Method and system for a color matching process that corrects for coating process biases - Patents.com

Abstract

The present invention relates to a computer-implemented method for a color matching process in which an offset (310) is determined that removes application process bias. The present invention also provides a respective system.

Description

[Technical field]

[0001] The present invention relates to a method and system for a color matching process that compensates for coating process biases. [Background technology]

[0002] Most computer-aided color matching processes are based on physical models that describe the interaction of light with scattering or absorbing media (e.g. colorants in a paint layer). Each paint layer has specific light reflection properties depending on the colorants it contains. These colorants have specific optical properties that are represented by specific optical constants / specific optical data, respectively. Based on the information about the colorants it contains (respectively based on the information about the respective formulation), the physical model is able to predict the light reflection properties (color) of the paint layer / paint coating together with the corresponding specific optical properties (respectively corresponding specific optical constants).

[0003] The optical constants of a colorant describe the absorption and scattering properties (or effect flake orientation) of the colorant in the environment of the respective physical model, for example the K / S value in the well-known "Kubelka / Munk" model. However, the reflective properties of a paint layer do not only depend on the formulation. It also strongly depends on the painting process, i.e. how the paint is generally applied to the substrate.

[0004] The optical properties of the colorants are determined based on existing letdown / test piece sample data with known formulation and known reflectance data applied on the substrate by a common reference painting process. The physical model color prediction and color matching process is always relative to this reference painting process. The specific optical constants of the colorants include the effect of the reference painting process on the reflectance properties of each of the final paint layers, i.e. the final paint coating.

[0005] Physical model color predictions for different target painting processes are subject to significant systematic errors and lower accuracy.

[0006] Using a numerical optimization algorithm, the appropriate formulation for a given target color can be predicted based on a physical model that takes the existing optical constants of the available colorants and the reflectance data of the target color as input.

[0007] The resulting formulation should be able to match the target color as closely as possible, provided it is applied in a painting process equivalent to the reference painting process.

[0008] The color adjustment can be calculated based on the samples, e.g. search results in searching an existing coloring step or a formula database. The existing samples had to be applied with a reference painting process, since the color adjustment algorithm is based on one fundamental assumption, namely that the respective model bias is constant for all formulas close to the sample formula. As long as the adjusted formula of the color adjustment is similar to the sample formula, the model bias is expected to be similar. Generally, the color adjustment algorithm interprets the sample offset, i.e. the offset between the measured and predicted reflectance data of the respective sample, as a model bias. This model bias is automatically taken into account / corrected within the color adjustment algorithm, leading to a correction of the adjusted formula (e.g. patent EP2149038B1). However, besides the model bias, real samples may contain a systematic but non-constant bias caused by (small) differences within the respective painting process, also called application process bias. This application process bias propagates to the adjusted formula. Depending on the magnitude of the application process bias of the sample, the color adjustment result may be significantly inaccurate.

[0009] There are several methods available for applying paint to a substrate; namely, some examples of painting processes are: - automatic or manual spraying process, - Different types or varieties and configurations of spray guns in a laboratory or auto body shop; - Different spray lines or drying processes at OEM customer sites, - How to do drawdowns in the color development lab; It is.

[0010] Even if the same paint (same raw material) is applied in different ways, the resulting color of the paint layer (paint coating) can be strongly influenced by the respective application process. Reasons for color variation include: - Different orientation of effect flakes within the paint layer, - Overspray losses of effect flakes, especially large effect flakes, - Settling of effect flakes within the paint layer, - Change in thickness of non-hiding paint layer, - Changes in colour intensity (shear effects or flocculation) It is.

[0011] Therefore, it is important to use one common reference painting process to create the training data (let-down) for calculating the optical constants of the colorants, i.e., the specific optical data of each colorant.

[0012] In manual painting, e.g. with a spray gun, the resulting color depends on the type and configuration of the spray equipment as well as on the individual spray characteristics ("fingerprint") of the sprayer. Sprayer-dependent variations in one painting process are generally larger for manual painting processes than for automated painting processes. The spray profile of a manual painting process depends on the spray equipment (e.g. type of spray gun), the conditions in the spray booth (e.g. air temperature and humidity), the spray equipment configuration, the individual spray characteristics ("fingerprint") of the sprayer, and the drying process. These variations consist of a systematic part ("fingerprint") that is approximately constant for one individual sprayer / spray profile, and a statistical part.

[0013] Color matching is an iterative process: the matching method starts with a scratch match or search in a formula database for a given target color.

[0014] The term "matching from scratch" includes color matching processes that proceed without information about an existing sample coating as a first solution. This method is applied, for example, when a recipe database is not available or when no suitable first solution can be found in the recipe database. In practice, "matching from scratch" methods often start with a pre-selection step of ingredients expected to be present in the target color. This pre-selection step is not mandatory. The "matching from scratch" method / algorithm calculates one or more preliminary matching recipes for the target color as a first solution. This / these preliminary matching recipe(s) can be sprayed and / or adjusted in the next step.

[0015] Compared to "color adjustment methods" where sample coatings are available as a first solution that are used to improve the color prediction accuracy of a physical model (e.g. based on approximating model errors by analyzing "sample offsets"), the accuracy of "matching from scratch" methods is generally lower.

[0016] The first solution is typically not close enough to the target color. An adjustment of the first solution is applied where a sample offset between the predicted and measured colors of the first solution is taken into account. The sample offset consists of a systematic part (the "fingerprint" of the sprayer) and a statistical part. If the systematic part of the sample offset is not constant, for example, if the adjusted formula is sprayed by a different sprayer than the first solution (for example, if the formula is calculated based on the search results of a formula database containing solutions of many different sprayers with their individual spray profiles), the color adjustment result may be significantly inaccurate. The adjusted formula is a function of the target color and the sample offset. The non-constant sample offset is propagated to the adjusted formula.

[0017] The term "non-constant sample offset" includes that the offset of the adjusted formula (after application) is significantly different compared to the offset of the sample. [Prior art documents] [Patent documents]

[0018] [Patent Document 1] EP2149038B1 Summary of the Invention [Problem to be solved by the invention]

[0019] It is therefore an object of the present invention to provide the possibility to correct coating process biases in a color tuning method. [Means for solving the problem]

[0020] The above mentioned object is solved by a method and a system with the features of the respective independent claims. Further embodiments are given in the following description and in the respective dependent claims.

[0021] The present disclosure relates to a computer-implemented method for a color matching process, in which an offset free from application process bias (also referred to herein as a sample offset) is determined, the color matching process using a color adjustment algorithm including a color prediction model (also referred to simply as a physical model) implemented and executed on at least one computer processor, and a database including specific optical data of individual color components, the specific optical data of the individual color components being determined based on known reference paint coatings having known reference color formulas and known measured reference colors, each of the reference paint coatings being applied onto a substrate using a reference painting process, the method comprising: A. receiving, via at least one interface, data of a color formula of a sample paint coating as a first solution for a target color to be matched; B. retrieving from the database specific optical data for each color component used in the color formulation of the sample paint coating; and C. receiving, via at least one interface, a measured color of a sample paint coating applied onto a substrate using a sample painting process; D. calculating, using at least one computer processor and a numerical method implemented and executed on the at least one computer processor, application fit parameters for the sample painting process by minimizing a predetermined cost function starting from a set of predetermined initial application fit parameters, and making the calculated application fit parameters available as input parameters to the application fit module; E. predicting the color of the sample paint coating for a given sample painting process using a color prediction model, the data of the color formula of the sample paint coating, and the retrieved specific optical data of the individual color components used in the color formula of the sample paint coating as model input parameters, and including application match parameters calculated by an application match module in conjunction with the color prediction model; F. calculating an offset as the difference between the measured color and the predicted color of the sample paint coating; Includes.

[0022] The measured color of the sample paint coating can be retrieved from a database that includes data of measured colors of a number of paint coatings, including the sample paint coating, the paint coating having been previously applied onto the substrate using the sample painting process. Alternatively, the sample paint coating can be applied onto the substrate using the sample painting process when performing the proposed method, i.e. as a further step of the proposed method, i.e. during the performance of the proposed method.

[0023] Within the scope of the present invention, a sample coating process defines a particular spray profile that defines the spray equipment, the spray conditions, the configuration of the spray equipment, the individual spray characteristics ("fingerprint") of each individual sprayer, and / or the drying process.

[0024] The terms "specific optical data of individual color components", "specific optical data of the individual color components" or "specific optical data of the colorants" are used synonymously herein and comprise the specific optical properties and specific optical constants of each individual color component, i.e. the colorants. The individual color components used in the color formulation of the respective paint coating are selected from the group consisting of at least: color pigments, i.e. so-called solid pigments, effect pigments, binders, solvents and additives, e.g. matt pastes.

[0025] The terms "color," "color data," "reflectance," "reflectance data," and "reflectance characteristics" are used interchangeably herein. The terms "color formula," "paint formula," and "formula" are used interchangeably herein. The terms "processor" and "computer processor" are used interchangeably herein.

[0026] Known approaches to color blend calculations based on radiative transfer models can be found in the literature, see for example "Farbenphysik fur industrielle Anwendungen" by Georg A. Klein.

[0027] The basic idea of ​​color formula calculation is the characterization of specific optical data, i.e. optical properties and / or optical constants of all relevant individual color components (e.g. all color pastes / colorants), based on previously calibrated coatings, i.e. based on the respective measurements of such calibration coatings. These calibration coatings correspond to existing let-downs with known formulas and known reflectance data, all of which were applied by a common reference painting process. Color prediction using physical models (also referred to herein as color prediction models), and color matching processes, are always associated with this reference painting process. The specific optical constants of the colorants include the influence of the reference painting process on the predicted reflectance properties of the respective final paint coating / layer.

[0028] According to the invention, the physical model for predicting the reflectance properties (related to a reference painting process) of a sample paint coating, i.e. for predicting the color of the sample paint coating, is extended by an additional application adaptation module: the additional application adaptation module is configured to work in interaction with the physical model and to adapt the predicted reflectance data to a specific target painting process. The additional module is configurable by the input of calculated application adaptation parameters. These application adaptation parameters describe the differences, or rather specific transfer functions, between the sample painting processes compared to the respective reference painting process. Examples of application adaptation parameters are: - Coating thickness adaptation: thicker / thinner (Applicable to non-hiding paint layers; adjusts the hiding power of the paint layer) - Effect flake orientation matching: better flake orientation / worse flake orientation (Applicable to effect colors; adjusts the brightness / color flop of the paint layer) - Efficacy of solid colorants: Higher efficacy / lower efficacy (Adjusts the color strength difference of solid colorants that may occur due to shear effects and agglomerates, etc.) - Effect of colorant: Higher effect / lower effect (Adjusts for differences in reflectivity of effect colorants that may result from overspray losses or settling or residual effects) It is.

[0029] The application suitability parameters of the sample painting process can be implicitly determined based on an analysis of existing sample coating(s) applied in the sample painting process (e.g., one or more existing coloring step(s) of a color matching process). Also, a list of (one or more) existing sample coating(s) from a database (associated with the sample painting process, and in particular with an individual human sprayer) can be used for the determination of the application suitability parameters.

[0030] Typically, the sample application suitability parameters are determined based on sample data from a search result of a formula database. Alternatively, the sample application suitability parameters can be loaded from a database in which the application suitability parameters ("spray profiles") are stored in association with individual sprayers. The latter option is possible when information about the individual sprayers of the respective samples is available / given.

[0031] It is possible that a search result sample is sprayed by a first sprayer and the adjusted formula is sprayed by a different second sprayer. The spray profiles of the first and second sprayers are significantly different. This leads to systematic errors in the color adjustment algorithm that must be corrected. The application adaptation module "learns" the individual spray profile of the first sprayer from the search result sample and corrects the associated application process bias in the calculation of the sample offset that is used as the basis information in the color adjustment algorithm.

[0032] Alternatively, the coating suitability parameters (describing the spray profile) of the first sprayer may be loaded from a database containing coating suitability parameters corresponding to individual sprayers.

[0033] As already mentioned above, the application suitability parameters for the sample painting process are calculated using a numerical method implemented and executed on at least one computer processor. The numerical calculation method is configured to minimize a predetermined cost function. The numerical method and the color prediction model form part of the color adjustment algorithm.

[0034] According to one embodiment of the proposed method, the predefined cost function is selected as the color distance between the measured color of the second sample paint coating applied on the substrate using the sample painting process and the predicted color of the second sample paint coating, and a color prediction model is used to perform a color prediction of the second sample paint coating starting from a set of predefined initial application fit parameters using as input parameters the color formula of the second sample paint coating, the specific optical data of the individual color components used in the color formula of the second sample paint coating retrieved from the database, and the respective preliminary application fit parameters obtained in the course of the minimization, the application fit parameters being calculated by comparing the recursively predicted color of the second sample paint coating with the measured color of the second sample paint coating until the predefined cost function falls below a predefined threshold. The initial application fit parameters are neutral parameters, meaning that the use of the initial application fit parameters results in a color prediction equal to that of the use of the reference painting process. The predefined threshold can also be dynamically determined and can, for example, indicate a certain state of the minimization where no further improvement is possible. Application fit parameters can be calculated from a number of second sample paint coatings (all described by their respective color formulas) and the measured colors and treated as described above. Due to the large amount of data used, reliable application fit parameters can be obtained.

[0035] According to a further embodiment, the sample paint coating and the second paint coating are the same, or the sample paint coating and the second paint coating are different but applied by the same sample application compatible process.

[0036] According to a further embodiment of the proposed method, the application adaptation module is configurable by direct input, e.g. human input, of specific application adaptation parameters.

[0037] According to another embodiment of the proposed method, the application adaptation module is configurable by loading specific application adaptation parameters from a database in which the application adaptation parameters are stored in relation to a spray profile ID, e.g., an individual sprayer.

[0038] Each coating adaptation parameter is assigned to one adaptation means of a number of different adaptation means, the number of different adaptation means including at least one of: adapting layer thickness, adapting effect flake orientation distribution, adapting effect of solid color components, and adapting effect of effect color components.

[0039] According to a second aspect of the present invention, the method further comprises the steps of: - providing a color formulation calculation algorithm (also called a color adjustment algorithm), implemented and executed on at least one computer processor, for determining a target color formulation of a target paint coating that matches the target color when applied onto a substrate using a reference painting process; - using the target color and the calculated offset (from the sample paint coating) as input parameters of a color recipe calculation algorithm to calculate a color recipe having optimized concentrations of individual color components as a target color recipe for the target paint coating when the target paint coating is applied onto a substrate using a reference painting process; Includes.

[0040] According to an embodiment of the proposed method, the color blend calculation algorithm is realized by a numerical method and a color prediction model. The numerical method is configured to optimize the concentrations of the individual color components of the preliminary color blend in relation to the target color by minimizing a predetermined cost function, starting from a predetermined initial color blend, the predetermined cost function being specifically selected as the color distance between the received target color and the predicted color of the preliminary color blend. The color prediction model is configured to predict the color of the preliminary color blend by using as input parameters the calculated offsets of the respective sample paint coatings, the concentrations of the individual color components used in the preliminary color blend, the specific optical data of the individual color components used in the preliminary color blend and retrieved from a database. The optimized concentrations of the color components are calculated by comparing the recursively predicted color of the preliminary color blend with the target color until the predetermined cost function falls below a predetermined threshold. Each numerical method includes a number of successive approximation steps. In each approximation step, a preliminary color blend is assumed / provided and input to a physical model to predict the respective reflectance data, i.e. color, and then compared with the target color using the cost function. The predetermined threshold may also be dynamically determined and may, for example, indicate a particular state of minimization that cannot be further improved upon.

[0041] As already shown, in real samples, the measured reflectance data (measured color) always differs (slightly) from the predicted reflectance data (predicted color) of the physical model ("sample offset"). Reasons for this sample offset between reality and theory include, for example: - Model bias: no model is 100% accurate - Biases in the application process: e.g. special characteristics of human sprayers ("fingerprints") - Statistical errors of the instrument: e.g. caused by temperature There is.

[0042] Previously, the color adjustment algorithm interprets the entire sample offset as a model bias and corrects the adjusted paint formula so that each sample offset is corrected. Here, the application process bias is part of the sample offset and is expected to be constant. If the application process bias is not constant, it acts as an instability factor in the color adjustment algorithm. For example, if the measured color of the sample is too light due to the application process bias, the color adjustment algorithm calculates an adjusted formula that is too dark when applied with the reference painting process. Depending on the magnitude of the sample offset, the color adjustment results may be significantly inaccurate due to error propagation.

[0043] The proposed method of the present invention serves to eliminate such application process biases in the sample offset (also referred to herein simply as offset). Improving the accuracy of the sample offset directly improves the quality / accuracy of the adjusted paint formula. The sample offset between the measured color and the predicted color is analyzed with respect to the application process biases potentially contained in the measured color. The basic idea is to decompose the sample offset into an application process part and a residual part using an application fitting model. The residual part includes, for example, model biases and statistical errors. The application process part is removed from the sample offset since it is defined as non-stationary and systematic: that is, it represents the systematic application differences between the individual sprays and the reference painting process. The residual part of the sample offset mainly consists of model biases, which are correctly handled within the color adjustment algorithm (also referred to herein as color formula calculation algorithm).

[0044] As already mentioned above, the search result sample is sprayed with a first sprayer and the adjusted formula can be sprayed with a different second sprayer. The spray profiles of the first and second sprayers are significantly different. This leads to systematic errors in the color adjustment algorithm that must be corrected. The application adaptation module "learns" the individual spray profile of the first sprayer from the search result sample and corrects the associated application process biases in the calculation of the sample offset that is used as the basis information in the color adjustment algorithm.

[0045] Alternatively, the coating suitability parameters (describing the spray profile) of the first sprayer can be loaded from a database containing coating suitability parameters corresponding to individual sprayers.

[0046] The present invention also relates to a system, the system comprising at least: - a database comprising individual color components, such as pigments and / or pigment classes, and specific optical data associated with each individual color component, the specific optical data of each individual color component being determined based on known reference paint coatings having a known reference color formula and a known measured reference color, each reference paint coating being applied to a substrate using a reference painting process; - at least one computer processor, communicatively connected to the database and programmed to execute the proposed methods of the invention described herein; Includes.

[0047] The system may further comprise an input device configured to receive data input via a suitable interface, such as a USB. Such an input device may be a computer keyboard, a microphone, a video camera, a data carrier, or any combination thereof. The system may further comprise an output device configured to output, and in particular display, a respective result calculated by executing an embodiment of the method described above. The output device is one of the group including at least an acoustic device, a tactile device, a display device, and any combination thereof. The output device is communicatively connected to the at least one computer processor via a respective interface.

[0048] The present invention further relates to a non-transitory computer readable medium comprising a computer program having a program code configured and programmed therein, which when loaded and executed by at least one computer processor communicatively connected to a database comprising individual color components, such as pigments and / or pigment classes, and specific optical data associated with each individual color component, the specific optical data of the individual color components are determined based on a known reference paint coating having a known reference color formulation and a known measured reference color, and the reference paint coating is applied onto a substrate using a painting process to perform the proposed method of the invention described herein.

[0049] Each communication connection between the different components may be a direct or indirect connection. Each communication connection may be a wired or wireless connection. Any suitable communication technology may be used. The database and the at least one computer processor may each include one or more communication interfaces for communicating with each other. Such communication may be performed using a wired data transmission protocol, such as Fiber Distributed Data Interface (FDDI), Digital Subscriber Line (DSL), Ethernet, Asynchronous Transfer Mode (ATM) or other wired transmission protocol. Alternatively, the communication may be performed wirelessly over a wireless communication network using any of a variety of protocols, such as General Packet Radio Service (GPRS), Universal Mobile Telecommunications System (UMTS), Code Division Multiple Access (CDMA), Long Term Evolution (LTE), Wireless Universal Serial Bus (USB), and / or any other wireless protocol. Each communication may be a combination of wireless and wired communication.

[0050] Computer-readable media suitable for storing computer program instructions (i.e., program code) and data include all forms of non-volatile memory, media, and memory devices, including, by way of example, semiconductor memory devices such as erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; optical disks; CD-ROM, DVD+R, DVD-R, DVD-RAM, and DVD-ROM disks, or one or more combinations thereof. Such memory devices may store a variety of objects or data, including caches, classes, applications, backup data, jobs, web pages, web page templates, database tables, repositories for storing dynamic information, and other suitable information, including any parameters, variables, algorithms, instructions, rules, constraints, and / or references thereto. Additionally, memory may include any other suitable data, such as policies, logs, security or access data, report files, and the like. Computer processors and memory devices may be supplemented by, or incorporated in, special purpose logic circuitry.

[0051] The computer program instructions may be program software, a software application, a module, a software module, a script, or code, and may be written in any type of programming language, including compiled or interpreted languages, or declarative or procedural languages, and may be deployed in any form, such as as a stand-alone computer program or as a module, component, subroutine, or other unit suitable for use in a computing environment. In one embodiment, the computer executable instructions (i.e., program code) of the present disclosure are written in HTML, TypeScript (TS), and Cascading Style Sheets (CSS).

[0052] A computer program can correspond to a file in a respective file system, but need not. A computer program can be stored as part of a file that holds other programs or data, e.g., in one or more scripts stored in a markup language document, in a single file dedicated to the computer program, or in multiple coordination files, e.g., files that store one or more modules, subprograms, or code portions. A computer program can be deployed to run on one computer, or on multiple computers located at one site, or distributed across multiple sites and interconnected by a communication network. Some computer programs can be designed as individual modules that implement various features and functions through various objects, methods, or other processes. Alternatively, a computer program can instead include multiple submodules, third-party services, components, libraries, etc., as appropriate. Conversely, features and functions of various components can be combined into a single component as appropriate.

[0053] A system suitable for carrying out the disclosed method can be based on a general-purpose or special-purpose microprocessor, or both, or any other type of CPU. In general, the CPU will receive instructions and data from a read-only memory (ROM) or a random access memory (RAM), or both. The essential elements of the system are a CPU for executing or executing instructions (i.e., program code) and one or more memory devices (such as databases) for storing instructions (i.e., program code) and data. In general, the system includes at least one memory device, or is operatively coupled to at least one memory device, and configured to receive data from the one or more memory devices for storing data, or to transfer data to the at least one memory device, or both. The at least one memory device may include, for example, a magnetic, magneto-optical, or optical disk. However, the system itself need not have such a memory device. Furthermore, the system can be incorporated into other devices, such as a mobile phone, a personal digital assistant (PDA), or a portable storage device, such as a universal serial bus (USB) flash drive, to name just a few examples.

[0054] The following description is presented and provided in the context of one or more specific embodiments. Various modifications to the disclosed embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other implementations and applications without departing from the scope of the disclosure.

[0055] Implementations of the subject matter and functional operations described in this disclosure can be implemented in digital electronic circuitry, in tangibly embodied computer software, in computer hardware, including the structures disclosed in this disclosure and their structural equivalents, or in one or more combinations thereof. Implementations of the subject matter described in this disclosure can be implemented as one or more computer programs, i.e., as one or more modules of computer program instructions encoded on a tangible, non-transitory computer-readable medium for execution by, or control of operation by, a data processing device. Alternatively or additionally, the computer program instructions can be encoded on an artificially generated propagated signal, such as a machine-generated electrical, optical, or electromagnetic signal generated to encode information for transmission to a suitable receiving device executed by at least one computer processor. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random access or serial access memory device, or one or more combinations thereof.

[0056] The details of one or more embodiments of the subject matter of this specification are set forth in the accompanying drawings and description. Other features, aspects, and advantages of the subject matter will become apparent from the description, drawings, and claims. [Brief description of the drawings]

[0057] [Figure 1] FIG. 2 shows a schematic block diagram illustrating one embodiment of a method for providing a sample offset for a color matching process in accordance with the present invention; [Diagram 2] FIG. 2 shows a schematic block diagram illustrating another embodiment of a method for correcting coating process bias according to the present invention. [Diagram 3] FIG. 2 shows a schematic block diagram illustrating a further embodiment of a color matching process according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0058] Detailed Description of the Drawings Identical units or components are given the same reference numbers throughout all the figures.

[0059] Fig. 1 shows a schematic block diagram illustrating the offsets that application process biases must be removed according to an embodiment of the method according to the invention. In real samples, the measured sample color 100 always differs (slightly) from the predicted sample color predicted using a physical model. The measured sample color 100 can be represented as a combination of a true color 101 and an offset 110. The offset 110, also called sample offset 110, corresponds to the difference between the measured sample color 100 and the predicted sample color. The reasons for this sample offset 110 between reality (measurement) and theory (physical model) can be, for example: - Model Bias 114: No model is 100% accurate - bias in the application process 115: i.e. how the sample was applied to the substrate at the time the sample color is measured, e.g. special characteristics of a human sprayer ("fingerprint"); - statistical errors 112 of the equipment 113, i.e. application equipment such as spray guns: e.g. those caused by temperature It is.

[0060] Up to now, the color adjustment algorithm, i.e., the paint color formula calculation algorithm, interprets the entire sample offset 110 as a model bias and modifies the adjusted paint color formula in such a way that the respective sample offset 110 is compensated. This means that the application process bias 115 is part of the sample offset 110. If the application process bias 115 is not constant, e.g., for different human sprayers, it acts as an instability factor. For example, if the measured color of the sample is too light due to the application process bias 115, the color adjustment algorithm calculates an adjusted formula that is too dark when applied with the respective reference painting process. The magnitude of the sample offset 110 can lead to significant inaccuracies in the color adjustment results due to error propagation.

[0061] The proposed method of the present invention serves to remove such application process biases 115 in the sample offset 110 (also referred to herein simply as offset 110). Improving the accuracy of the sample offset 110 directly improves the quality / accuracy of the adjusted paint formula. The sample offset 110 between the measured color 100 and the predicted color is analyzed with respect to the application process biases 115 potentially included in the measured color 100. As mentioned before, the offset 110 includes a systematic bias 111 and a statistical bias 112, also referred to as a statistical error 112. The statistical bias 112 is caused by an instrument 113, such as a spray gun. The systematic bias 111 includes a model bias 114 (which is expected to be constant) and an application process bias 115 (which may not be constant). The basic idea of ​​the proposed method is to decompose the sample offset 110 into an application process bias 115 and a residual part including the model bias 114 and the statistical bias 112 using an application fitting model. Application process bias 115 is defined as not being constant and should therefore be removed from sample offset 110. The remaining portion of the sample offset consists primarily of model bias 114, which is properly handled within the color adjustment algorithm.

[0062] 2 illustrates one embodiment of a system 200 according to the present invention. The system includes a computer processor 210 and a database 220. The database 220 stores information about individual color components, such as pigments and / or pigment classes, Colorant 1, Colorant 2, Colorant 3, ..., Colorant n , and specific optical data associated with each individual color component, constant 1, constant 2, constant 3, ..., constant nThe specific optical data of the individual color components are determined based on known reference paint coatings having known reference color formulations and known measured reference colors, each of which is applied onto a substrate using a common reference painting process. The computer processor 210 is communicatively connected to the database 220 and is programmed to perform embodiments of the method as described herein.

[0063] In order to free the sample offset 110 from the application process bias 115 assigned to the sample painting process, it is proposed to determine the difference between the measured color of each sample coating applied on a substrate using the sample painting process and the predicted color of each sample coating determined using the physical model 240. Therefore, since the actual physical model 240 is based on the assumption of using a reference painting process, when making predictions using the physical model, it is important to take the sample painting process into account and to use a database 220 containing specific optical properties of colorants determined based on data of existing let-downs / test pieces with known formulations and known reflectance data, all applied on substrates by a common reference painting process. Taking the sample painting process into account is achieved by determining sample application fit parameters as further input parameters of the physical model 240. These sample application fit parameters are calculated based on data of existing coloring steps, respectively, of existing sample paint coatings.

[0064] These sample paint coatings are applied onto a substrate using a sample painting process. A sample color 203 of each of the sample paint coatings is measured. Data for a respective color formula 202 of each of the sample paint coatings is provided, where each color formula 202 includes all of the colorants included: Colorant 1, Colorant 2, Colorant 3, ... Colorant n and their respective concentrations, c1, c2, c3, . . . , c n This section stipulates:

[0065] Data for a color recipe 202 of a sample paint coating is received via at least one interface 211 of a computer processor 210. Additionally, measured colors 203 of the sample paint coating are received via at least one interface 211 of the computer processor 210.

[0066] A numerical method 230 and a physical model 240 are provided and implemented on the computer processor 210. The numerical method 230 is configured to optimize the application fit parameters by minimizing a predetermined cost function starting from a set of predetermined initial application fit parameters. The initial application fit parameters are neutral parameters, meaning that the use of the initial application fit parameters results in a color prediction equal to that of using a reference painting process. The predetermined cost function is selected as the color distance between the measured color 203 of each one of the existing sample paint coatings and the predicted color of each sample paint coating. The physical model 240 is configured to predict the color of each sample paint coating by using as input parameters the color formula 202 of each sample paint coating, the specific optical data of the individual color components used in the color formula 202 of each sample paint coating, and the respective preliminary application fit parameters resulting from the optimization process. The specific optical data is retrieved from the database 220. The specific optical properties of the colorants are determined based on data of existing letdowns / test pieces with known formulas and known reflectance data, all applied on a substrate by a common reference painting process. Therefore, the color predictions of the physical model 240 are related to this reference painting process. The colorant optical constants / data contain the effect of the reference painting process on the reflective properties of the final paint layer.

[0067] Using a computer processor 210, and using a numerical method 230 and a physical model 240 implemented and executed on the computer processor 210, the application suitability parameters 205 are calculated by comparing the recursively predicted color of each sample paint coating with the measured color 203 of each sample paint coating until a predefined cost function falls below a predefined threshold. The predefined threshold may also be dynamically determined and may, for example, indicate a certain state of minimization where no further improvement is possible.

[0068] The calculated optimized coating fit parameters 205 are made available and optionally output to an output device such as a display via a further interface 212. These calculated optimized coating fit parameters 205 are characteristic to the sample painting process. The coating fit parameters 205 are used as input parameters of the physical model 240 when determining a sample offset as the difference between the measured sample color and the predicted sample color of the first solution in the color adjustment method, as shown in FIG. 3. By using the coating fit parameters 205, both the measured sample color and the predicted sample color are related to the same sample painting process. Therefore, the difference between the measured sample color and the predicted sample color can be removed from the influence of the sample painting process.

[0069] The color adjustment process for a given target color 300 starts with a sample 301, e.g., an existing tinting step or a search result in a recipe database, as a first solution. So far, the existing sample has had to be applied with a reference painting process, since the color adjustment algorithm is based on the assumption that the model bias is constant for all recipes close to the sample recipe. However, as already explained, real samples or sample coatings are generally not applied with a reference painting process, but with a sample painting process that contributes to a systematic bias in the sample offset. If the contribution of the sample painting process to this sample offset is not taken into account, the results of the color adjustment process will be significantly inaccurate.

[0070] The first solution 301 is typically not close enough to the target color 300. An adjustment to the first solution 301 is applied where an offset 310 between the predicted reflectance data 306 and the measured reflectance data 303 for the first solution 301 is taken into account.

[0071] The adjusted formula is therefore a function of the target color 300 and the offset 310 between the predicted reflectance data 306 and the measured reflectance data 303 of the first solution 301. If the measured reflectance data 303 of the first solution 301 contains a bias caused by variations in the painting process, this error will propagate to the next formula in the iterative color matching process.

[0072] It is therefore proposed to avoid such painting process bias 115 by taking into account the painting process variability already in the first iteration step, i.e. when considering the first solution 301 .

[0073] A painting process independent offset 310 is calculated based on a first solution 301. A sample formulation 302 of the first solution 301 is known. The first solution 301 is applied as a paint coating on a substrate using a sample painting process and its color is measured. A measured color 303 of the first solution 301 is provided. The measured color 303 includes a true color 304, a systematic bias 305, and a statistical error 306. Furthermore, a physical model 240 is used to predict the color of the first solution 301 based on the known formulation 302. Since the physical model 240 uses the database 220 and is thus related to the reference painting process, the sample painting process is taken into account by combining the physical model 240 with the sample application fit parameters 205 determined as described in FIG. 2. A predicted color 306 of the first solution 301 is now predicted under the assumption that the underlying formulation 302 is applied as a paint coating on a substrate using the sample painting process. Thus, both the measured color 303 and the predicted color 307 refer to the same sample painting process. Thus, the offset 310, as the difference between the measured color 303 and the predicted color 307, is independent of the underlying sample painting process. This offset 310 can now be used in the iterative color adjustment process.

[0074] Since the first solution 301 is typically not close enough to the target color 300, a physical model 240 is used in combination with a numerical optimization algorithm 230 to obtain an iterative optimized blend 350. The optimized blend 350 is a blend of all the included colorants, Colorant 1, Colorant 2, Colorant 3, ... Colorant n and their respective concentrations, c1, c2, c3, . . . , c n The physical model 240 also uses the database 220 as the basis for color prediction. The target color 300 and the calculated offset 310 are used in combination to account for model bias and statistical errors. It is assumed that the combination of the sample and the target color 300 are similar.

[0075] The target color 300 and offset 310 are received via interface 211 by computer processor 210, in which physical model 240 and numerical optimization algorithm 230 are implemented and executed. The target color 300, offset 310 and specific optical constants of available colorants from database 220 are used to determine a formula 350 for a paint coating whose color matches the target color 300 when applied on a substrate using a reference painting process, and an optimized formula 350 is iteratively determined. This formula 350 and a predicted color 351 when applied using the reference painting process can be output to an output device via interface 212. The predicted color 351 is composed of the true color 352 of the optimized formula 350 when applied on a substrate using the reference painting process and the statistical error 353 of the sample. Due to the inclusion of offset 310 with the sample application bias 115 removed, the sample painting process bias 354 is no longer present. [Explanation of symbols]

[0076] 100 measured sample colors 101 True Colors 110 Offset 111 Systematic Bias 112 Statistical Error, Statistical Bias 113 Equipment 114 Model Bias 115 Coating process bias 200 Systems 202 color combinations 203 Measured Color 205 Sample Application Compatible Parameters 210 Computer Processors 211 (input) interface 212 (Output) Interface 220 Database 230 Numerical Optimization Algorithms 240 Physical Model 300 target color 301 Samples, Sample Coating 302 Sample Composition 303 Measurement sample color 304 True Colors 305 Systematic Bias 306 Statistical Error 307 Predicted sample color (for sample painting process) 310 Sample application compatibility parameters 350 Optimized Formulation 351 Predicted colors in a reference painting process 352 True Colors Statistical error in 353 samples 354 Sample Application Process Bias

Claims

1. 1. A computer-implemented method for a color matching process in which an offset (310) with application process bias (115, 354) removed is determined for a sample paint coating (301), said color matching process using a color prediction model (240) implemented and executed on at least one computer processor (210) and a database (220) containing specific optical data of individual color components, said specific optical data of said individual color components being determined based on known reference paint coatings having known reference color formulas and known measured reference colors, each of said reference paint coatings being applied onto a substrate using a reference painting process, said method comprising: A. receiving, via at least one interface (211) of said at least one computer processor, data of a color formula (302) of a sample paint coating (301) as a first solution for a target color (300) to be matched; B. retrieving from said database (220) specific optical data for each color component used in the color formula (302) of said sample paint coating (301); C. receiving, via said at least one interface (211), a measured color (303) of a sample paint coating (301) applied onto a substrate using a sample painting process; D. calculating, using said at least one computer processor (210) and a numerical method (230) implemented and executed on said at least one computer processor (210), application adaptation parameters (205) of said sample painting process by minimizing a predetermined cost function starting from a set of predetermined initial application adaptation parameters, and making said calculated application adaptation parameters (205) available as input parameters to an application adaptation module; E. predicting a color (307) of the sample paint coating (301) for the sample painting process using the color prediction model (240), the data of the color formula (302) of the sample paint coating (301), and the retrieved specific optical data of the individual color components used in the color formula (302) of the sample paint coating (301) as model input parameters, and including the calculated application match parameters (205) by an application match module in conjunction with the color prediction model (240); F. Calculating an offset (310) of the sample paint coating as the difference between the measured color (303) and a predicted color (307) of the sample paint coating (301); Including, The computer-implemented method, wherein the predetermined cost function is selected as a color distance between a measured color (203) of a second sample paint coating applied on a substrate using the sample painting process and a predicted color of the second sample paint coating.

2. The method of claim 1, wherein the color prediction model (240) is used to predict the color of the second sample paint coating starting from a predetermined set of initial application fit parameters by using as input parameters the color formula (202) of the second sample paint coating, specific optical data of the individual color components used in the color formula (202) of the second sample paint coating retrieved from the database (220), and respective preliminary application fit parameters obtained in the minimization process, wherein the application fit parameters are calculated by comparing the recursively predicted color of the second sample paint coating with the measured color (203) of the second sample paint coating until a predetermined cost function falls below a predetermined threshold.

3. The method of claim 2 , wherein the sample paint coating and the second sample paint coating are the same.

4. The method of any one of claims 1 to 3, wherein the application adaptation module is configurable by input of specific application adaptation parameters.

5. The method according to any one of claims 1 to 4, wherein each application adaptation parameter is assigned to one adaptation means of a number of different adaptation means, said number of different adaptation means including at least one of: adaptation of layer thickness, adaptation of orientation distribution of effect flakes, adaptation of effect of solid color components, adaptation of effect of effect color components, adaptation of individual characteristics of a human sprayer.

6. Further steps: - providing a color formulation calculation algorithm implemented and executed on said at least one computer processor (210) for determining a target color formulation (350) of a target paint coating that matches said target color (300) when applied onto a substrate using said reference painting process; - using said target color (300) and said calculated offset (310) as input parameters of said color blend calculation algorithm, calculating a target color blend (350) of said target paint coating when said target paint coating is applied onto a substrate using said reference painting process, together with optimized concentrations of the individual color components; The method according to any one of claims 1 to 5, comprising:

7. 7. The method of claim 6, wherein the color blend calculation algorithm is realized by a numerical method (230) and a color prediction model (240), the numerical method (230) being configured to optimize the concentrations of the individual color components of a preliminary color blend in relation to the target color (300) by minimizing a predetermined cost function, starting from a predetermined initial color blend, the predetermined cost function being specifically selected as the color distance between the received target color (300) and the predicted color of the preliminary color blend, the color prediction model (240) being configured to predict the color of the preliminary color blend by using as input parameters the calculated offset (310) of the sample paint coating, the concentrations of the individual color components used in the preliminary color blend, and specific optical data of the individual color components used in the preliminary color blend and retrieved from a database (220), the optimized concentrations of the color components being calculated by comparing the recursively predicted color of the preliminary color blend with the target color (300) until the predetermined cost function falls below a predetermined threshold.

8. at least: a database (220) comprising individual color components, such as pigments and / or pigment classes, and specific optical data associated with each of said individual color components, said specific optical data being determined on the basis of known reference paint coatings having a known reference color formulation and a known measured reference color, each of said reference paint coatings being applied to a substrate using a reference painting process; at least one computer processor (210), communicatively connected to said database (220) and programmed to execute the method according to any one of claims 1 to 7; Including, the system.

9. 8. A non-transitory computer readable medium having a computer program having a program code configured and programmed to perform the method of any one of claims 1 to 7 when loaded and executed by at least one computer processor (210) communicatively connected to a database (220) comprising individual color components, such as pigments and / or pigment classes, and specific optical data associated with each individual color component, wherein the specific optical data is determined based on a known reference paint coating having a known reference color formula and a known measured reference color, and wherein the reference paint coating is applied onto a substrate using a reference painting process.

Images