing copy method are: (1) each pixel of an image file is analyzed and converted into a color data value having an XYZ/La*b* value under a specific light source through a server device; (2) the server device converts the color data value into a conversion spectrum data; and (3) the server device finds one or more predicted color matching ink set data corresponding to the conversion spectrum data, and then uses the predicted color matching ink set data to print a copy of the printed product through the printing device.

More specifically, the server device stores a first color data set and performs model training on the first color data set to establish a spectral imaging model, which is used to convert the color data value into the converted spectral data, wherein the first color data set at least includes spectral data of each pixel of multiple ink set samples and a default color data value with an XYZ/La*b* value under a specific light source.

More specifically, the server device stores a second color data set and performs model training on the second color data set to establish a color matching conversion model, which is used to find one or more predicted color matching ink set data corresponding to the converted spectral data, wherein the second color data set at least includes spectral data of each pixel point of multiple ink set samples, a color matching color data set and hue angle data.

More specifically, the server device selects a color matching color data set that matches the converted hue angle value from a plurality of predicted color matching ink set data by corresponding the hue angle data to the color matching color data set.

Other technical contents, features and effects of the present invention will be clearly presented in the following detailed description of the preferred embodiments with reference to the drawings.

Please refer to FIG. 1 , the high-fidelity printing and copying system includes an input device 1 , a server device 2 and a printing device 3 , wherein the input device 1 is used to input at least one picture image file.

As shown in FIG. 2A , the server device 2 includes at least a processor 21 and a readable recording medium 22 (the readable recording medium 22 is a memory element with a memory function). The readable recording medium 22 stores a color conversion application 221. In addition, the image file, the first color data set, and the second color data set input by the input device 1 or other external devices are also stored in the readable recording medium 22.

As shown in FIG. 2B , the color conversion application 221 has a color space analysis unit 2211, which is used to analyze the image file to analyze the color data value of the image file. The color data value has the XYZ/La*b* value and/or the converted hue angle value of each pixel under a specific light source.

The specific light source in this case can be D50, D65 or other light sources. The choice of different light sources varies according to needs.

The color space analysis unit 2211 is used to analyze the image file to obtain XYZ values, and then convert the color into La*b* values, and further convert it into LCH to obtain the converted hue angle value.

The converted hue angle value is a representation of converting the LAB color space into the LCH (lightness-hue-saturation) color space, and the conversion formula is as follows: C = sqrt(A ² + B ² ) H = arctan(B / A) (1) Where C is chromaticity, H is hue angle, and A and B are the color opposite axes in the LAB color space.

As shown in FIG. 2B , the color conversion application 221 has a first depth training unit 2212 . The server device 2 stores a first color data set. The first depth training unit 2212 performs model training on the first color data set to establish the spectral imaging model.

The color data value and the default color data value are in RGB color space format, such as AdobeRGB format, sRGB format, etc.

The first color data set at least includes spectral data of each pixel of a plurality of ink set samples and a preset color data value after spectrum conversion, and the preset color data value has an XYZ/La*b* value under a specific light source.

The spectral data is obtained by detecting the spectral data of each pixel of a plurality of ink set samples by a spectrum detection device.

The ink set sample includes at least four color inks, wherein the ink set sample is CMYK or CMYK plus one or more special color inks (eg, five-color CMYKR, six-color CMYKRG, seven-color CMYKRGB).

After model training, it is possible to establish the association model data between the spectral data and the default color data value with the XYZ/La*b* value. As shown in FIG. 2B , the color conversion application 221 has a color tracing unit 2213. The color tracing unit 2213 converts the color data value (XYZ/La*b* value) stored in the readable recording medium 22 into the converted spectral data (SR, Spectral Reflectance) through the spectral imaging model.

As shown in FIG. 2B , the color conversion application 221 has a second depth training unit 2214, which performs model training on the second color data set to establish the color matching conversion model. The second color data set at least includes spectral data of each pixel of multiple ink set samples, a color matching color data set (C value, M value, Y value, K value or/and R, G, B, wherein C value, M value, Y value, K value are basic color values, and R, G, B are key color material usage dot values) and hue angle data (hue angle range).

The ink set samples are first collected from four-color (CMYK), five-color (CMYKR, CMYKG, CMYKB), six-color (CMYKRG, CMYKGB, CMYKBR), and seven-color (CMYKRGB) samples, and the above samples are tested for the spectrum data of each pixel through spectrum detection equipment. The ink set samples (CMYK, CMYKR, CMYKG, CMYKB, CMYKRG, CMYKGB, CMYKBR, CMYKRGB) all correspond to different hue angle data (primary hue angle and hue angle range).

As shown in Figures 3A and 3B, if four colors (CMYK) are used, since the C value 41, the M value 42, and the Y value 43 can be mixed and replaced by the K value 44, the C value 41 is reduced to the C value 41', the M value 42 is reduced to the M value 42', and the Y value 43 is reduced to the Y value 43', among which the K value 44 can be regarded as the main key colorant usage dot value.

As shown in Figure 4, if five colors (CMYKR) are used, since the M value 42' and the Y value 43' can replace the R value (red), the M value 42' can be reduced to M value 42'', and the Y value 43' can be reduced to Y value 43'', where the R value 45 can be regarded as the main key color material usage dot value.

After model training, it is possible to establish the associated model data of the color matching color data set and the spectral data. As shown in FIG. 2B , the color conversion application 221 has a color matching unit 2215. The color matching unit 2215 finds one or more predicted color matching ink set data corresponding to the converted spectral data through the color matching conversion model.

Since each transformed spectral data may find one or more sets of predicted color matching ink set data, if there are multiple sets of predicted color matching ink set data, they can be filtered by the transformed hue angle value.

As shown in FIG. 2B , the color conversion application 221 further has a color gamut determination unit 2216. The color gamut determination unit selects a color matching color data set that matches the converted hue angle value by corresponding the hue angle data with the color matching color data set. Since the hue angle data is a hue angle range, it first determines in which color ink/material area the chromaticity value falls, and then performs a ratio calculation (Compute Ratio Distance) on the converted hue angle value and the hue angle data to determine which set of predicted color matching ink set data should be selected.

Each of the predicted color matching ink set data has a color matching color data set (C value, M value, Y value, K value or/and the main key color material usage dot value). Therefore, after determining which set of predicted color matching ink set data it is, the color matching color data set (C value, M value, Y value, K value or/and the main key color material usage dot value) can be used to copy the color closest to the image file, and finally the copy printing product can be printed out through the printing device 3.

The high-fidelity printing and copying method of the present invention, as shown in FIG. 5, comprises the following steps: (1) analyzing and converting each pixel of an image file into a color data value having an XYZ/La*b* value 501 through a server device; (2) the server device converts the color data value into a conversion spectrum data 502; and (3) the server device finds one or more predicted color matching ink set data corresponding to the conversion spectrum data, and then uses the predicted color matching ink set data for printing 503 through the printing device.

The advantages of the high-fidelity printing reproduction system and method provided by the present invention compared with other conventional technologies are as follows: (1) Through deep learning, the present invention analyzes each pixel of the image file to obtain a color data value with an XYZ/La*b* value, and then converts the color data value into a conversion spectrum data, and finally finds the conversion spectrum data and the conversion hue angle value as a color matching data. (2) When the present invention is converted to six or seven-color printing output, the resulting six or seven-color high-fidelity complex image reproduction print can present more dazzling, more powerful, and more shocking colors. (3) In this case, after the high-dynamic image is separated by the color characteristic rendering mode of the high-fidelity printing equipment, the printed color is reproduced by multi-color high-fidelity, which has greatly reduced the impact of metamerism and made it closer to the optimal situation of isomerism.

The present invention has been disclosed as above through the above-mentioned embodiments, but they are not used to limit the present invention. Anyone familiar with this technical field and having common knowledge can make some changes and modifications without departing from the spirit and scope of the present invention after understanding the above-mentioned technical features and embodiments of the present invention. Therefore, the scope of patent protection of the present invention shall be determined by the definition of the claim items attached to this specification.

1: Input device

2: Server equipment

21: Processor

22: Readable recording media

221:Color conversion application

2211: Color space analysis unit

2212: First Depth Training Unit

2213: Color Origin Unit

2214: Second Depth Training Unit

2215: Color matching unit

2216: Color gamut determination unit

3: Printing equipment

41: C value

41’: C value

42:M value

42’: M value

42’’: M value

43:Y value

43’: Y value

43’’: Y value

44:K value

45: R value

[Figure 1] is a schematic diagram of the overall structure of the high-fidelity printing and copying system and method of the present invention. [Figure 2A] is a schematic diagram of the structure of the server equipment of the high-fidelity printing and copying system and method of the present invention. [Figure 2B] is a schematic diagram of the structure of the color conversion application of the high-fidelity printing and copying system and method of the present invention. [Figure 3A] is a schematic diagram of the implementation of the key color of the four-color printing of the high-fidelity printing and copying system and method of the present invention. [Figure 3B] is a schematic diagram of the implementation of the key color of the four-color printing of the high-fidelity printing and copying system and method of the present invention. [Figure 4] is a schematic diagram of the implementation of the key color of the five-color printing of the high-fidelity printing and copying system and method of the present invention. [Figure 5] is a schematic diagram of the process of the high-fidelity printing and copying system and method of the present invention.

1: Input device

2: Server equipment

3: Printing equipment

Claims (8)

A high-fidelity printing and copying system is applied to a printing device. The high-fidelity printing and copying system includes: an input device for inputting at least one picture image file; and a server device for receiving the picture image file. The server device has a built-in color conversion application program. The color conversion application program is used to analyze each pixel of the picture image file to obtain a color data value. The color data value has at least an XYZ/La*b* value and a conversion hue angle value of each pixel. The color conversion application program converts the color data value into a conversion spectrum data through a spectral imaging model, and then finds the corresponding color spectrum through a color matching conversion model. One or more predicted color matching ink set data of the data, and then the predicted color matching ink set data is used for printing through the printing device: wherein the color conversion application has a first depth training unit and a color tracing unit, the server device stores a first color data set, the first depth training unit performs model training on the first color data set to establish the spectral imaging model, and the color tracing unit converts the color data value into the converted spectral data through the spectral imaging model, wherein the first color data set at least includes the spectral data of each pixel of multiple ink set samples and the default color data value with XYZ/La*b* value. A high fidelity printing and copying system as described in claim 1, wherein the color data value and the default color data value are in RGB color space format. A high-fidelity printing and copying system as described in claim 1, wherein the first color data set is a spectrum data of each pixel of a plurality of ink set samples detected by a spectrum detection device, the ink set samples contain at least four color inks, and the ink set samples are CMYK or CMYK plus one or more special color inks. The high-fidelity printing and copying system as described in claim 1, wherein the color conversion application has a second depth training unit and a color matching unit, the server device stores a second color data set, the depth training unit performs model training on the second color data set to establish the color matching conversion model, and the color matching unit finds one or more predicted color matching ink set data corresponding to the conversion spectral data through the color matching conversion model, wherein the second color data set at least includes spectral data of each pixel of multiple ink set samples, color matching color data set and hue angle data. As described in claim 4, the high-fidelity printing and copying system, wherein the color conversion application further has a color gamut determination unit, and the color gamut determination unit selects a color matching color data set that meets the conversion hue angle value from a plurality of predicted color matching ink set data by corresponding the hue angle data with the color matching color data set. A high-fidelity printing and copying method is applied to a printing device. The steps of the high-fidelity printing and copying method are: a server device analyzes each pixel of a picture image file to obtain a color data value with an XYZ/La*b* value; the server device converts the color data value into a converted spectrum data, wherein the server device stores a first color data set, and performs model training on the first color data set to establish a spectrum type The imaging model is used to convert the color data value into the converted spectral data, wherein the first color data set at least includes the spectral data of each pixel of multiple ink set samples and the default color data value with XYZ/La*b* value; and the server device finds one or more predicted color matching ink set data corresponding to the converted spectral data, and then uses the predicted color matching ink set data for printing through the printing device. The high-fidelity printing copying method as described in claim 6, wherein the server device stores a second color data set, and performs model training on the second color data set to establish a color matching conversion model, the color matching conversion model is used to find one or more predicted color matching ink set data corresponding to the conversion spectral data, wherein the second color data set at least includes spectral data of each pixel of multiple ink set samples, color matching color data set and hue angle data. A high-fidelity printing and copying method as described in claim 7, wherein the server device selects a color matching color data set that meets the converted hue angle value from a plurality of predicted color matching ink set data by corresponding the hue angle data to the color matching color data set.

Images