Method, apparatus, and system for artificial intelligence-based photograph color correction for printing

Abstract

One embodiment of the present invention relates to an artificial intelligence-based method, apparatus, and system for color correction of a photograph for printing, which uses artificial intelligence to predict the surface characteristics and ink absorption behavior of a printing medium and dynamically optimizes the color profile of the photograph accordingly. According to one embodiment, an AI-based photo color correction method for printing is provided, wherein the device includes a processor, memory, a communication module, and a non-transient storage medium, and the method is performed by the processor by executing a program stored in the non-transient storage medium, the method comprising: a step of acquiring an original photo file; a step of generating a preprocessed photo file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black and white conversion on the original photo file; and a step of generating a corrected photo file by using an AI color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set to be suitable for print output.

Description

Technology Field

[0001] The following embodiments relate to an artificial intelligence-based method, apparatus, and system for color correction of a photograph for printing, which uses artificial intelligence to predict the surface characteristics and ink absorption behavior of a printing medium and dynamically optimizes the color profile of the photograph accordingly. Background Technology

[0002] In the traditional newspaper and publishing printing industry, editors or designers go through a process of converting photos edited in an RGB environment on a monitor into a CMYK format for actual printing. However, there are limitations in that the color of the output varies significantly depending on the skill level of the operator or the calibration status of the monitor, and even for the same image, the physical properties of the roll paper used differ depending on the newspaper's edition or page allocation, resulting in uneven print quality.

[0003] In particular, in high-speed multi-color rotary printing processes using rotary presses, the variability in ink absorption rate and dot gain is very high due to environmental variables such as paper surface porosity, temperature, and humidity, as well as the ink die-cutting sequence.

[0004] Conventional general-purpose color conversion software has been unable to predict and compensate for physical and chemical ink absorption behaviors or the material characteristics of individual paper surfaces in advance. This has resulted in critical print quality degradation, such as set-off defects caused by excessive ink application or blurring of gradations in shadow areas. Therefore, there is an urgent need for the introduction of an intelligent automation solution capable of multidimensionally quantifying environmental variables throughout the entire printing process and analyzing them using artificial intelligence to automatically generate color conversion parameters optimized for print output. Prior art literature

[0005] (Patent Document 0001) KR 10-1513431 B (Patent Document 0002) KR 10-1731258 B The problem to be solved

[0006] The problem that an embodiment of the present invention aims to solve is to provide an AI-based photo color correction method, apparatus, and system for printing that predicts ink absorption behavior in a time series based on publishing metadata and environmental variables and dynamically calculates CMYK conversion weights optimized for paper characteristics through artificial intelligence, in order to overcome the limitations of a simple color conversion method that cannot reflect the physical printing environment (paper material, temperature, humidity, etc.) and relies on the subjective judgment of a conventional operator as described above. means of solving the problem

[0007] According to one embodiment, an AI-based photo color correction method for printing is provided, wherein the device includes a processor, memory, a communication module, and a non-transient storage medium, and the method is performed by the processor by executing a program stored in the non-transient storage medium, the method comprising: a step of acquiring an original photo file; a step of generating a preprocessed photo file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black and white conversion on the original photo file; and a step of generating a corrected photo file by using an AI color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set to be suitable for print output.

[0008] Additionally, the step of generating the correction photo file comprises: a step of obtaining publication metadata including newspaper edition information, page information, and date setting information associated with the original photo file; a step of determining a target CMYK profile and a total ink amount limit value for each printing condition based on the publication metadata; a step of calculating predicted values ​​for the paper material characteristics and ink absorption rate of the target page based on the publication metadata; a step of calculating correction input features including a histogram per RGB channel, a luminance histogram, a saturation distribution, a color temperature deviation, a local contrast value, an edge sharpness value, and a shake index from the preprocessed photo file; and a step of dividing the preprocessed photo file into at least two of a human skin area, a sky area, a vegetation area, a shadow area, a highlight area, a text area, and a graphic area. The method may include the step of inputting the correction input feature quantity, region segmentation result, target CMYK profile, and total ink amount limit value into an artificial intelligence color correction model to generate a set of color correction parameters including region-specific brightness correction coefficients, saturation correction coefficients, white balance correction coefficients, sharpening intensity, black generation amount, and CMYK channel-specific conversion weights; the step of dynamically adjusting the CMYK channel-specific conversion weights by reflecting the paper material characteristics and ink absorption rate prediction values; and the step of generating a corrected photo file by using the set of color correction parameters to limit the color deviation of the human skin area to within a preset reference range, suppressing gradation loss in dark and bright areas, performing dot gain pre-compensation for the target CMYK profile, and then mapping the preprocessed photo file to the target CMYK profile.

[0009] And, the step of calculating the predicted values ​​for the paper material characteristics and ink absorption rate comprises: a step of querying paper history information including the lot number, paper supplier, paper basis weight, coating type, calendaring processing history, manufacturing date, and storage period of the paper roll to be used for printing the target page, based on the newspaper company plate information, page information, and date setting information included in the publishing metadata; a step of generating a basic material characteristic vector including the fiber orientation, surface porosity, coating layer thickness deviation, surface roughness, and degree of whiteness degradation of the target paper, based on the paper history information; a step of generating an environmental equilibration vector including the ambient temperature immediately before printing, relative humidity, average humidity of the paper storage space, elapsed time since opening the paper, and the printing press preheating status, based on the date information and printing time information; and a step of querying the reference ink coverage rate, reference color load index, and reference dot gain index corresponding to the target page from a previously stored print history database by page type, based on the plate information and page information. A step of calculating a set of absorption behavior parameters including an initial penetration velocity of the paper surface, a time-delayed diffusion coefficient, a transverse blurring coefficient, an anisotropic absorption coefficient according to the winding direction, and a saturation critical absorption amount, using the above-mentioned basic material characteristic vector, moisture equilibrium vector, reference ink coverage rate, reference color load index, and reference dot gain index as inputs; and a step of calculating a first-order absorption rate, a stabilized absorption rate, and a dot expansion rate for each CMYK channel based on the above-mentioned set of absorption behavior parameters.The method may include: a step of determining the set of fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the basic material characteristic vector, and the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount as the paper material characteristics of the target surface based on the stabilization absorption rate and dot expansion rate for each CMYK channel; a step of calculating the ink absorption contribution for each channel by weightedly combining the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the basic material characteristic vector, the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount, and the stabilization absorption rate and dot expansion rate for each CMYK channel by channel; and a step of calculating the predicted ink absorption rate for each channel for the C channel, M channel, Y channel, and K channel, and the predicted integrated ink absorption rate for the entire target surface based on the ink absorption contribution for each channel.

[0010] In addition, the step of calculating the above absorption behavior parameter set comprises: generating a material normalized feature vector by converting the fiber orientation, surface porosity, coating layer thickness deviation, surface roughness, and degree of whiteness degradation included in the above basic material characteristic vector into a preset normalization range; generating moisture-related state values ​​by normalizing the ambient temperature, relative humidity, average humidity of the paper storage space, and elapsed time after opening the paper included in the above moisture equilibrium vector, and calculating a paper moisture activity index by weighted combining the moisture-related state values; generating heat-related state values ​​by normalizing the ambient temperature, relative humidity, printing press preheating state, and elapsed time after opening the paper included in the above moisture equilibrium vector, and calculating a thermal stabilization index by weighted combining the heat-related state values; and calculating a directional penetration coefficient for the main orientation direction of the paper and a direction orthogonal to the main orientation direction, respectively, based on the fiber orientation and surface porosity included in the above material normalized feature vector, and generating a biaxial transmission matrix from the directional penetration coefficient. A step of calculating a coating delay coefficient indicating the degree of delay in the initiation of ink penetration into the substrate layer based on the coating layer thickness deviation and surface porosity included in the material normalization feature vector and the paper moisture activity index; a step of calculating a surface retention coefficient indicating the degree of ink retention on the surface based on the surface roughness included in the material normalization feature vector and the paper moisture activity index; a step of calculating a moisture swelling coefficient indicating the degree of wet swelling of the paper surface based on the paper moisture activity index, thermal stabilization index, and the surface porosity and coating layer thickness deviation included in the material normalization feature vector; a step of calculating a saturation limit correction coefficient indicating the amount of variation in the ink acceptance limit based on the moisture swelling coefficient, thermal stabilization index, and the degree of whiteness reduction included in the material normalization feature vector;A step of generating a reference load vector representing the reference printing load state of a target surface using the reference ink coverage rate, reference color load index, and reference dot gain index; a step of calculating a directional load response vector by combining the reference load vector and a biaxial transmission matrix; a step of inputting the directional load response vector, coating lag coefficient, surface residence coefficient, moisture swelling coefficient, and saturation limit correction coefficient into a time response function generation model to calculate the capillary penetration gradient immediately after ink application, the diffusion decay gradient over time, the lateral blur sensitivity, and the critical time to reach saturation, and generating a set of time response functions based on the capillary penetration gradient, diffusion decay gradient, lateral blur sensitivity, and critical time to reach saturation; and a step of calculating, respectively, the initial penetration velocity at a first reference time point, the time-delayed diffusion coefficient, the lateral blur coefficient, the anisotropic absorption coefficient, and the critical absorption amount for saturation at a second reference time point. and may include the step of generating the set of absorption behavior parameters by combining the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount.;

[0011] Additionally, the step of calculating the first absorption rate, stabilization absorption rate, and dot expansion rate for each of the above CMYK channels comprises: a step of setting a reference ink application amount, a reference dot area, and a reference dot diameter for each of the C channel, M channel, Y channel, and K channel based on the above reference ink coverage rate, reference color loading index, and target CMYK profile; a step of generating an hourly cumulative penetration curve, a surface residue curve, a first dot expansion curve in the main orientation direction, and a second dot expansion curve in a direction orthogonal to the main orientation direction for each of the above C channel, M channel, Y channel, and K channel by applying the above initial penetration rate, time-delayed diffusion coefficient, transverse blurring coefficient, anisotropic absorption coefficient, and saturation threshold absorption amount per channel; and a step of calculating a first cumulative penetration amount per channel and a first surface residue amount per channel from the ink application time point to a first reference time point based on the above hourly cumulative penetration curve and surface residue curve. A step of normalizing the first cumulative penetration amount per channel by the reference ink application amount per channel and performing a subtraction correction on the first surface residue amount per channel to calculate the first absorption rate for each of the C channel, M channel, Y channel, and K channel; a step of calculating the additional penetration amount per channel and the stabilized surface residue amount per channel from the first reference time point to the second reference time point based on the time-based cumulative penetration amount curve, the surface residue amount curve, and the saturation threshold absorption amount; a step of normalizing the stabilized cumulative penetration amount, which is the sum of the first cumulative penetration amount per channel and the additional penetration amount per channel, by the reference ink application amount per channel, and performing a correction on the stabilized surface residue amount per channel and the reference dot gain index to calculate the stabilized absorption rate for each of the C channel, M channel, Y channel, and K channel;The method may include the steps of: calculating a major orientation direction expansion amount and an orthogonal direction expansion amount for each of the C channel, M channel, Y channel, and K channel based on the first dot expansion curve and the second dot expansion curve, and calculating an equivalent elliptical area increase amount or an equivalent diameter increase amount for each channel from the major orientation direction expansion amount and the orthogonal direction expansion amount; and normalizing the equivalent elliptical area increase amount or the equivalent diameter increase amount for each channel with respect to the reference dot area or reference dot diameter for each channel, and performing weighted correction according to the reference color load index and the reference dot gain index to calculate a dot expansion rate for each of the C channel, M channel, Y channel, and K channel.

[0012] A device according to one embodiment may be combined with hardware and controlled by a computer program stored on a medium to execute the method of any one of the methods described above. Effects of the invention

[0013] According to one embodiment, by introducing an artificial intelligence color correction model, consistent high-quality corrected images can be automatically obtained independently of operator skill level or device variation, and the efficiency of the entire publishing process can be maximized by performing preprocessing suitable for print media in batches prior to color correction.

[0014] In addition, by subdividing the target image into semantic areas such as the person's skin, sky, and shadows, and by fusing the newspaper's edition, page, and date setting information, it is possible to achieve ultra-precise customized color balance and gradation correction that simultaneously considers the physical characteristics of the page and the attributes of the subject, going beyond simple global correction.

[0015] Furthermore, by tracking the history information of paper rolls to identify microscopic material characteristics such as fiber orientation and surface porosity, and by quantifying the moisture equilibrium state at the time of printing, it provides the effect of perfectly constructing and pre-simulating the physical behavioral environment of ink occurring at the actual printing site using digital data.

[0016] In addition, by normalizing the collected material and environmental data to generate a set of time response functions representing capillary penetration and diffusion decay over time, analog fluid dynamic phenomena such as anisotropic ink bleeding or saturation critical points can be replaced with quantitative indicators that the system can pre-control.

[0017] Furthermore, by precisely calculating and correcting the ink application amount and halftone expansion rate for each channel through time-dependent penetration / residue curves and elliptical geometric models, it fundamentally prevents ink interference (trapping failure) and back-smudging during multi-color printing and optimizes the usage of colored ink to derive an eco-friendly printing control effect that reduces costs. Brief explanation of the drawing

[0018] FIG. 1 is a schematic diagram showing the overall processing flow of an artificial intelligence-based photo color correction system for printing according to an embodiment of the present invention. FIG. 2 is a flowchart showing the overall processing procedure of an artificial intelligence-based photo color correction method for printing according to an embodiment of the present invention. FIG. 3 is a flowchart showing a detailed processing procedure for the step of generating a corrected photo file according to an embodiment of the present invention. FIG. 4 is a flowchart showing a detailed processing procedure for the step of calculating predicted values ​​for paper material characteristics and ink absorption rate according to an embodiment of the present invention. Specific details for implementing the invention

[0019] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, various modifications may be made to the embodiments, and thus the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, and substitutions to the embodiments are included within the scope of the rights.

[0020] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, the embodiments are not limited to the specific disclosed forms, and the scope of this specification includes modifications, equivalents, or substitutions that fall within the technical concept.

[0021] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0022] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or coupled with that other component, or that there may be other components in between.

[0023] The terms used in the embodiments are for illustrative purposes only and should not be interpreted as intended to be limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0024] In particular, where a 'step' in this specification is described as 'comprising' one or more detailed steps or sub-steps, said 'step' may be interpreted as including its own basic processing step while simultaneously performing the described detailed steps as well.

[0025] For example, if it is stated that 'a step of doing B to A' includes 'a step of doing D to C; a step of doing F to E; and a step of doing H to G,' the 'step of doing B to A' may be interpreted not merely as the basic operation of doing B to A, but as a configuration that performs detailed procedures together, such as a step of doing D to C, a step of doing F to E, and a step of doing H to G.

[0026] Accordingly, the above configuration does not exclude various sub-procedures included within the scope of execution of the corresponding step, and may be included within the scope of the present invention even if other procedures or means performing substantially the same or equivalent functions are substituted.

[0027] Expressions such as 'end part', 'both ends', 'one end', 'other end', and 'side end' of a component can be interpreted as referring to at least / any one of the end parts of that component.

[0028] In the description of the present invention, 'a method in which a device comprises a processor, a memory, a communication module, and a non-transient storage medium, and a program stored in the non-transient storage medium is executed by the processor,' the term 'method' may be interpreted as referring to the program stored in the non-transient storage medium itself or a part of the program.

[0029] The term 'Return' as used in the description of the present invention may refer to a result value being output, returned, or returned from a method, procedure, function, etc. used in a given program language / structure.

[0030] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the embodiments pertain. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application. For example, the term 'artificial intelligence model' may be selected from one or more of known general artificial intelligence models.

[0031] In addition, when describing with reference to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the embodiments, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the embodiments, such detailed description is omitted.

[0032] According to one embodiment, an AI-based photo color correction method for printing is provided, wherein the device includes a processor, memory, a communication module, and a non-transient storage medium, and the method is performed by the processor by executing a program stored in the non-transient storage medium, the method comprising: a step of acquiring an original photo file; a step of generating a preprocessed photo file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black and white conversion on the original photo file; and a step of generating a corrected photo file by using an AI color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set to be suitable for print output.

[0033] The step of acquiring the original photo file is a step in which the device receives a digital image file to be color-corrected from an external input source, or loads an image file stored on an internal or external storage medium and registers it as data to be subsequently processed.

[0034] Here, the device may be implemented as a server, cloud system, desktop PC, smartphone, or a combination thereof that performs computation and data processing according to embodiments of the present invention, and may transmit and receive video data and control signals to and from an external worker terminal, editing terminal, newspaper editing system, photo transmission terminal, network storage, or printing house system through a communication module.

[0035] The step of acquiring the original photo file may be a step of loading initial image data into memory that was generated through a digital camera, smartphone, scanner, tablet, or PC, or transmitted from an external database and cloud.

[0036] Here, the original photo file may refer to photo data in an initial state prior to being optimized for print color standards, and may be a general RGB-based image file stored in one or more of known file formats such as JPEG, PNG, TIF, BMP, RAW format, or PSD. In addition, resolution information, color space information, shooting time information, file identification information, and other metadata may be obtained together with the original photo file.

[0037] The step of acquiring the original photo file may be a step of selectively loading one of a plurality of original photo files, or acquiring a file to be printed from among a plurality of candidate photo files corresponding to the same article or the same page. For example, the device may receive an article identification number, submission time, editing department information, and page information scheduled for use along with the photo file from an external editing terminal through a communication module, and based on this, register the corresponding original photo file as target data for subsequent color correction. At this time, if the original photo file is damaged, the file header is abnormal, the resolution is below a preset minimum threshold value, or it is an unsupported format, an exception handling routine may be performed to generate an error log and request retransmission.

[0038] The step of generating a preprocessed photo file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black and white conversion on the original photo file is a step of primarily processing the original photo file to suit the specifications of the actual print page, editing direction, and subsequent correction purpose prior to full-scale artificial intelligence correction calculation.

[0039] Here, the preprocessed photo file refers to a result file in which at least one preprocessing operation has been applied to the original photo file, and can be used as input data for an artificial intelligence color correction model in a subsequent step. The preprocessed photo file may be maintained in the same file format as the original photo file, or it may be converted into a separate intermediate format for processing to improve the internal processing efficiency of the device.

[0040] The step of generating a preprocessed photo file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black-and-white conversion on the original photo file can enable subsequent AI-based color correction to be performed more stably and purposefully by organizing the shape, composition, visual noise, and color expression method of the original photo file in advance. In particular, by performing appropriate preprocessing in advance depending on whether the surface on which the photo is to be placed is a color printing surface or a black-and-white printing surface, or a surface having a specific number of layers and area ratio, it is possible to prevent a subsequent model from performing calculations on unnecessary areas or mismatched color modes, thereby reducing the computational load of the system and improving processing speed.

[0041] The above-mentioned size adjustment may be a process of changing the width, height, resolution, or pixel density of a photo file to suit the purpose of printing. For example, if the original photo file is excessively large, it may be reduced and resized to match the specified page size, number of columns, or processing performance of the printing device; conversely, for small files, it may be expanded to a print-allowable resolution level through an upscaling model or interpolation algorithm. Additionally, the above-mentioned size adjustment may include a process of readjusting the aspect ratio to match the area ratio of the actual newspaper or magazine in which the target photo will be inserted.

[0042] The above cropping process may refer to removing the surrounding area while leaving only the portion of the entire photo containing meaningful objects or key subjects related to the article. The above cropping process may be performed according to center alignment criteria, object detection criteria, editor-specified coordinate criteria, or page layout criteria, thereby enhancing the communicative power of the key subject and optimizing its visual placement within the page. Additionally, the above mosaic processing may be a process intended to de-identify portrait rights, personal information, specific individuals requiring security, vehicle license plates, trademarks, or sensitive areas, and may be implemented using one or more of the following methods: blurring, pixelation, insertion of masking patterns, or solid color overlay.

[0043] The above color conversion and black-and-white conversion may be processes that modify the basic color channel structure of the image in advance according to the printing method of the surface on which the photograph is to be placed. For example, in the case of a photograph to be placed on a color surface, the original photo file stored in the camera's raw color space or a specific device's proprietary color space may be converted to a standard RGB working color space, or normalized to an intermediate reference color space to facilitate CMYK profile mapping in a subsequent step. On the other hand, in the case of a photograph to be placed on a black-and-white surface, the conversion may not stop at simply removing saturation, but may be converted into a grayscale file by applying luminance-based weighting to ensure visibility during printing while maintaining the separation of tonal values ​​between highlights and shadows. For example, grayscale may be generated by assigning different weights to the red, green, and blue channels, or the image may be converted into a black-and-white file after applying a correction curve to maintain contrast in skin areas or text areas.

[0044] The step of generating a corrected photo file by using an artificial intelligence color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set for print output is a step of inputting the preprocessed photo file into a learned model to produce a result file having a color distribution and gradation structure suitable for a printing environment.

[0045] Here, an artificial intelligence color correction model may refer to a learning-based model capable of generating correction parameters or direct output images by analyzing the color state, brightness distribution, channel balance, and regional visual characteristics of an input image. The artificial intelligence color correction model may be implemented as a convolutional neural network, a vision transformer, a generative neural network, a regression-based deep learning model, or a combination thereof, and may be a model that has been pre-trained using pairs of multiple original images and high-quality print images that have been retouched and optimized for printing by experts as training data.

[0046] The step of generating a corrected photo file by using an artificial intelligence color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set for print output may be a step of restoring a photo with degraded image quality or color distortion caused by lighting conditions to an optimal state and converting it into a format suitable for actual printing standards. The saturation correction may be a process to mitigate the problem of specific colors being expressed as excessively vivid or, conversely, excessively dull, and the brightness correction may be performed by rearranging midtones and highlights to prevent key objects from being obscured, by considering not only the average brightness of the entire photo but also the regional distribution of dark and bright areas. Additionally, the color balance correction may be a process to stabilize the white standard and reduce color deviation by adjusting the relative ratios or complementary relationships of red, green, and blue components.

[0047] The step of generating a corrected photo file by using an artificial intelligence color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set for print output can automatically correct local brightness, saturation, and white balance of photos that are dark or have incorrect color temperature depending on the shooting environment by analyzing them on a pixel-by-pixel or area-by-area basis. For example, the artificial intelligence color correction model can analyze the overall saturation distribution contained in the preprocessed photo file and individually calculate the amount of saturation increase or decrease to ensure visual clarity in sky or vegetation areas while limiting excessive saturation in human skin areas. Additionally, if a photo taken under indoor lighting is skewed toward a yellow tone, the white balance can be corrected by increasing the correction amount for blue tones and decreasing the correction amount for red tones.

[0048] The step of generating a corrected photo file by using an artificial intelligence color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photo file and performing color conversion to correspond to a target color profile set for print output may include the step of converting the output result of the artificial intelligence model into a color space that corresponds to the actual printing device, printing method, and ink characteristics.

[0049] Here, the target color profile refers to a reference color space or color mapping rule set to match the physical characteristics of the printing press and ink from which the final image will be output, and may be, for example, a CMYK profile for newspaper printing, a CMYK profile for general commercial printing, or an in-house standard output profile. Generally, since the range and characteristics of the light-based RGB color space visible on a monitor and the CMYK color space implemented by the combination of actual inks do not match, simple conversion alone may result in the colors of the printed material becoming dull or distorted. Accordingly, in the above step, the AI ​​correction results are converted to match the color gamut and channel structure required by the target color profile, and if necessary, colors outside the color gamut may be compressed into adjacent colors or a rendering intent may be applied to prevent discontinuity in gradation.

[0050] The step of generating a corrected photo file is to save the result, in which the saturation correction, brightness correction, color balance correction, and target color profile corresponding conversion are completed, or to output it in a form that can be delivered to a subsequent printing process.

[0051] Here, a corrected photo file may refer to a final or intermediate result file whose color characteristics have been adjusted to be suitable for print output. The corrected photo file may be a single image file, or it may be generated in multiple versions to allow comparison of before and after differences relative to the original file. For example, the device may store the original photo file, the preprocessed photo file, and the corrected photo file with linked identification numbers to enable an editor to review changes at each stage, and the corrected photo file may record the applied correction history, the target color profile used, the processing time, model version information, and correction intensity information together in the form of metadata.

[0052] The step of generating a corrected photo file can be particularly useful when, for example, the original photo file was taken in a low-light environment, resulting in a strong overall blue tint, clustered shadows, and abnormally dull skin tones. In this case, the device can first acquire the original photo file, crop unnecessary peripheral areas, and adjust the size to match standard printing resolutions to generate a pre-processed photo file. Then, an AI color correction model can generate a corrected photo file by performing saturation correction, brightness correction, and white balance correction on the pre-processed photo file to naturally restore skin tones, recover gradations in shadow areas, and finally perform color conversion to correspond to a target color profile for newspaper printing. Accordingly, it is possible to secure photo quality that is not only visually appealing on screen but also highly effective in conveying the article in the actual print result.

[0053] Additionally, the step of generating the correction photo file comprises: a step of obtaining publication metadata including newspaper edition information, page information, and date setting information associated with the original photo file; a step of determining a target CMYK profile and a total ink amount limit value for each printing condition based on the publication metadata; a step of calculating predicted values ​​for the paper material characteristics and ink absorption rate of the target page based on the publication metadata; a step of calculating correction input features including a histogram per RGB channel, a luminance histogram, a saturation distribution, a color temperature deviation, a local contrast value, an edge sharpness value, and a shake index from the preprocessed photo file; and a step of dividing the preprocessed photo file into at least two of a human skin area, a sky area, a vegetation area, a shadow area, a highlight area, a text area, and a graphic area. The method may include the step of inputting the correction input feature quantity, region segmentation result, target CMYK profile, and total ink amount limit value into an artificial intelligence color correction model to generate a set of color correction parameters including region-specific brightness correction coefficients, saturation correction coefficients, white balance correction coefficients, sharpening intensity, black generation amount, and CMYK channel-specific conversion weights; the step of dynamically adjusting the CMYK channel-specific conversion weights by reflecting the paper material characteristics and ink absorption rate prediction values; and the step of generating a corrected photo file by using the set of color correction parameters to limit the color deviation of the human skin area to within a preset reference range, suppressing gradation loss in dark and bright areas, performing dot gain pre-compensation for the target CMYK profile, and then mapping the preprocessed photo file to the target CMYK profile.

[0054] The step of generating a corrected photo file is not limited to merely applying uniform color correction to a preprocessed photo file, but involves comprehensively reflecting the publishing conditions of the page where the photo file will actually be placed, the color reproduction characteristics of the printing equipment, the physical properties of the paper used, the ink penetration and bleeding characteristics, and the visual structure of the photo itself to produce an output image suitable for the final printing result. In other words, the above step may not be a step of generating an image that looks good on a digital screen, but rather a step of precisely correcting the photo file to ensure color reproduction and gradation preservation expected during actual printing in a newspaper editorial environment and a rotary printing environment.

[0055] The step of obtaining publication metadata including newspaper edition information, page information, and date setting information linked to the original photo file is a step of collecting additional publication-related information indicating under what page conditions the original photo file is scheduled to be used.

[0056] Here, newspaper edition information may refer to information classified according to printing time or publication number, such as morning edition, evening edition, or 1st, 2nd, and 3rd editions; page information may refer to specific page locations where the relevant photo is placed, such as page 1, business page, sports page, or full-page advertisement page; and date setting information may refer to the year, month, and day of the week information for publication. Additionally, publication metadata may refer to a data set that combines the above edition information, page information, and date setting information with article identification numbers, editorial department information, printing house identification information, and regional edition classification information, and may be inserted in the form of meta tags in a file header or managed as a text data set linked to a separate database record.

[0057] The step of obtaining publication metadata, including newspaper edition information, page information, and date setting information associated with the original photo file, may be a preliminary step for setting reference values ​​for a color profile, total ink limit, page-specific editing policy, and printing conditions to be applied in a subsequent step. For example, even if the same photo file is used, the required color reproduction range and allowable ink amount may differ depending on whether the photo is placed on a full-page color advertisement or inserted into a part of a general article page; therefore, the device may first obtain the publication metadata and then perform a subsequent correction procedure based on it. Additionally, if there are missing values ​​or abnormal values ​​in the publication metadata, the device may call default publication conditions according to the basic editing policy or set the photo file to a pending state to request reassignment by the administrator.

[0058] Based on the above-mentioned publishing metadata, the step of determining the target CMYK profile and total ink volume limit for each printing condition is a step of setting color space standards and ink usage limits suitable for the printing environment in which the above-mentioned photo file will be finally output. Here, the target CMYK profile may refer to a standard color profile set to correspond to a specific printing press, ink set, rotary speed, page type, and internal editing policy, and the profile may include output characteristics of each of the C, M, Y, and K channels, gray balance maintenance conditions, black generation rules, allowable total ink volume range, and dot gain correction conditions. Additionally, the total ink volume limit may refer to the maximum allowable sum when the four inks of C, M, Y, and K overlap at the same point or in a local area, and may be used as a standard value to prevent backsmudging, drying delay, smudging, or page blurring caused by excessive ink loading.

[0059] Based on the above-mentioned publishing metadata, the step of determining the target CMYK profile and total ink volume limit for each printing condition may select or generate different sets of references according to plate information, page information, and date setting information. For example, under morning mass printing conditions, the total ink volume limit may be set relatively low considering the ink drying speed and page absorption rate, and for high-quality color special pages, a target profile with an expanded color gamut may be selected. Additionally, since the date setting information may indirectly reflect variations in the season, ambient temperature, relative humidity, and the internal environment of the printing room, the total ink volume limit, black generation ratio, or allowable density range per channel may be further adjusted according to the hot and humid conditions of summer and the cold and dry conditions of winter.

[0060] Based on the above-mentioned publishing metadata, the step of calculating predicted values ​​for the paper material characteristics and ink absorption rate of the target page is a step of predicting the physical characteristics of the paper on which the photo file will actually be printed and the tendency for ink penetration and diffusion on the paper. Here, paper material characteristics are not limited to a simple paper name, but may refer to multiple numerical values ​​or vector-shaped data representing the surface roughness, thickness, presence or absence of a coating layer, surface porosity, fiber orientation, surface roughness, degree of whiteness reduction, moisture status, and ink acceptance tendency of the substrate layer. Additionally, the predicted value for ink absorption rate may be a numerical value representing the degree to which ink penetrates into the substrate layer or remains on the surface at the level of the C channel, M channel, Y channel, and K channel, or at the entire page level, and may be reflected in the subsequent step of adjusting the conversion weights for each channel.

[0061] The step of calculating predicted values ​​for the paper material characteristics and ink absorption rate of the target page based on the above-mentioned publishing metadata may be a step intended to reflect the fact that the output color may vary depending on the actual paper characteristics even if the same target CMYK profile is applied. For example, even within the same newspaper or magazine company, the type of paper or coating used for page 1 and page 20 may differ, and the speed at which ink penetrates the paper and the degree of bleeding may vary depending on the seasonal and humidity conditions corresponding to the date setting information. Accordingly, since paper with a high surface porosity may have ink penetrate quickly and the color density may be relatively low, and paper with a large variation in coating layer thickness may experience increased spreading of specific channels or dot expansion, the above step may function as a step of providing foundational data that enables a subsequent artificial intelligence color correction model to generate correction parameters including predictions of the actual printing results.

[0062] The step of calculating correction input features including histograms by RGB channels, luminance histograms, saturation distribution, color temperature deviation, local contrast values, edge sharpness values, and shake index from the above-mentioned preprocessed photo file is a step for quantifying the visual state of the above-mentioned preprocessed photo file and providing it as input features for an artificial intelligence color correction model.

[0063] Here, the RGB channel histogram may be data representing the distribution of pixel values ​​for each of the red, green, and blue channels, and the luminance histogram may be a histogram representing the brightness distribution of the entire image. The saturation distribution may refer to the distribution of color intensity across the entire photograph or by region, and the color temperature deviation may be a value representing the degree of deviation from the reference white balance. Additionally, the local contrast value may be a value quantifying the difference in brightness between adjacent pixels or local blocks, the edge sharpness value may be a value indicating how clearly the outline boundaries of an object are maintained, and the shake index may be an indicator quantifying the degree of blur caused by camera shake or subject shake.

[0064] The step of calculating correction input features, including histograms by RGB channels, luminance histograms, saturation distribution, color temperature deviation, local contrast values, edge sharpness values, and shake index, from the preprocessed photo file may be a step of subdividing and identifying the type of problem possessed by the photo file. For example, if the luminance histogram is excessively concentrated in the low-luminance range, it may be determined to be a low-light photograph; if the saturation distribution and color temperature deviation are significantly skewed in a specific direction, it may be determined that the entire image is biased toward red or blue light. Additionally, if the local contrast value is low and the edge sharpness value is low, it may be determined that the subject is rendered blurry; and if the shake index is above a reference value, it may be used as a basis for separately increasing the sharpening intensity or the amount of local correction. As such, the above step may be a step of composing correction input features in the form of a multidimensional vector by synthesizing a number of quantitative values.

[0065] The step of dividing the above-mentioned preprocessed photo file into at least two of a person skin area, a sky area, a vegetation area, a shadow area, a light area, a text area, and a graphic area is a step of separating visual objects or gradation areas included in the photo file by separating them by semantic unit.

[0066] Here, the human skin area may refer to an area containing a person's face, neck, arms, or other exposed skin, and the sky area may refer to an upper background area containing a clear sky or clouds. The vegetation area may correspond to an area containing trees, grass, forests, or other green natural elements, and the shadow area and light area may refer to areas with relatively low and high luminance, respectively. Additionally, the text area may be an area containing article captions, guide text, signage characters or numbers, etc., and the graphic area may be an area containing charts, icons, logos, illustrations, or artificially synthesized visual elements.

[0067] The step of dividing the above-mentioned preprocessed photo file into at least two of the human skin area, sky area, vegetation area, shadow area, highlight area, text area, and graphic area may be a step intended to prevent unnatural results that may occur when the same amount of correction is applied uniformly to areas having different visual attributes. For example, both the skin area and the sky area may require saturation correction, but the skin area may prioritize maintaining a natural complexion, while the sky area may prioritize maintaining sharpness in blue tones. Additionally, for the text area, securing clear contrast may be more important than excessive sharpening for readability, and the direction of tonal restoration may differ between the shadow area and the highlight area. Therefore, the area division step may generate pixel-unit or block-unit masks through an artificial intelligence segmentation network, etc., and be utilized as reference information to enable area-specific customized correction in the subsequent parameter generation step.

[0068] The step of inputting the correction input feature quantity, region segmentation result, target CMYK profile, and total ink amount limit value into an artificial intelligence color correction model to generate a color correction parameter set including region-specific brightness correction coefficients, saturation correction coefficients, white balance correction coefficients, sharpening intensity, black generation amount, and CMYK channel-specific conversion weights is a step of deriving a correction operation rule that simultaneously reflects the current state of the input image and actual printing conditions. Here, the color correction parameter set may refer to a set of multiple correction coefficients, weights, and control values ​​to be used in the subsequent color correction and color space mapping process. The brightness correction coefficient may be a value that controls the increase or decrease in brightness of each region or the entire image, the saturation correction coefficient may be a value that increases or decreases color vividness, and the white balance correction coefficient may be a channel-specific reference correction amount for correcting color temperature deviation. Additionally, sharpening intensity may be a value for controlling the degree of edge emphasis, black generation amount may be a value for adjusting the sharing ratio of CMY channels and K channels, and CMYK channel-specific transformation weights may be values ​​for adjusting the channel-specific mapping intensity corresponding to the target CMYK profile.

[0069] The step of inputting the correction input feature quantity, region segmentation result, target CMYK profile, and total ink amount limit value into an artificial intelligence color correction model to generate a set of color correction parameters including region-specific brightness correction coefficients, saturation correction coefficients, white balance correction coefficients, sharpening intensity, black generation amount, and CMYK channel-specific conversion weights may be a step that enables the artificial intelligence model to perform correction decisions that consider reproducibility and print stability in actual printed materials, rather than aiming solely for image quality improvement. For example, a relatively high sharpening intensity may be assigned to regions with a high shake index, and on surfaces with a low total ink amount limit value, a black generation amount may be calculated by increasing the utilization ratio of the achromatic K channel instead of limiting the increments of the C and M channels for highly saturated red regions. Additionally, if gradation loss in dark areas is expected, the brightness correction coefficient for midtone repositioning and the black generation amount may be corrected simultaneously.

[0070] The step of dynamically adjusting the CMYK channel-specific conversion weights by reflecting the paper material characteristics and ink absorption rate prediction values ​​is a step of recalibrating the initial channel-specific conversion weights generated by the artificial intelligence color correction model to match actual paper conditions and ink behavior conditions. Here, dynamic adjustment does not mean uniform adjustment based on a fixed lookup table, but rather means variably changing the weights of the C channel, M channel, Y channel, and K channel according to the paper material characteristics corresponding to the paper and the channel-specific or integrated ink absorption rate prediction values. For example, if the ink absorption rate prediction value is abnormally high or the material is identified as having a high possibility of bleeding, such as newsprint paper, the channel-specific weights can be dynamically updated by attenuating the conversion weights of the C channel or M channel, which are the cause of bleeding, by a predetermined ratio relative to the normal value, and amplifying the conversion weights of the K channel, which ensures contour and brightness stability.

[0071] The step of dynamically adjusting conversion weights for each CMYK channel by reflecting the predicted values ​​of the paper material characteristics and ink absorption rate may be a compensation step to minimize differences in output results on actual paper even if the target CMYK profile is the same. In particular, in paper with high absorption rates, ink may penetrate rapidly into the substrate layer, which can lower the surface color density, and since the spreading characteristics for each channel may vary depending on the coating layer thickness variation or porosity difference, the above step can perform pre-prediction-based print quality compensation by reflecting a correction amount for each channel corresponding to the predicted values ​​of the paper material characteristics and ink absorption rate.

[0072] The step of generating a corrected photo file by using the above color correction parameter set to limit the color deviation of the human skin area to within a preset reference range, suppressing the loss of gradation in the dark and bright areas, performing dot gain pre-compensation for the target CMYK profile, and then mapping the preprocessed photo file to the target CMYK profile is a step of performing final area-specific color correction and color space conversion for printing on the preprocessed photo file.

[0073] Here, limiting the color deviation of the human skin area to within a reference range may mean performing color conversion within a pre-set color difference range so that the skin color does not become excessively red or skew toward an abnormal blue tone or grayish-green. Additionally, suppressing the loss of gradation in the shadow and highlight areas may mean maintaining midtones and boundary gradations by readjusting the luminance distribution and tone curve so that dark areas are not blurred or bright areas are not completely blown out.

[0074] The step of generating a corrected photo file by mapping a preprocessed photo file to a target CMYK profile after limiting the color deviation of the human skin area to within a preset reference range using the above set of color correction parameters, suppressing tonal loss in the shadow and highlight areas, and performing dot gain pre-compensation for the target CMYK profile, may be a step of performing final color space mapping by taking into account in advance the dot expansion phenomenon expected during the actual rotary printing process. Here, dot gain pre-compensation may refer to a process of adjusting the density or occupied area per channel in advance at the digital file stage in anticipation of the phenomenon where ink dots spread on the paper due to the pressure of the rotary press and the diffusion characteristics of the ink during printing, resulting in them being printed larger and darker than their original size. For example, if it is predicted that a large K-channel dot gain will occur under specific surface conditions, compensation may be performed by partially lowering the K-channel weight of the preprocessed photo file and rearranging the combination of CMY channels.

[0075] The step of generating a corrected photo file by using the above set of color correction parameters to limit the color deviation of the human skin area to within a preset reference range, suppressing the loss of gradation in the dark and bright areas, performing dot gain pre-compensation for the target CMYK profile, and then mapping the preprocessed photo file to the target CMYK profile may be a step of generating an output image file that can ultimately be delivered to a printing device. The corrected photo file may be a single image file, or it may be a multiple-version structure in which the original file, the preprocessed file, and the corrected result file are stored together with identification information linked to each other. In addition, the applied target CMYK profile, total ink amount limit value, channel-specific conversion weight adjustment history, skin area color difference control information, dot gain compensation history, and processing time may be recorded together in the corrected photo file in the form of metadata.

[0076] For example, it can be assumed that the above-mentioned preprocessed photo file is a night sports scene in which the overall blue tone is strong due to stadium lighting, the dark areas in the spectator seating area are clustered, and the skin color of the players is expressed somewhat dull with gray tones. In this case, the device can first obtain publication metadata including plate information, surface information, and date setting information associated with the photo, and determine a target CMYK profile and a total ink amount limit value suitable for the rotary printing conditions of the sports page. Subsequently, it can calculate paper material characteristics and predicted ink absorption rates, extract correction input features from the preprocessed photo file, and distinguish the player skin area, vegetation area corresponding to grass, lighting background area, dark area, and character area. Next, the artificial intelligence color correction model can generate a set of color correction parameters including region-specific brightness correction coefficients, saturation correction coefficients, white balance correction coefficients, and channel-specific conversion weights based on the above input values, and then dynamically adjust them by reflecting the above paper material characteristics and ink absorption rate prediction values, and generate a corrected photo file by mapping to a target CMYK profile while limiting color deviation in the skin area, restoring shadow gradations, and performing dot gain pre-compensation. Accordingly, the corrected photo file can secure output quality in which not only is the image aesthetically pleasing on the screen, but the player's skin color, grass color, uniform color, and background gradations are stably reproduced on the actual printed surface.

[0077] And, the step of calculating the predicted values ​​for the paper material characteristics and ink absorption rate comprises: a step of querying paper history information including the lot number, paper supplier, paper basis weight, coating type, calendaring processing history, manufacturing date, and storage period of the paper roll to be used for printing the target page, based on the newspaper company plate information, page information, and date setting information included in the publishing metadata; a step of generating a basic material characteristic vector including the fiber orientation, surface porosity, coating layer thickness deviation, surface roughness, and degree of whiteness degradation of the target paper, based on the paper history information; a step of generating an environmental equilibration vector including the ambient temperature immediately before printing, relative humidity, average humidity of the paper storage space, elapsed time since opening the paper, and the printing press preheating status, based on the date information and printing time information; and a step of querying the reference ink coverage rate, reference color load index, and reference dot gain index corresponding to the target page from a previously stored print history database by page type, based on the plate information and page information. A step of calculating a set of absorption behavior parameters including an initial penetration velocity of the paper surface, a time-delayed diffusion coefficient, a transverse blurring coefficient, an anisotropic absorption coefficient according to the winding direction, and a saturation critical absorption amount, using the above-mentioned basic material characteristic vector, moisture equilibrium vector, reference ink coverage rate, reference color load index, and reference dot gain index as inputs; and a step of calculating a first-order absorption rate, a stabilized absorption rate, and a dot expansion rate for each CMYK channel based on the above-mentioned set of absorption behavior parameters.The method may include: a step of determining the set of fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the basic material characteristic vector, and the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount as the paper material characteristics of the target surface based on the stabilization absorption rate and dot expansion rate for each CMYK channel; a step of calculating the ink absorption contribution for each channel by weightedly combining the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the basic material characteristic vector, the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount, and the stabilization absorption rate and dot expansion rate for each CMYK channel by channel; and a step of calculating the predicted ink absorption rate for each channel for the C channel, M channel, Y channel, and K channel, and the predicted integrated ink absorption rate for the entire target surface based on the ink absorption contribution for each channel.

[0078] The step of querying paper history information, including the lot number, paper supplier, paper basis weight, coating type, calendaring processing history, manufacturing date, and storage period of the paper roll to be used for printing the target page, based on the newspaper edition information, page information, and date setting information included in the above-mentioned publishing metadata, is a step of specifying the source and production and storage history of the physical medium on which the photograph will finally be printed.

[0079] Here, a paper roll may refer to a roll of paper loaded into a rotary press for mass printing of newspapers or magazines, and lot number and paper supplier may refer to information identifying which production batch the paper was manufactured in and through which manufacturer or supply chain it was imported. Additionally, paper basis weight refers to the weight of the paper per unit area and can serve as a standard for inferring thickness and density, and calendering history may refer to whether a process was performed and the intensity of a process in which the paper surface is pressed with rollers to adjust smoothness and gloss.

[0080] Based on the newspaper company edition information, page information, and date setting information included in the above-mentioned publishing metadata, the step of querying paper history information, including the lot number, paper supplier, paper basis weight, coating type, calendaring processing history, manufacturing date, and storage period of the paper roll to be used for printing the target page, may be a preliminary process for quantifying paper material deviation in a subsequent step. For example, since the paper used for the sports page of the morning edition 2nd edition and the paper used for the advertising page of the evening edition may have different basis weights or coating conditions even within the same newspaper company, and the whiteness and moisture content tendencies may vary depending on the manufacturing date and storage period, the device may identify the paper roll most likely to be fed into the target page by querying the paper feeding records of the rotary line, the material management database, or the warehouse management system. In addition, if multiple lot numbers are mixed on the same date, a representative roll may be selected by comparing the feeding history by page or the distribution history by copy, or the average value of the physical properties of multiple candidate rolls may be set as the representative value.

[0081] The step of generating a basic material characteristic vector including fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation of the target paper based on the above paper history information is a step of converting the retrieved paper history information into a numerical characteristic set suitable for print response analysis.

[0082] Here, a basic material characteristic vector may refer to data composed of multiple basic characteristic values ​​representing the material tendencies of the target paper in the form of a single vector. Fiber orientation may be a value representing the direction in which internal pulp fibers are predominantly arranged, i.e., the paper grain, and surface porosity may be a value representing the ratio of fine pores or the pore distribution density of the paper surface or near-surface layer. Additionally, coating layer thickness variation may be a value representing the degree of uniformity across the entire surface rather than the absolute thickness of the coating layer, surface roughness may be a value quantifying fine roughness, and the degree of whiteness degradation may refer to the amount of reduction in the current optical whiteness of the paper compared to the manufacturing standard whiteness or initial whiteness.

[0083] Based on the above paper history information, the step of generating a basic material property vector including the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation of the target paper can be performed by referring to physical property tables provided by the manufacturer, quality control logs, historical measurements, or an in-house paper specification database. For example, once the paper basis weight and supplier are specified, the device can retrieve the average surface roughness and porosity ranges by querying the standard physical property table of the supplier, and can estimate the degree of surface consolidation and coating uniformity based on the coating type and calendering history. Additionally, the degree of whiteness degradation may be assigned relatively higher as the manufacturing date and storage period increase, or additional correction values ​​related to surface porosity or roughness may be reflected if there is a history of long-term storage in a high-temperature and high-humidity environment.

[0084] Based on the above date information and printing time information, the step of generating a moisture equilibrium vector including the ambient temperature immediately before printing, relative humidity, average humidity of the paper storage space, elapsed time since opening the paper, and the printing press preheating state is a step of generating an environmental characteristic vector indicating to what extent the paper has reached a moisture and thermal equilibrium state in the environment immediately before printing.

[0085] Here, the moisture equilibrium vector may refer to a set of multiple state values ​​that quantitatively represent the current moisture content trend of the target paper and the thermal state immediately prior to printing. Ambient temperature and relative humidity may be real-time sensor values ​​inside the rotary press or printing house, and the average humidity of the paper storage space may be calculated based on sensor records installed in the warehouse, feed waiting room, or storage rack. Additionally, the elapsed time since opening the paper may refer to the time elapsed from when the wrapped paper roll is opened until the point of printing, and the printing press preheating status may be a state value corresponding to roller temperature, drive time, or whether warming up is complete.

[0086] Based on the above date information and printing time information, the step of generating a moisture equilibrium vector including the ambient temperature immediately before printing, relative humidity, average humidity of the paper storage space, elapsed time since opening the paper, and the printing press preheating status may be a step intended to reflect the influence of seasonal and real-time operating environments on the interaction between paper and ink. For example, in a high-humidity environment during the summer, the likelihood of moisture content in the paper surface and substrate layer increases, which may alter the initial penetration pattern and surface residue amount of ink; and if the printing press is not sufficiently preheated, the viscosity and drying speed of the ink may fluctuate. Therefore, the device may compose the above environmental values ​​into a single moisture equilibrium vector and reflect them in the subsequent analysis of absorption behavior.

[0087] Based on the above plate information and surface information, the step of querying the reference ink coverage rate, reference color load index, and reference dot gain index corresponding to the target surface from the previously stored print history database by surface type is a step of retrieving the print load reference values ​​typically used in the corresponding surface type.

[0088] Here, the reference ink coverage rate may refer to the ratio of the average ink coverage area used in a specific surface type or the level of area coverage per channel, the reference color load index may be a numerical value representing the color density burden or color concentration of the entire surface, and the reference dot gain index may be a reference value for the statistically observed dot expansion trend in the same surface type. For example, the sports surface may have a high color load index because it includes large-area color photos, uniform colors, and advertising elements, while the general political surface may have a relatively low reference ink coverage rate because it is a text-centered surface.

[0089] Based on the above plate information and surface information, the step of querying the reference ink coverage rate, reference color load index, and reference dot gain index corresponding to the target surface from the previously stored print history database by surface type may be a step for securing a relative standard by comparing it with the print history of the same type of surface in the past, rather than determining the correction amount of the current photo file alone. For example, if a number of past print results on a specific advertising surface repeatedly showed an increase in dot gain near the total ink amount threshold, the device may induce conservative color correction by loading the reference dot gain index corresponding to the advertising surface type higher than that of a general article surface.

[0090] The step of calculating a set of absorption behavior parameters including an initial penetration rate of the paper surface, a time-delayed diffusion coefficient, a transverse diffusing coefficient, an anisotropic absorption coefficient according to the winding direction, and a saturation critical absorption amount by using the above-mentioned basic material characteristic vector, moisture equilibrium vector, reference ink coverage rate, reference color load index, and reference dot gain index as inputs is a step of generating a set of dynamic parameters indicating how the ink penetrates, diffuses, and remains in the target paper according to time and direction.

[0091] Here, the initial penetration rate may refer to the speed at which ink begins to penetrate into the substrate layer immediately after being applied to the surface of the paper, and the time-delayed diffusion coefficient may be a value that quantifies the characteristic of the amount of ink diffusion changing over time. Additionally, the lateral spreading coefficient may be a value indicating the degree to which ink spreads laterally along the surface, and the anisotropic absorption coefficient may be a value reflecting the degree to which different penetration behaviors appear depending on the winding direction or fiber orientation direction, and the saturation critical absorption amount may refer to a reference amount at which the paper surface and near-surface layer can no longer effectively accommodate ink.

[0092] The step of calculating a set of absorption behavior parameters, including the initial penetration rate of the paper surface, time-delayed diffusion coefficient, transverse blurring coefficient, anisotropic absorption coefficient according to the winding direction, and saturation critical absorption amount, using the above-mentioned basic material characteristic vector, moisture equilibrium vector, reference ink coverage rate, reference color load index, and reference dot gain index as inputs, can be performed through rule-based computation, a regression model, a physical simulation model, or a combination thereof. For example, if the surface porosity is high and the paper moisture state in the moisture equilibrium vector is low, the initial penetration rate may be set high, and if the coating layer thickness variation is large, the transverse blurring coefficient or the anisotropic absorption coefficient may be calculated in a direction that increases. Additionally, a paper with a high reference color load index may be modeled as a condition that reaches the saturation critical absorption amount more quickly because the ink supply amount per unit area is relatively large.

[0093] Based on the above set of absorption behavior parameters, the step of calculating the first-order absorption rate, the stabilization absorption rate, and the dot expansion rate for each CMYK channel is a step of quantifying, by channel, how the inks of the C channel, M channel, Y channel, and K channel in the paper remain on the surface, penetrate into the substrate layer, and change the halftone shape over time.

[0094] Here, the first absorption rate may refer to the penetration rate per channel that appears in the initial section immediately after ink application, the stabilization absorption rate may refer to the final penetration rate per channel when the effective penetration of the ink reaches a stable state after a certain period of time has elapsed, and the dot expansion rate may be a value indicating how much the halftone dots of each channel are enlarged and reproduced from their original digital size due to surface diffusion and pressure influence during the printing process.

[0095] Based on the above set of absorption behavior parameters, the step of calculating the first-order absorption rate, stabilization absorption rate, and dot expansion rate for each CMYK channel may be a step of deriving independent predicted values ​​by reflecting the fact that the pigment characteristics, viscosity, and visual influence of the ink for each channel are different. For example, even under the same paper conditions, the Y channel may show a sensitive decrease in visual density, and the K channel may show a significant dot gain influence directly related to contour sharpness; therefore, the device may calculate the first-order absorption rate, stabilization absorption rate, and dot expansion rate for each channel separately based on the set of absorption behavior parameters.

[0096] Based on the above-mentioned stabilized absorption rate and dot expansion rate for each CMYK channel, the step of determining the set of fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the above-mentioned basic material characteristic vector, and the above-mentioned initial penetration rate, time-delayed diffusion coefficient, transverse blurring coefficient, anisotropic absorption coefficient, and saturation critical absorption amount as the paper material characteristics of the target surface is a step of determining the final set of surface-corresponding material characteristics by integrating the basic material characteristic values ​​and absorption parameters reflecting actual printing behavior. That is, rather than defining paper characteristics solely based on paper information in simple manufacturing specifications, it may be a step of combining and fixing a unique material profile corresponding to the corresponding printing unit by considering the environment immediately before actual printing and the printing behavior for each channel together. Accordingly, even if paper has the same paper model name, different paper material characteristics may be determined if storage conditions or the working environment on the day differ.

[0097] The step of calculating the ink absorption contribution per channel by weightedly combining the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness reduction included in the basic material characteristic vector, the initial penetration rate, time-delayed diffusion coefficient, transverse spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount, and the stabilization absorption rate and dot expansion rate per CMYK channel, is a step of calculating the contribution indicating how much the paper affects each channel by quantifying the correlation between the paper material characteristics and the printing response per channel.

[0098] Here, the ink absorption contribution per channel may refer to the relative influence that the physical material and absorption behavior of the paper contribute to the final color reproduction of a specific channel. For example, the absorption contribution of a channel can be derived by summing the value obtained by multiplying the fiber orientation value by a first weight, the value obtained by multiplying the initial penetration rate by a second weight, and the value obtained by multiplying the stabilization absorption rate of a specific channel by a third weight.

[0099] The step of calculating the ink absorption contribution per channel by weighting and combining the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the above basic material characteristic vector, the above initial penetration rate, time-delayed diffusion coefficient, transverse blurring coefficient, anisotropic absorption coefficient, and saturation critical absorption amount, and the above CMYK channel-specific stabilization absorption rate and dot expansion rate, may be a step of generating a basis value for determining in which direction the channel-specific conversion weights will be adjusted in a subsequent step. For example, if both the surface porosity and the initial penetration rate are high and the stabilization absorption rate of the Y channel is high, the absorption contribution of the Y channel may be calculated to be relatively large, and if the dot expansion rate of the K channel is large, the surface blurring contribution of the K channel may be reflected as high. In addition, if the absorption contribution of the M channel is high and the dot expansion rate is also high, a correction in the direction of subtracting the M channel conversion weight may be induced, and conversely, if the surface retention of the C channel is low and a decrease in color density is expected, a correction in the direction of increasing the C channel conversion weight may be possible.

[0100] Based on the above-mentioned ink absorption contribution per channel, the step of calculating the predicted ink absorption rate per channel for the C channel, M channel, Y channel, and K channel, and the predicted integrated ink absorption rate for the entire target surface, is a step of generating an output value that finally summarizes the paper material characteristics and the reaction characteristics per channel.

[0101] Here, the predicted ink absorption rate per channel may be a final predicted value reflecting the effective penetration and surface retention rates expected to actually be seen on the target surface of each channel ink, and the integrated ink absorption rate predicted value may be a representative value calculated by summing, averaging, or normalizing and integrating the above-mentioned predicted values ​​per channel based on the entire surface. The integrated ink absorption rate predicted value may be used as a reference value to determine the print correction intensity, total ink amount correction amount, or the direction of increase or decrease per channel at the level of the entire photo file.

[0102] Based on the ink absorption contribution per channel, the step of calculating the predicted ink absorption rate per channel for the C, M, Y, and K channels and the predicted integrated ink absorption rate for the entire target surface may be a step of generating key intermediate output values ​​to simultaneously satisfy printability and color reproduction accuracy throughout the entire step of generating the correction photo file. For example, if the predicted integrated ink absorption rate is calculated to be high on a specific surface, it is highly likely that ink penetration will occur rapidly throughout and saturation will decrease, so the channel-specific conversion weight or black generation amount can be corrected more aggressively in a subsequent step. Conversely, if the predicted integrated ink absorption rate is low and the dot expansion rate is high, it is determined that there is a high possibility of turbidity due to surface residue and smudging, so the channel density can be adjusted more conservatively.

[0103] For example, it can be assumed that the target page is a summer evening sports page, and that large-area color photos and advertisements are placed together according to page information. In this case, the device can first look up the lot number, basis weight, coating type, and storage period of the paper roll to be used based on the corresponding plate information, page information, and date setting information. Then, if the paper is determined from the paper history information to have a relatively large variation in surface porosity and coating layer thickness, and if the relative humidity of the rotary chamber is high immediately before printing, causing the moisture equilibrium vector to form in a high-humidity state, the device can calculate both the initial penetration rate and the transverse bleed coefficient to be high. Subsequently, if the dot expansion rate of the C channel and M channel exceeds a reference value and the stabilization absorption rate of the Y channel appears high, the paper material characteristics are determined, the ink absorption contribution per channel is calculated, and based on this, predicted ink absorption rates per channel can be derived, with the Y channel in the direction of concentration supplementation and the C channel and M channel in the direction of bleed suppression. Accordingly, an integrated ink absorption rate prediction value for the entire target surface is generated, and a subsequent color correction step can compensate in advance for color bleeding, skin tone dullness, and background saturation reduction in actual sports surface printing by reflecting the prediction value.

[0104] In addition, the step of calculating the above absorption behavior parameter set comprises: generating a material normalized feature vector by converting the fiber orientation, surface porosity, coating layer thickness deviation, surface roughness, and degree of whiteness degradation included in the above basic material characteristic vector into a preset normalization range; generating moisture-related state values ​​by normalizing the ambient temperature, relative humidity, average humidity of the paper storage space, and elapsed time after opening the paper included in the above moisture equilibrium vector, and calculating a paper moisture activity index by weighted combining the moisture-related state values; generating heat-related state values ​​by normalizing the ambient temperature, relative humidity, printing press preheating state, and elapsed time after opening the paper included in the above moisture equilibrium vector, and calculating a thermal stabilization index by weighted combining the heat-related state values; and calculating a directional penetration coefficient for the main orientation direction of the paper and a direction orthogonal to the main orientation direction, respectively, based on the fiber orientation and surface porosity included in the above material normalized feature vector, and generating a biaxial transmission matrix from the directional penetration coefficient. A step of calculating a coating delay coefficient indicating the degree of delay in the initiation of ink penetration into the substrate layer based on the coating layer thickness deviation and surface porosity included in the material normalization feature vector and the paper moisture activity index; a step of calculating a surface retention coefficient indicating the degree of ink retention on the surface based on the surface roughness included in the material normalization feature vector and the paper moisture activity index; a step of calculating a moisture swelling coefficient indicating the degree of wet swelling of the paper surface based on the paper moisture activity index, thermal stabilization index, and the surface porosity and coating layer thickness deviation included in the material normalization feature vector; a step of calculating a saturation limit correction coefficient indicating the amount of variation in the ink acceptance limit based on the moisture swelling coefficient, thermal stabilization index, and the degree of whiteness reduction included in the material normalization feature vector;A step of generating a reference load vector representing the reference printing load state of a target surface using the reference ink coverage rate, reference color load index, and reference dot gain index; a step of calculating a directional load response vector by combining the reference load vector and a biaxial transmission matrix; a step of inputting the directional load response vector, coating lag coefficient, surface residence coefficient, moisture swelling coefficient, and saturation limit correction coefficient into a time response function generation model to calculate the capillary penetration gradient immediately after ink application, the diffusion decay gradient over time, the lateral blur sensitivity, and the critical time to reach saturation, and generating a set of time response functions based on the capillary penetration gradient, diffusion decay gradient, lateral blur sensitivity, and critical time to reach saturation; and a step of calculating, respectively, the initial penetration velocity at a first reference time point, the time-delayed diffusion coefficient, the lateral blur coefficient, the anisotropic absorption coefficient, and the critical absorption amount for saturation at a second reference time point. and may include the step of generating the set of absorption behavior parameters by combining the initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount.;

[0105] The step of generating a material normalization feature vector by converting the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the above basic material characteristic vector into a preset normalization range, is a step of unifying basic material-related numerical values ​​having different units and ranges into a common scale suitable for subsequent calculations.

[0106] Here, a material normalized feature vector may refer to vector data that expresses the basic material characteristics of the target paper as a set of normalized numerical values. For example, fiber orientation can be measured as a directional value in an angular range, surface porosity can be measured as a percentage or fractional value, and coating layer thickness deviation can be expressed in length units; therefore, the device may not use these values ​​as they are but convert them into normalized values ​​between zero and one or between minus one and one according to a preset minimum and maximum value range.

[0107] The step of generating a material normalized feature vector by converting the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the above basic material characteristic vector into a preset normalization range, respectively, may be a step intended to prevent subsequent models or mathematical calculation results from being distorted due to absolute unit differences of specific characteristic values. For example, if the surface roughness value is recorded in a relatively large numerical range and the degree of whiteness degradation is recorded in decimal units, the surface roughness value may have an excessive influence compared to other characteristic values ​​if the normalization process is not performed. Accordingly, the device generates a material normalized feature vector by applying a predefined normalization range to each characteristic value and can utilize the material normalized feature vector as a common input value for the subsequent calculation of the directional penetration coefficient, coating retardation coefficient, and surface retention coefficient.

[0108] The step of generating moisture-related state values ​​by normalizing the ambient temperature, relative humidity, average humidity of the paper storage space, and the elapsed time after opening the paper, respectively, included in the moisture equilibrium vector, and calculating the paper moisture activity index by weighted combination of the moisture-related state values, is a step of expressing the potential for the current paper to absorb or release moisture as a single representative indicator.

[0109] Here, the paper moisture activity index refers to a normalized index value indicating how much the paper surface and substrate layer are affected by moisture from the environment immediately prior to printing. For example, if the ambient temperature and relative humidity are high, the average humidity of the paper storage space is also high, and a sufficiently long time has elapsed since opening the paper, it is determined that the paper has reached moisture equilibrium with the outside air, allowing the paper moisture activity index to be calculated as relatively high.

[0110] The step of generating moisture-related state values ​​by normalizing the ambient temperature, relative humidity, average humidity of the paper storage space, and the elapsed time after opening the paper, each included in the moisture equilibrium vector, and calculating the paper moisture activity index by weightedly combining the moisture-related state values, may be a step of interpreting the comprehensive wet state by assigning weights proportional to the influence on ink diffusion and penetration to each state value, rather than simply listing individual environmental values ​​such as temperature or humidity. Accordingly, the paper moisture activity index may be used as a reference value for calculating the coating delay coefficient, surface residence coefficient, and moisture swelling coefficient in subsequent steps.

[0111] The step of generating heat-related state values ​​by normalizing the ambient temperature, relative humidity, printing press preheating state, and elapsed time after opening the paper, respectively, included in the moisture equilibrium vector, and calculating a thermal stabilization index by weightedly combining the heat-related state values ​​is a step of summarizing the influence of the thermal environment immediately before printing on the viscosity, fluidity, and drying conditions of the ink into a single indicator.

[0112] Here, the thermal stabilization index may be a value indicating the extent to which the paper and printing equipment have currently reached a thermally stable state. For example, if the printing press is sufficiently preheated, the ambient temperature is close to the reference range, and a certain amount of time has elapsed since the paper was opened, the thermal stabilization index can be calculated as high, allowing for a more conservative interpretation of diffusion attenuation trends or drying stability in subsequent stages.

[0113] The step of generating heat-related state values ​​by normalizing the ambient temperature, relative humidity, printing press preheating status, and elapsed time after opening the paper, respectively, included in the moisture equilibrium vector, and calculating a thermal stabilization index by weighted combination of the heat-related state values, may be a step for quantifying the influence of the thermal environment of the rotary press on the micro-deformation of the paper and the coagulation and drying timing of the ink. For example, if preheating is insufficient or the ambient temperature changes rapidly, the thermal stabilization index may be set low to reflect unstable ink behavior; conversely, if sufficient thermal stability is secured, it is determined that the range of variable fluctuation during printing is small, allowing the response of the subsequent model to be set to a more stable condition.

[0114] The step of calculating a directional penetration coefficient for each of the main orientation direction of the paper and the direction orthogonal to the main orientation direction based on the fiber orientation and surface porosity included in the material normalization feature vector, and generating a biaxial transmission matrix from the directional penetration coefficient is a step of mathematically expressing the phenomenon in which ink penetrates and diffuses asymmetrically along the paper grain.

[0115] Here, the primary orientation direction may refer to the machine direction in which pulp fibers are predominantly arranged during the paper manufacturing or winding process, and the orthogonal direction may refer to a transverse direction perpendicular to the primary orientation direction. Additionally, the directional penetration coefficient may be a value indicating the relative ease of ink penetration in the corresponding direction, and the biaxial transmission matrix may refer to data expressing the directional transmission characteristics of the paper for the primary orientation direction and the orthogonal direction, respectively, in the form of a matrix.

[0116] The step of calculating a directional penetration coefficient for each of the main orientation direction of the paper and the direction orthogonal to the main orientation direction based on the fiber orientation and surface porosity included in the material normalization feature vector, and generating a biaxial transmission matrix from the directional penetration coefficient, may be a step to reflect the fact that the penetration speed and spreading width may vary depending on the orientation of the paper even when the same ink load is applied. For example, for a paper with strong fiber orientation and high surface porosity, the penetration coefficient for the main orientation direction may be calculated to be greater than that for the orthogonal direction, and the device may structure the difference into a biaxial transmission matrix and utilize it for the subsequent calculation of a directional load response vector.

[0117] The step of calculating a coating delay coefficient, which indicates the degree of delay in the initiation of ink penetration into the substrate layer based on the coating layer thickness deviation and surface porosity included in the material normalization feature vector and the paper moisture activity index, is a step of quantifying the degree to which the ink cannot penetrate into the substrate layer immediately after being applied to the paper surface and is delayed for a certain period of time by the coating layer or surface structure.

[0118] Here, the coating retardation factor refers to the initial penetration delay resulting from the combination of the coating layer's barrier effect, surface pore structure, and moisture status. For example, if the coating layer thickness variation is large, the surface porosity is low, and the paper's moisture activity index is high, the ink remains on the surface for a longer time, which can result in a relatively large coating retardation factor.

[0119] The step of calculating a surface retention coefficient indicating the degree of ink retention on the surface based on the surface roughness included in the material normalization feature vector and the paper moisture activity index is a step of quantifying the possibility that the ink will not immediately penetrate into the substrate layer but will remain on the paper surface and spread.

[0120] Here, the surface residence coefficient may be a value representing the tendency for surface retention determined by the surface roughness and moisture status of the paper. For example, if the surface roughness is high and the paper's moisture activity index is high, the surface residence time may increase as the ink interacts with fine irregularities or micro-water films, so a high surface residence coefficient can be calculated. On the other hand, if the surface is smooth and dry, the ink does not remain on the surface for a long time and can penetrate relatively quickly, so the surface residence coefficient can be set low.

[0121] The step of calculating a moisture swelling coefficient representing the degree of wet swelling of the paper surface based on the paper moisture activity index, thermal stabilization index, and surface porosity and coating layer thickness deviation included in the material normalization feature vector is a step of quantifying how much the current paper is swollen at the surface or near-surface layer due to moisture and thermal environments.

[0122] Here, the moisture swelling coefficient may be a value representing the degree of microstructural expansion of the paper in response to external humidity and thermal conditions. For example, if the paper has a high moisture activity index, a low thermal stabilization index, and a high surface porosity, the paper surface may swell more easily due to a humid environment, so the moisture swelling coefficient may be calculated to be high. Additionally, if there is a large variation in the thickness of the coating layer, there is a possibility that the swelling behavior may appear non-uniform in specific local areas, so a variation correction term may be included in the coefficient.

[0123] The step of calculating a saturation limit correction coefficient representing the amount of variation in ink acceptance limits based on the moisture swelling coefficient, thermal stabilization index, and the degree of whiteness reduction included in the material normalization feature vector is a step of correcting how quickly the paper approaches a saturated state at the time of actual printing.

[0124] Here, the saturation limit correction factor may be a value indicating how much the ink acceptance limit has shifted from the default value by reflecting the current state of the paper. For example, if the moisture swelling coefficient is high, the degree of whiteness degradation is high, and the thermal stabilization index is low, it can be seen that the ink acceptance characteristics of the paper surface and substrate layer have deteriorated; therefore, the saturation limit correction factor may be set high to advance the saturation critical condition.

[0125] The step of generating a reference load vector representing the reference print load state of a target surface using the above reference ink coverage rate, reference color load index, and reference dot gain index is a step of expressing the print intensity and color load degree typically required by the surface as a numeric load vector.

[0126] Here, the reference load vector may refer to vector data in which the reference ink coverage rate, reference color load index, and reference dot gain index are arranged in a specific dimensional structure. For example, since color advertising surfaces may have high reference ink coverage rates and reference color load indices, as well as high dot gain risk, the magnitude or directionality of the reference load vector may be configured differently from that of general article surfaces.

[0127] The step of calculating a directional load response vector by combining the above reference load vector and the 2-axis transmission matrix is ​​a step of generating a response vector that indicates how much absorption or diffusion response is expected in which direction when the printing load of the surface is combined with the anisotropic transmission characteristics of the paper.

[0128] Here, the directional load response vector may be a value expressed in vector form as a result of projecting a reference print load state in different ways depending on the main orientation direction and the orthogonal direction of the paper. For example, if a 2-axis transmission matrix is ​​combined with the reference load vector using matrix multiplication, a response value can be calculated in which the initial penetration or diffusion amount in that direction appears relatively larger when the transmission coefficient in the main orientation direction is larger, even under the same ink load conditions.

[0129] The step of inputting the above-mentioned directional load response vector, coating delay coefficient, surface residence coefficient, moisture swelling coefficient, and saturation limit correction coefficient into a time response function generation model to calculate the capillary penetration gradient immediately after ink application, the diffusion decay gradient over time, the lateral spreading sensitivity, and the critical time to reach saturation, and generating a set of time response functions based on the capillary penetration gradient, the diffusion decay gradient, the lateral spreading sensitivity, and the critical time to reach saturation, is a step of modeling the penetration and diffusion behavior over time after ink application in the form of a function.

[0130] Here, the time response function generation model may refer to a computational model that derives function parameters describing ink behavior on the time axis from the above input variables. The above model may be a pre-trained artificial intelligence model or a model based on complex fluid dynamics equations. Additionally, the capillary penetration gradient may be a value representing the gradient of the speed at which ink enters the paper by capillary action immediately after application, the diffusion decay gradient may be a value representing how much the diffusion speed decreases over time, the lateral spreading sensitivity may be a value representing the sensitivity to spreading laterally along the surface, and the saturation threshold time may refer to the expected time until the ink acceptance limit is reached.

[0131] The step of inputting the above-mentioned directional load response vector, coating delay coefficient, surface residence coefficient, moisture swelling coefficient, and saturation limit correction coefficient into a time response function generation model to calculate the capillary penetration gradient immediately after ink application, the diffusion decay gradient over time, the lateral spreading sensitivity, and the threshold time to reach saturation, and generating a set of time response functions based on the capillary penetration gradient, the diffusion decay gradient, the lateral spreading sensitivity, and the threshold time to reach saturation, may be a step for expressing the ink behavior that changes over time as a continuous or discrete set of functions rather than as a single numerical value.

[0132] Here, the set of time response functions may refer to a set of multiple functions or function coefficients for calculating the cumulative infiltration amount, surface retention amount, lateral diffusion amount, and the degree of approach to saturation over time. For example, a function may be generated that exhibits a steep infiltration trend in the initial period and becomes dominated by diffusion attenuation after a certain point in time, and when the surface retention coefficient is high, the slope of the decrease in the surface retention amount function may be formed relatively gently.

[0133] Based on the set of time response functions above, the step of calculating the initial penetration rate at the first reference point, the time-delayed diffusion coefficient at the second reference point, the transverse spreading coefficient, the anisotropic absorption coefficient, and the saturation critical absorption amount, respectively, is a step of extracting representative parameter values ​​that can be directly utilized for actual printing correction from the set of time response functions.

[0134] Here, the first reference point may be a time point representing the initial reaction period immediately after ink application, and the second reference point may be a subsequent time point where the ink exhibits more stable diffusion behavior between the surface and the substrate layer. The device can calculate the initial penetration rate at the first reference point and the time-delayed diffusion coefficient at the second reference point by sampling or integrating the set of time response functions, and can recalculate the transverse spreading coefficient and the anisotropic absorption coefficient by reflecting the directional response difference of the function, and can calculate the saturation critical absorption amount based on the saturation approach curve and the saturation attainment critical time.

[0135] The step of generating the set of absorption behavior parameters by combining the initial penetration rate, time-delayed diffusion coefficient, transverse spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount is a step of grouping the representative physical parameters derived in the preceding steps into a single consistent set and providing it to the subsequent step of calculating the absorption rate and dot expansion rate per channel.

[0136] Here, the absorption behavior parameter set may refer to a standardized set of parameters describing the penetration, diffusion, spreading, and saturation characteristics expected to be observed by the ink under actual printing conditions of the target paper and target surface. The device may reconstruct and store the parameters in the form of a vector or table with a constant dimensional structure, and may use the absorption behavior parameter set as common base data for calculating the first-order absorption rate, stabilized absorption rate, and dot expansion rate of each of the C channel, M channel, Y channel, and K channel.

[0137] For example, it can be assumed that a specific paper has a high surface porosity and a large variation in coating layer thickness, and that the paper moisture activity index is calculated to be high due to a high humidity environment during the summer. In this case, the device can generate a material normalization feature vector and then calculate both the coating delay coefficient and the surface residence coefficient as high, and can also set the moisture swelling coefficient and the saturation limit correction coefficient in an increasing direction. Subsequently, a directional load response vector is calculated by combining a reference load vector and a biaxial transmission matrix, and a set of time response functions can be generated through a time response function generation model in which the penetration gradient is steep initially but diffusion decay occurs rapidly after a certain point in time. Then, the initial penetration velocity at a first reference point, the time-delayed diffusion coefficient at a second reference point, the lateral spreading coefficient, the anisotropic absorption coefficient, and the saturation critical absorption amount can be calculated and configured into a single set of absorption behavior parameters. Accordingly, in the subsequent step, the ink absorption rate and dot expansion rate of each channel can be predicted more precisely based on the above set of absorption behavior parameters, and color bleeding and saturation degradation that may occur in the actual printing result can be compensated for in advance.

[0138] Additionally, the step of calculating the first absorption rate, stabilization absorption rate, and dot expansion rate for each of the above CMYK channels comprises: a step of setting a reference ink application amount, a reference dot area, and a reference dot diameter for each of the C channel, M channel, Y channel, and K channel based on the above reference ink coverage rate, reference color loading index, and target CMYK profile; a step of generating an hourly cumulative penetration curve, a surface residue curve, a first dot expansion curve in the main orientation direction, and a second dot expansion curve in a direction orthogonal to the main orientation direction for each of the above C channel, M channel, Y channel, and K channel by applying the above initial penetration rate, time-delayed diffusion coefficient, transverse blurring coefficient, anisotropic absorption coefficient, and saturation threshold absorption amount per channel; and a step of calculating a first cumulative penetration amount per channel and a first surface residue amount per channel from the ink application time point to a first reference time point based on the above hourly cumulative penetration curve and surface residue curve. A step of normalizing the first cumulative penetration amount per channel by the reference ink application amount per channel and performing a subtraction correction on the first surface residue amount per channel to calculate the first absorption rate for each of the C channel, M channel, Y channel, and K channel; a step of calculating the additional penetration amount per channel and the stabilized surface residue amount per channel from the first reference time point to the second reference time point based on the time-based cumulative penetration amount curve, the surface residue amount curve, and the saturation threshold absorption amount; a step of normalizing the stabilized cumulative penetration amount, which is the sum of the first cumulative penetration amount per channel and the additional penetration amount per channel, by the reference ink application amount per channel, and performing a correction on the stabilized surface residue amount per channel and the reference dot gain index to calculate the stabilized absorption rate for each of the C channel, M channel, Y channel, and K channel;The method may include the steps of: calculating a major orientation direction expansion amount and an orthogonal direction expansion amount for each of the C channel, M channel, Y channel, and K channel based on the first dot expansion curve and the second dot expansion curve, and calculating an equivalent elliptical area increase amount or an equivalent diameter increase amount for each channel from the major orientation direction expansion amount and the orthogonal direction expansion amount; and normalizing the equivalent elliptical area increase amount or the equivalent diameter increase amount for each channel with respect to the reference dot area or reference dot diameter for each channel, and performing weighted correction according to the reference color load index and the reference dot gain index to calculate a dot expansion rate for each of the C channel, M channel, Y channel, and K channel.

[0139] Based on the above-mentioned reference ink coverage rate, reference color load index, and target CMYK profile, the step of setting the reference ink coverage amount, reference dot area, and reference dot diameter for each of the C channel, M channel, Y channel, and K channel is a step of setting how much and geometric size each channel ink basically participates in printing under target surface and target paper conditions.

[0140] Here, the reference ink coverage amount may refer to the volume or mass of the reference ink supplied to each channel per unit area, the reference dot area may refer to the basic area occupied by a single halftone dot assuming a state before the ink spreads, and the reference dot diameter may refer to the initial diameter value when the dot is converted into a circle or an equivalent circle. For example, in the case of a surface where the proportion of the C channel in the target CMYK profile is higher than that of the M channel, and both the reference ink coverage rate and the reference color load index are large, the reference ink coverage amount and reference dot area of ​​the C channel may be set to be relatively larger than those of the M channel.

[0141] Based on the above-mentioned reference ink coverage rate, reference color load index, and target CMYK profile, the step of setting the reference ink coverage amount, reference dot area, and reference dot diameter for each of the C channel, M channel, Y channel, and K channel may be a step of providing a reference axis for normalizing the penetration amount, surface residue amount, and dot expansion amount calculated in a subsequent step. For example, even for the same channel, the absolute value of the reference coverage amount may be set differently depending on whether the target CMYK profile is a low-ink profile for newspaper printing or a profile for commercial printing with a relatively wide color gamut; therefore, the device may calculate the reference ink coverage amount, reference dot area, and reference dot diameter by referring to the channel-specific density table, black generation rules, and total ink amount constraints included in the target CMYK profile. Additionally, for surfaces where character outline preservation is important or fine sharpness is required, the reference dot diameter of the K channel may be separately corrected and set.

[0142] The step of generating a time-based cumulative penetration curve, a surface residue curve, a first dot expansion curve in the main orientation direction, and a second dot expansion curve in the direction orthogonal to the main orientation direction for each of the C channel, M channel, Y channel, and K channel by applying the above initial penetration rate, time-delayed diffusion coefficient, lateral spreading coefficient, anisotropic absorption coefficient, and saturation critical absorption amount for each channel is a step of modeling in the form of curves how much each channel ink penetrates into the paper, how much remains on the surface, and how much spreads in which direction over time.

[0143] Here, the cumulative penetration amount curve over time may be a curve representing the total amount of ink that has penetrated into the paper from the time of ink application to a specific time point in the form of a time function, and the surface residue amount curve may be a curve representing the change in the amount of ink remaining on the surface that has not yet sufficiently penetrated into the substrate layer on the same time axis. In addition, the first dot expansion curve may show the pattern of halftone dots spreading along the main orientation direction, and the second dot expansion curve may show the pattern of halftone dots spreading along a transverse direction orthogonal to the main orientation direction.

[0144] The step of generating a time-dependent cumulative penetration curve, a surface residue curve, a first dot expansion curve in the main orientation direction, and a second dot expansion curve in the direction orthogonal to the main orientation direction for each of the C, M, Y, and K channels by applying the initial penetration velocity, time-delayed diffusion coefficient, transverse spreading coefficient, anisotropic absorption coefficient, and saturation threshold absorption amount for each channel may be a step that reflects the fact that each channel may exhibit different behaviors because the pigment characteristics, viscosity, and visual influence of the ink for each channel differ even under the same paper conditions. For example, since the Y channel is relatively sensitive to visual density degradation, changes in the initial penetration amount can have a significant impact on the final color density, and the K channel may show prominent changes in contour sharpness due to dot expansion, the device may generate different sets of curves by using the same set of absorption behavior parameters but reflecting the reference coating amount and the reference dot value for each channel. Accordingly, the analog fluid behavior in which the ink spreads in an elliptical or asymmetric shape rather than a circular shape can be converted into digital curve data for each channel.

[0145] Based on the above-mentioned time-based cumulative penetration curve and surface residue curve, the step of calculating the first cumulative penetration amount and the first surface residue amount per channel from the time of ink application to the first reference time is a step of quantifying how much of each channel ink has moved into the paper and how much remains on the surface during the initial reaction section.

[0146] Here, the first reference point may be a time point where the initial kinetic reaction immediately after ink application is completed, and may correspond, for example, to a very short time interval immediately after stamping or a interval before initial drying begins. The first cumulative penetration amount may be calculated by integrating the cumulative penetration amount curve over time up to the first reference point or by extracting a value at a specific point in time, and the first surface residue amount may be calculated as the surface residue amount curve value at the same point in time or the average value over a certain interval.

[0147] Based on the above-mentioned time-based cumulative penetration curve and surface residue curve, the step of calculating the first cumulative penetration amount and the first surface residue amount for each channel from the time of ink application to the first reference time may be a step of separating the basic physical quantity required for calculating the actual absorption rate in the initial section. For example, even at the same first reference time, the surface residue amount for channel C may be relatively low and the first cumulative penetration amount for channel Y may be calculated to increase more rapidly; therefore, the device can calculate the initial response values ​​distinguished by each channel by reflecting the slope of the curve and the difference in cumulative amount.

[0148] The step of normalizing the first cumulative penetration amount for each channel by the standard ink application amount for each channel and performing a subtraction correction on the first surface residue amount for each channel to calculate the first absorption rate for each of the C channel, M channel, Y channel, and K channel is a step of normalizing and calculating the ratio that contributed to actual penetration in the initial reaction section for each channel.

[0149] Here, the first absorption rate may refer to the ratio of effective penetration into the paper relative to the standard ink amount supplied during the initial time period. The normalization may be a process of converting the first cumulative penetration amount into a relative ratio value by dividing it by the standard ink application amount, and the subtraction correction may be a process to prevent the amount of ink remaining on the surface from being mistaken for the actual penetration amount.

[0150] The step of normalizing the first cumulative penetration amount for each channel by the standard ink application amount for each channel and performing a subtraction correction on the first surface residue amount for each channel to calculate the first absorption rate for each of the C, M, Y, and K channels may be a step of evaluating relative absorption efficiency relative to the supply standard rather than the absolute amount of initial penetration. For example, even if the first cumulative penetration amount is the same, a channel with a larger standard ink application amount may have a relatively lower first absorption rate, while conversely, a channel with a smaller surface residue amount may be evaluated as having a higher first absorption rate under the same penetration amount conditions. Additionally, a channel with a large first surface residue amount may be judged to have a relatively high risk of future smudging or back-smudging, and thus the first absorption rate may be adjusted in a reduction direction.

[0151] Based on the above-mentioned cumulative infiltration curve, surface residue curve, and saturation critical absorption amount, the step of calculating the additional infiltration amount and the stabilized surface residue amount per channel from the first reference point to the second reference point is a step of calculating the additional infiltration and final residue state that proceed in the stabilization section after the initial reaction for each channel.

[0152] Here, the second reference point may be a subsequent time point where the ink reaches a more stable state between the surface and the substrate layer or drying is substantially completed. The additional penetration amount per channel may be a value indicating how much the cumulative penetration amount has increased from the first reference point to the second reference point, and the stabilized surface residue amount per channel may be the amount of ink still remaining on the surface at the second reference point or a corresponding representative value.

[0153] Based on the above-mentioned cumulative penetration curve, surface retention curve, and saturation critical absorption amount, the step of calculating the additional penetration amount per channel and the stabilized surface retention amount per channel from the first reference time point to the second reference time point may be a step that reflects subsequent penetration and retention behavior that cannot be explained by initial penetration alone. For example, under conditions where coating delay or a surface retention tendency is significant, a significant additional penetration amount may occur due to capillary action even after the first reference time point, and when approaching the saturation critical absorption amount, the increase in additional penetration amount slows down, leaving a relatively large amount of stabilized surface retention. Therefore, the above step may be a step of separately extracting subsequent behavior by distinguishing between the initial section and the stabilization section.

[0154] The step of normalizing the stabilization cumulative penetration amount, which is the sum of the first cumulative penetration amount per channel and the additional penetration amount per channel, by the reference ink application amount per channel, and performing corrections on the stabilization surface residue amount per channel and the reference dot gain index to calculate the stabilization absorption rate for each of the C channel, M channel, Y channel, and K channel is a step of calculating a ratio indicating how effectively each channel ink was absorbed in a finally stable state.

[0155] Here, the stabilization absorption rate may be the final absorption ratio per channel obtained by applying a correction considering the stabilization surface residue amount and dot gain risk to a value normalized with respect to the reference ink coating amount, which reflects both initial and subsequent penetration amounts.

[0156] The step of normalizing the stabilization cumulative penetration amount, which is the sum of the first cumulative penetration amount per channel and the additional penetration amount per channel, by the reference ink application amount per channel, and performing corrections on the stabilization surface residue amount per channel and the reference dot gain index to calculate the stabilization absorption rate for each of the C channel, M channel, Y channel, and K channel, may be a step of deriving a final absorption index that reflects actual printing stability rather than a simple cumulative penetration ratio. For example, even if the stabilization cumulative penetration amount is high, if the stabilization surface residue amount is excessively large, the possibility of surface bleeding or back bleeding remains, so the stabilization absorption rate may be reduced, and in a surface type with a high reference dot gain index, a more conservative stabilization absorption rate may be calculated even under the same penetration amount conditions.

[0157] The step of calculating the expansion amount in the main orientation direction and the expansion amount in the orthogonal direction for each of the C channel, M channel, Y channel, and K channel based on the first dot expansion curve and the second dot expansion curve, and calculating the equivalent elliptical area increase amount or equivalent diameter increase amount per channel from the expansion amount in the main orientation direction and the expansion amount in the orthogonal direction, is a step of converting the phenomenon of dots spreading asymmetrically in each direction into an area increase value or diameter increase value per channel.

[0158] Here, the expansion in the main orientation direction may refer to the dot length or radius component increased along the fiber orientation direction, and the expansion in the orthogonal direction may refer to the dot length or radius component increased in the direction perpendicular thereto. Additionally, the equivalent elliptical area increase may be a value indicating how much the area has increased relative to the initial area when the dot is approximated as an ellipse by reflecting the two expansions in the above directions, and the equivalent diameter increase may refer to the amount of diameter increased when converted into a circle having the same area.

[0159] The step of calculating the expansion amount in the main orientation direction and the expansion amount in the orthogonal direction for each of the C channel, M channel, Y channel, and K channel based on the first dot expansion curve and the second dot expansion curve, and calculating the equivalent elliptical area increase or equivalent diameter increase amount per channel from the expansion amount in the main orientation direction and the expansion amount in the orthogonal direction, may be a step for evaluating a dot deformation amount close to the actual printing result by reflecting an anisotropic expansion shape rather than simple linear blurring. For example, if the expansion amount in the main orientation direction appears larger than in the orthogonal direction, it can be interpreted that the halftone dots have been deformed into an ellipse having a major axis and a minor axis rather than a circle, and the device can quantify the dot increase amount per channel based on area or diameter using the deformation amount.

[0160] The step of normalizing the increase in the equivalent elliptical area or the increase in the equivalent diameter for each channel with respect to the reference dot area or reference dot diameter for each channel, and performing weighted correction according to the reference color load index and the reference dot gain index to calculate the dot expansion rate for each of the C channel, M channel, Y channel, and K channel is a step of converting the final degree of expansion of each channel dot into a relative ratio relative to the reference value.

[0161] Here, the dot expansion rate may refer to the ratio of the actual expected dot expansion amount to the reference dot area or reference dot diameter. The weighted correction based on the reference color load index may be intended to reflect that the interaction between dots and pressure effects may be amplified as the color load increases on the surface, and the weighted correction based on the reference dot gain index may be intended to reflect the dot expansion tendency typically observed in the corresponding surface type.

[0162] The step of normalizing the increase in the equivalent elliptical area or the increase in the equivalent diameter for each channel with respect to the reference dot area or reference dot diameter for each channel, and performing weighted correction based on the reference color load index and the reference dot gain index to calculate the dot expansion rate for each of the C, M, Y, and K channels, may be a step of generating a channel-specific dot expansion index to be directly used for subsequent channel-specific conversion weight adjustment and dot gain pre-compensation. For example, if the increase in the equivalent diameter of the K channel is large and the reference dot gain index is also high, the K channel dot expansion rate is calculated to be large, which can be used as a basis for correction in a subsequent step to pre-calculate the K channel density or halftone occupancy rate. Conversely, if the dot expansion rate of a specific channel is calculated to be relatively low, it may be utilized as a basis for increasing the channel-specific conversion weight to compensate for the reduction in color density of that channel.

[0163] For example, for a photograph to be inserted into a specific sports page, a condition can be assumed where the reference color loading index is high and the reference dot gain index is also high. In this case, the device can first set the reference ink application amount, reference dot area, and reference dot diameter for each channel based on the target CMYK profile. Then, by applying a set of absorption behavior parameters per channel, the device can generate a time-based cumulative penetration curve, a surface residue curve, and dot expansion curves in the main orientation direction and the orthogonal direction. Subsequently, the first cumulative penetration amount and the first surface residue amount up to the first reference point are calculated to derive the first absorption rate, and the stabilization absorption rate can be calculated by reflecting the additional penetration amount and the stabilization surface residue amount up to the second reference point. Additionally, the dot expansion rate for each channel can be calculated by obtaining the equivalent elliptical area increase or equivalent diameter increase using the dot expansion curves in both directions, normalizing this with respect to the reference dot area or reference dot diameter, and then applying weighted corrections based on the reference color loading index and reference dot gain index. Accordingly, in the subsequent step, the predicted ink absorption rate and the conversion weight for each channel can be adjusted more precisely based on the primary absorption rate, stabilization absorption rate, and dot expansion rate for each channel, and color bleeding, density reduction, and contour spreading in the actual printing result can be compensated in advance.

[0164] An apparatus according to one embodiment includes a processor and memory. The processor may include at least one apparatus described above through the drawings or perform at least one method described above through the drawings. The memory may store information related to the method described above or store a program in which the method described above is implemented. The memory may be volatile memory or non-volatile memory.

[0165] The processor can execute a program and control the device. The code of the program executed by the processor can be stored in memory. The device can be connected to an external device (e.g., a personal computer or a network) through an input / output device (not shown in the drawing) and exchange data.

[0166] The embodiments described above may be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible.

[0167] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0168] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or instruct the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium, or device so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0169] Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, suitable results may be achieved even if the described techniques are performed in a different order than described, and / or if the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0170] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below. Explanation of the symbols

[0171] S100: Step to acquire the original photo file S200: Step to generate preprocessed photo files S300: Step to generate a corrected photo file S310: Step to obtain publication metadata S320: Step for determining target CMYK profiles and total ink quantity limit values ​​for each printing condition S330: Step for calculating predicted values ​​for paper material characteristics and ink absorption rate S331: Step to retrieve paper history information S332: Step to generate basic material property vectors S333: Step to generate the moisture equilibrium vector S334: Step to query reference ink coverage rate, reference color load index, and reference dot gain index S335: Step of calculating the set of absorption behavior parameters S336: A step of calculating the first-order absorption rate, stabilization absorption rate, and dot expansion rate for each CMYK channel S337: Step of determining the paper material characteristics of the target surface S338: Step for calculating ink absorption contribution per channel S339: Step for calculating the predicted value of the integrated ink absorption rate S340: Step for calculating correction input features S350: A step of dividing a preprocessed photo file into at least two of a person skin area, sky area, vegetation area, shadow area, highlight area, text area, and graphic area. S360: Step to generate a set of color correction parameters S370: Step for dynamically adjusting CMYK channel-specific transform weights S380: Step to generate a corrected photo file

Claims

Claim 1 An artificial intelligence-based method for color correction of a photograph for printing, comprising a device including a processor, memory, a communication module, and a non-transient storage medium, wherein the method is performed by the processor by executing a program stored in the non-transient storage medium, the method comprises: a step of acquiring an original photograph file; a step of generating a preprocessed photograph file by performing at least one of size adjustment, cropping, mosaic processing, color conversion, and black-and-white conversion on the original photograph file; and a step of generating a corrected photograph file by using an artificial intelligence color correction model to automatically correct the saturation, brightness, and color balance of the preprocessed photograph file and performing color conversion to correspond to a target color profile set to be suitable for print output; wherein the step of generating the corrected photograph file comprises: a step of acquiring publishing metadata including newspaper edition information, page information, and date setting information associated with the original photograph file; a step of determining a target CMYK profile and a total ink amount limit value for each printing condition based on the publishing metadata; and a step of calculating predicted values ​​for the paper material characteristics and ink absorption rate of a target page based on the publishing metadata. A step of calculating correction input features including a histogram per RGB channel, a luminance histogram, a saturation distribution, a color temperature deviation, a local contrast value, an edge sharpness value, and a shake index from the preprocessed photo file; a step of dividing the preprocessed photo file into at least two of a human skin area, a sky area, a vegetation area, a shadow area, a highlight area, a text area, and a graphic area; a step of inputting the correction input features, the area division result, a target CMYK profile, and a total ink amount limit value into an artificial intelligence color correction model to generate a set of color correction parameters including a brightness correction coefficient per area, a saturation correction coefficient, a white balance correction coefficient, a sharpening intensity, a black generation amount, and a conversion weight per CMYK channel;An AI-based method for color correction of a photo for printing, comprising: a step of dynamically adjusting conversion weights for each CMYK channel by reflecting the predicted values ​​of the paper material characteristics and ink absorption rate; and a step of generating a corrected photo file by using the color correction parameter set to limit the color deviation of a person's skin area to within a preset reference range, suppressing gradation loss in dark and bright areas, performing dot gain pre-compensation for a target CMYK profile, and then mapping the preprocessed photo file to the target CMYK profile. Claim 2 In claim 1, the step of calculating predicted values ​​for paper material characteristics and ink absorption rate comprises: a step of querying paper history information including the lot number of the paper roll to be used for printing the target page, the paper supplier, the paper basis weight, the coating type, the calendaring processing history, the manufacturing date, and the storage period, based on newspaper company plate information, page information, and date setting information included in the publishing metadata; a step of generating a basic material characteristic vector including the fiber orientation, surface porosity, coating layer thickness deviation, surface roughness, and degree of whiteness degradation of the target paper, based on the paper history information; a step of generating an environmental equilibration vector including the ambient temperature immediately before printing, relative humidity, average humidity of the paper storage space, elapsed time after opening the paper, and the printing press preheating state, based on the date information and printing time information; and a step of querying a reference ink coverage rate, a reference color load index, and a reference dot gain index corresponding to the target page from a previously stored print history database by page type, based on the plate information and page information. A step of calculating a set of absorption behavior parameters including an initial penetration velocity of the paper surface, a time-delayed diffusion coefficient, a transverse blurring coefficient, an anisotropic absorption coefficient according to the winding direction, and a saturation critical absorption amount, using the above-mentioned basic material characteristic vector, moisture equilibrium vector, reference ink coverage rate, reference color load index, and reference dot gain index as inputs; and a step of calculating a first-order absorption rate, a stabilized absorption rate, and a dot expansion rate for each CMYK channel based on the above-mentioned set of absorption behavior parameters.A step of determining the set of fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the basic material characteristic vector, and the initial penetration rate, time-delayed diffusion coefficient, lateral blurring coefficient, anisotropic absorption coefficient, and saturation critical absorption amount as the paper material characteristics of the target surface based on the above-mentioned CMYK channel-specific stabilization absorption rate and dot expansion rate; a step of calculating the channel-specific ink absorption contribution by weightedly combining the fiber orientation, surface porosity, coating layer thickness variation, surface roughness, and degree of whiteness degradation included in the above-mentioned basic material characteristic vector, the above-mentioned initial penetration rate, time-delayed diffusion coefficient, lateral blurring coefficient, anisotropic absorption coefficient, and saturation critical absorption amount, and the above-mentioned CMYK channel-specific stabilization absorption rate and dot expansion rate by channel; and, based on the ink absorption contribution per channel, a step of calculating a predicted ink absorption rate per channel for C channel, M channel, Y channel, and K channel and a predicted integrated ink absorption rate for the entire target surface; comprising an artificial intelligence-based photo color correction method for printing; Claim 3 delete

Images