Color tone prediction device, color tone prediction method, and color tone prediction program

Abstract

This color tone prediction device comprises an acquisition unit that acquires the formulation of a composition, and a prediction unit that predicts the absorption coefficient and the scattering coefficient of the composition on the basis of a prescription of the composition.

Description

Color prediction device, color prediction method, and color prediction program

[0001] The present invention relates to a color prediction device, a color prediction method, and a color prediction program.

[0002] Traditionally, the color matching process for manufacturing products using colorants involved a person looking at a color sample, selecting the primary colors to use, considering the colorant mixture based on experience and past examples, creating a colored sample, and comparing it with the color sample, repeating this process until the color tone matched. However, this process has been streamlined by information technology, and with computer color matching technology, color samples are measured with a colorimeter and the colorant mixture is calculated by computer.

[0003] However, even if the blending ratio of coloring materials required to create a desired color is calculated and toned using computer color matching technology, a color tone different from the desired color tone may be created due to various factors such as the thickness of the colored layer, the manufacturing process, etc. Therefore, methods for reducing the gap, which is the color difference between the target color tone, which is the desired color tone, and the color created using computer color matching technology, have been studied (see, for example, Patent Documents 1 and 2).

[0004] JP 2023-075718 A JP 08-105797 A

[0005] In color matching calculations using computer color matching, information on the absorption characteristics (K) and scattering characteristics (S) of the colorants used is required. Normally, to measure these, the colorants must be applied to two types of substrates so that the film thickness is the same. Therefore, when it is difficult to control the film thickness, there are problems with measurement accuracy, and measurements of the absorption characteristics (K) and scattering characteristics (S) taking the film thickness into consideration have not been sufficient.

[0006] Furthermore, in the case of a coating film, such as a solder resist, which is formed through a process in which a composition made up of a mixture of multiple components is applied to a substrate and then heated and dried, there is a risk of a large discrepancy between the actual usage conditions and the conditions under which the absorption characteristics (K) and scattering characteristics (S) of the coloring materials contained in the composition are measured. For this reason, conventional computer color matching has not been able to predict color tones under actual usage conditions with sufficient accuracy.

[0007] An embodiment of the present invention has been made in consideration of the above-mentioned problems, and aims to provide a color tone prediction method, a color tone prediction system, and a color tone prediction program that are capable of predicting color tone from a prescription with high accuracy.

[0008] A color prediction device according to one aspect of the present invention includes an acquisition unit that acquires a composition formula, and a prediction unit that predicts an absorption coefficient and a scattering coefficient of the composition based on the composition formula.

[0009] According to one embodiment of the present invention, it is possible to provide a color prediction method, a color prediction system, and a color prediction program that are capable of predicting color from a prescription with high accuracy.

[0010] It is a schematic diagram showing the configuration of a color tone prediction system of this embodiment. It is a schematic diagram of the hardware configuration and functional configuration of a color tone prediction device of this embodiment. It is a flowchart showing an example of a process for creating a prediction model. It is a flowchart showing an example of a color tone prediction process. It is an image of creating a coating film sample. It is a schematic diagram showing an example of learning data.

[0011] Hereinafter, an embodiment of the present invention (hereinafter referred to as "the present embodiment") will be described in detail with reference to the drawings, but the present invention is not limited to this, and various modifications are possible within the scope of the gist thereof.

[0012] 1. Color Tone Prediction System Fig. 1A is a schematic diagram showing the configuration of a color tone prediction system 1 according to one embodiment of the present invention. As shown in Fig. 1A, in this example of the color prediction system 1, a server 100 (hereinafter also referred to as "color tone prediction device 100") that serves as a color tone prediction device and a user device 200 are communicably connected via a network N such as the Internet.

[0013] The color prediction device 100 is an information processing device realized by a color prediction program, and may transmit processing results to the user device 200 via the communication interface 120 and the network N in response to a processing request received from the user device 200. For example, the color prediction device 100 predicts a predicted absorption coefficient and a predicted scattering coefficient of a composition based on information regarding the formulation of the composition received from the user device 200. Then, for example, the color prediction device 100 may calculate a predicted spectral reflectance of a coating film made of the composition to be formed on a substrate based on the predicted predicted absorption coefficient and predicted scattering coefficient, the film thickness of the coating film, and the spectral reflectance of the substrate to which the composition is applied, and transmit the calculated spectral reflectance to the user device 200.

[0014] The user device 200 is an information processing device used by a user to execute the color tone prediction process, and may be, for example, a computer, a smartphone, a tablet terminal, a personal computer, or the like.

[0015] FIG. 1A shows a client / server system including a color prediction device 100 and a user device 200. In the following, we will explain the various ways in which the server functions as the color prediction device 100. However, the system of this embodiment is not limited to this, and instead, the user device 200 may have the processing functions of the color prediction device, which will be described later.

[0016] 1.1 Color Tone Prediction Device The hardware configuration and functional configuration of the color tone prediction device 100 will be described below with reference to FIG. 1B, and then each control will be described in detail in association with the functional configuration of the color tone prediction device 100.

[0017] As shown in FIG. 1B, the color prediction device 100 includes, for example, a processor 110, a communication interface 120, an input / output interface 130, a memory 140, a storage 150, and one or more communication buses 160 for interconnecting these components.

[0018] The processor 110 executes processes, functions, or methods implemented by code or instructions included in a program stored in the storage 150. The processor 110 may include, for example and without limitation, one or more central processing units (CPUs), MPUs, GPUs, etc., and may implement the processes, functions, or methods disclosed in each embodiment by logic circuits (hardware) formed in an integrated circuit or the like, or by dedicated circuits.

[0019] As shown in FIG. 1B , the processor 110 of this embodiment may be configured to function as an estimation unit 111 , a learning unit 112 , an acquisition unit 113 , a prediction unit 114 , a calculation unit 115 , and a proposal unit 116 .

[0020] The communication interface 120 transmits and receives various data to and from other devices via the network N. The communication may be performed either wired or wirelessly, and any communication protocol may be used as long as mutual communication is possible. For example, the communication interface 120 may be implemented as hardware such as a network adapter, various types of communication software, or a combination of these.

[0021] Network N may be, by way of example and not limitation, an Ad Hoc Network, an Intranet, an Extranet, a Virtual Private Network (VPN), a Local Area Network (LAN), a Wireless LAN (WLAN), a Wide Area Network (WAN), a Wireless WAN (WWAN), a Metropolitan Area Network (MAN), a portion of the Internet, a portion of the Public Switched Telephone Network (PSTN), a cellular telephone network, an Integrated Services Digital Network (ISDN), or the like. The network may be a digital network, a wireless LAN, a Long Term Evolution (LTE), a Code Division Multiple Access (CDMA), a Bluetooth (registered trademark), a satellite communication, or the like, or a combination thereof. The network may include one or more networks.

[0022] The input / output interface 130 includes an input device for inputting various operations to the color prediction device 100, and an output device for outputting processing results processed by the color prediction device 100. For example, the input / output interface 130 includes information input devices such as a keyboard, a mouse, and a touch panel, and information output devices such as a display. Note that the color prediction device 100 may accept predetermined inputs and execute predetermined outputs by connecting an external input / output interface 130.

[0023] The memory 140 temporarily stores programs loaded from the storage 150 and provides a working area for the processor 110. The memory 140 also temporarily stores various data generated while the processor 110 is executing the programs. The memory 140 may be, for example, a high-speed random access memory such as a DRAM, an SRAM, a DDR RAM, or other random access solid-state storage device, or a combination of these.

[0024] The storage 150 stores programs, each functional unit, and various data. The storage 150 may be, for example, one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or nonvolatile memories such as other nonvolatile solid-state storage devices, or a combination thereof. Another example of the storage 150 may be one or more storage devices installed remotely from the processor 110.

[0025] Next, each functional unit of the color prediction device of this embodiment will be described in detail. The estimation unit 111 estimates the absorption coefficient Km and scattering coefficient Sm of a composition sample having a predetermined recipe based on the spectral reflectance Rm of a coating sample formed from the composition sample at a plurality of different film thicknesses and the film thicknesses. The learning unit 112 then creates a prediction model that predicts the absorption coefficient and scattering coefficient of a composition from its recipe based on the absorption coefficients and scattering coefficients estimated by the estimation unit 111 for the plurality of composition samples having different recipes and the recipes.

[0026] Below, we will explain in detail the process of creating a prediction model by the estimation unit 111 and the learning unit 112. However, before going into the details of this embodiment, we will first explain, as background knowledge, color material mixing and color tone prediction in general computer color matching.

[0027] As an example, consider a case where a paint film is formed by adjusting the amount of color material to match a certain color sample using general computer color matching. In this case, whether the color tone of the paint film to be formed matches that of the color sample can be evaluated by spectral reflectance. If the difference in spectral reflectance between the color sample and the paint film is small, the color tones of the two can be evaluated as being homologous. If the difference in spectral reflectance between the color sample and the paint film is large, the color tones of the two can be evaluated as being different. Furthermore, if the spectral reflectance of the paint film is known, the color space L*, a*, b* values ​​of the paint film can be calculated, and the color tone of the paint film can be calculated. Therefore, when a paint film is formed by adjusting the amount of color material to match a certain color sample, computer color matching has a computer perform calculations to adjust the amount of color material contained in the paint film, etc., so as to reduce the difference in spectral reflectance between the color sample and the paint film.

[0028] However, spectral reflectance is not additivity, so even if the spectral reflectance of a certain colorant is known, it is not possible to predict the spectral reflectance of a coating film obtained by changing the amount of that colorant.

[0029] Therefore, in computer color matching, instead of spectral reflectance, which does not have additive properties, absorption coefficients and scattering coefficients, which do have additive properties, are used. The following formula (1) is used in this case. Formula (1) below shows the relationship between the absorption coefficient Km and scattering coefficient Sm of a composition having a formula m composed of multiple substances i (=1, 2, ...). Specifically, the absorption coefficients and scattering coefficients of the object to be colored and the absorption coefficients and scattering coefficients of each colorant are investigated in advance and compiled into a database. Based on this, the absorption coefficients and scattering coefficients of a composition obtained by mixing the object to be colored with one or more colorants are calculated using formula (1). The data summarizing the absorption coefficients and scattering coefficients of each colorant used in computer color matching is also referred to as "basic data."

[0030] The following equation (2) based on the Kubelka-Munk theory is used when obtaining basic data. Equation (2) below shows the relationship between the spectral reflectance R of a colorant or a colored object at a specific wavelength (λ) and the absorption coefficient K and scattering coefficient S. Specifically, it shows the spectral reflectance Rm of a coating film m formed from a composition comprising formulation m, which has an absorption coefficient Km and a scattering coefficient Sm. By forming coating films of the same thickness on two types of substrates with different reflectances using a composition comprising formulation m and measuring the spectral reflectance of each coating film, the absorption coefficient Km and scattering coefficient Sm of a composition comprising formulation m can be obtained from equation (2). When formulation m consists of only substance i, basic data for substance i can be obtained.

[0031] Note that, since the recognition of color tone by the human eye is a consideration, the wavelength λ in this embodiment refers to a wavelength in at least the visible light region. Furthermore, although the spectral reflectance, absorption coefficient, and scattering coefficient all take values ​​that depend on the wavelength λ, λ will be omitted in this embodiment. When the wavelength λ is not specifically mentioned when referring to spectral reflectance, etc., it is understood that the description is of spectral reflectance, etc. in the visible light region, without being limited to a specific wavelength.

[0032]

[0033] For a colorant layer such as a paint, which has a large thickness and whose spectral reflectance is not affected by either the thickness or the substrate, the following simplified formula (3), which is derived by setting X to ∞ in the above formula (2) and setting the spectral reflectance Rg of the substrate to 0, is commonly used (see, for example, Patent Documents 1 and 2):

[0034] However, for example, when a thin film is formed on a metal substrate such as copper, as in the case of solder resist, the film thickness is thick and the assumption that the spectral reflectance is not affected by either the film thickness or the substrate cannot be applied, and formula (3) cannot be used. Furthermore, while solder resist is used as an example above, this is merely one example, and situations in which the film thickness is thick and the assumption that the spectral reflectance is not affected by either the film thickness or the substrate cannot be applied are not limited to solder resist. In the following discussion, formula (2) will be used as a premise.

[0035] When attempting to estimate the spectral reflectance of a coating film from the formulation of a composition using the above formulas (1) and (2) by conventional computer color matching, basic data on the absorption coefficient Ki and scattering coefficient Si of each component obtained in advance are essential.

[0036] As described above, in order to determine the absorption coefficient Km and scattering coefficient Sm of a composition comprising formula m, for example, in order to use formula (2), it is necessary to measure the spectral reflectance using coating films formed by applying the composition comprising formula m to two types of substrates, one white and one black, at the same thickness. However, in practice, it is extremely difficult to form coating films at exactly the same thickness on two types of substrates, and if there is a difference in the thickness of the coating film, the calculated absorption coefficient Km and scattering coefficient Sm will contain errors. Similar problems can also occur when obtaining basic data for each component. Therefore, if the basic data contains errors, the absorption coefficient Km and scattering coefficient Sm of the composition calculated using formula (1) will also contain errors, which may reduce the prediction accuracy.

[0037] Furthermore, for example, in the case of a coating film, such as a solder resist, which is formed through a process in which a composition containing a large number of components is applied to a substrate and then heated and dried, there is a risk of a large discrepancy between the actual use state and the state in which basic data on the components contained in the composition is measured. Specifically, for example, the coating film used to obtain basic data such as the coloring materials contained in the composition and the coating film whose color tone is actually predicted may have significantly different preconditions, such as the components contained in the coating film and the process such as the temperature and time of heating and drying. In this way, the color tone of the coating film predicted using basic data measured under significantly different preconditions does not reflect the actual components and process, and therefore the prediction accuracy may be reduced.

[0038] In contrast, this embodiment takes an approach that does not rely on the absorption coefficient Ki and scattering coefficient Si of each component. Specifically, the absorption coefficient Km and scattering coefficient Sm of a composition sample having a predetermined formulation are estimated from the spectral reflectance Rm(x) and thickness x of a coating film sample formed from a composition sample having a predetermined formulation at multiple different film thicknesses. Based on the absorption coefficients Km and scattering coefficients Sm estimated for multiple composition samples having different formulations and the formulations, a prediction model can be created that predicts the absorption coefficient Km and scattering coefficient Sm of a composition from its formulation. Because the prediction process is performed using a prediction model trained on the absorption coefficient Km and scattering coefficient Sm of the composition sample as a whole, it is not necessary to obtain the absorption coefficient Ki and scattering coefficient Si of each component. Furthermore, the absorption coefficient Km and scattering coefficient Sm of the composition sample estimated in this way incorporate the thickness and formation process of the coating film sample, as well as the formulation of the composition sample, thereby improving the prediction accuracy of the absorption coefficient and scattering coefficient. Hereinafter, one aspect of the color prediction method of this embodiment will be described with reference to the drawings.

[0039] FIG. 2A shows a flowchart for creating a prediction model of absorption coefficients and scattering coefficients used in the prediction process in the color tone prediction method of this embodiment.

[0040] As shown in FIG. 2A, in step S01, a plurality of coating film samples are prepared by forming a coating film on a predetermined substrate with a composition sample of a predetermined formulation A at different film thicknesses, and the spectral reflectance Rm of the coating film samples is measured. Ａ The film forming conditions other than the film thickness (for example, drying temperature and drying time) are the same, and each coating film sample is prepared. The same operation is also carried out for the composition samples of formulations B, C, etc., and the spectral reflectance Rm Ｂ , Rm Ｃ , ... are measured. An image of the creation of this coating sample is shown in Figure 3. Note that the "coating sample" refers to a coating created to acquire data when creating learning data and a prediction model, and the "composition sample" refers to a composition used to create the coating sample. The spectral reflectance Rm of the coating sample may be acquired and recorded by the acquisition unit 113.

[0041] Then, in step S02, the estimation unit 111 calculates the spectral reflectance Rm of the coating sample of formula A. Ａ From this, the absorption coefficient Km Ａ and the scattering coefficient Sm Ａ Specifically, as shown in FIG. 3, in step S01, coating film samples with film thicknesses of 20 μm, 30 μm, and 40 μm are formed using a composition sample of prescription A, and the spectral reflectance Rm Ａ (20), Rm Ａ (30), Rm Ａ (40) are measured. These spectral reflectances can be calculated using the formula (2) as the theoretical value R cal m Ａ (20), R cal m Ａ (30), R cal m Ａ (40) can be calculated as follows:

[0042] Here, Rg can be obtained by measuring the spectral reflectance of the substrate. Furthermore, a and b are expressed by the following formula: Ａ , Sm Ａ Since the value is expressed by cal m Ａ (20), R cal m Ａ (30), R cal m Ａ (40) is km Ａ and Sm Ａ It is a function of the actual measured value Rm Ａ (20), Rm Ａ (30), Rm Ａ (40) and the theoretical value R cal m Ａ (20), R cal m Ａ (30), R cal m Ａ Km that minimizes the difference between (40) Ａ and Sm Ａ The absorption coefficient Km of the composition sample of formulation A can be calculated by using a method such as curve fitting. Ａ and the scattering coefficient Sm ＡIt should be noted that such a calculation cannot be performed using equation (3) which assumes that X is ∞.

[0043] The measured spectral reflectance was corrected by Sanderson correction and expressed as Rm Ａ (20), Rm Ａ (30), Rm Ａ (40) By performing the Sanderson correction, the influence of reflection due to the difference in refractive index between the coating film and air can be eliminated, making it possible to predict color tone with higher accuracy.

[0044] Absorption coefficient Km Ａ and the scattering coefficient Sm Ａ The specific calculation method is not particularly limited, but for example, a curve fitting process may be performed. The curve fitting process is not particularly limited, but for example, the least squares method or the maximum likelihood method may be used.

[0045] In the above example, a composition sample with formula A was used as an example, but by applying the same operations to composition samples with different formulas such as formula B, C, etc., it is possible to obtain training data 151 for creating a predictive model.

[0046] In this way, the absorption coefficient Km and scattering coefficient Sm of the composition sample in a predetermined recipe obtained by the estimation unit 111 may be recorded for each recipe in the training data 151. Fig. 4 shows an example of the training data 151. As shown in Fig. 4, the training data 151 may record information such as a "coating sample ID" for uniquely identifying a coating sample, a "recipe" of the composition used to create the coating sample, a "forming process" and "thickness" of the coating using the composition, and the "spectral reflectance" of the coating sample, and the "absorption coefficient" and "scattering coefficient" of the composition sample.

[0047] Subsequently, in step S03, the learning unit 112 creates a prediction model that predicts the absorption coefficient Km and scattering coefficient Sm of a composition from the formulation of the composition, based on the thus-obtained learning data 151. The prediction model created at this time may predict the absorption coefficient Km and scattering coefficient Sm of the coating film further based on the formation process.

[0048] A known method such as machine learning can be used to create the prediction model. In this embodiment, since certain constraints are imposed on the absorption coefficient Km and the scattering coefficient Sm based on Equation (2) as in step S02, a highly accurate prediction model can be created even by a machine learning method using a linear model. An example of a linear model is the Huber regression model.

[0049] Furthermore, in creating the prediction model, the prediction model may be created without using the absorption coefficient Km and scattering coefficient Sm of the prescription for which the difference between the actual measured value and the theoretical value was large in the curve fitting in step S02 as learning data.

[0050] Next, the acquisition unit 113, the prediction unit 114, and the calculation unit 115, and the calculation process of the predicted spectral reflectance performed by these units will be described in detail.

[0051] The acquisition unit 113 acquires the composition recipe from the user device 200 via the communication interface 120 and the network N. The acquisition unit 113 may further acquire the film thickness of a coating film formed using the composition, or the spectral reflectance of a substrate to which the composition is applied. In this embodiment, the substrate refers to a base material to which the composition is applied.

[0052] The prediction unit 114 predicts the predicted absorption coefficient and predicted scattering coefficient of the composition based on the formulation of the composition. Furthermore, the prediction unit 114 may predict the predicted absorption coefficient and predicted scattering coefficient based on, for example, the prediction model described above, although not particularly limited thereto.

[0053] The calculation unit 115 may calculate the predicted spectral reflectance of the coating film based on the predicted absorption coefficient and scattering coefficient of the composition and the film thickness of the coating film formed using the composition, or may calculate the predicted spectral reflectance of the coating film based on the predicted absorption coefficient and scattering coefficient of the composition and the spectral reflectance of the substrate. The calculation unit 115 may calculate the predicted spectral reflectance of the coating film based on the predicted absorption coefficient and scattering coefficient of the composition, the film thickness of the coating film formed using the composition, and the spectral reflectance of the substrate. The calculation of the predicted spectral reflectance may use equation (2) based on the Kubelka-Munk theory. Furthermore, the calculation unit 115 may generate information such as Lab from the predicted spectral reflectance. Specifically, the L*a*b* values ​​can be calculated based on the predicted spectral reflectance, the spectrum of incident light, and color matching functions. The spectrum of incident light can be, for example, sunlight or the spectrum of illumination light in the usage environment. The color matching functions can be, for example, the color matching functions of a 2-degree standard observer or a 10-degree standard observer.

[0054] Next, the prediction process in the color tone prediction method of this embodiment using the above prediction model will be described with reference to FIG. 2B.

[0055] In step S11, the acquisition unit 113 receives input from a user via the input / output interface 130 or the like, acquires the composition recipe, and may also acquire the film thickness of the coating film and the spectral reflectance of the substrate to which the composition is applied.

[0056] Then, in step S12, the prediction unit 114 predicts the predicted absorption coefficient Kp and the predicted scattering coefficient Sp of the composition based on the acquired composition prescription. At this time, the prediction unit 114 may predict the predicted absorption coefficient Kp and the predicted scattering coefficient Sp of the composition using the above-mentioned prediction model.

[0057] Next, in step S13, the calculation unit 115 calculates a predicted spectral reflectance of the coating film based on the predicted absorption coefficient Kp and predicted scattering coefficient Sp of the composition, the film thickness of a coating film formed using the composition, and the spectral reflectance of a substrate to which the composition is applied. In this case, the calculation unit 115 may calculate the predicted spectral reflectance using the above-mentioned formula (2). Finally, in step S14, the calculation unit 115 may calculate L*a*b* values ​​based on the predicted spectral reflectance.

[0058] In this manner, in this embodiment, the predicted absorption coefficient Kp and predicted scattering coefficient Sp of a composition are predicted from the formulation of the composition using a prediction model, and based on these, it becomes possible to predict with high accuracy the spectral reflectance and L*a*b* values ​​of a coating film formed using the composition.

[0059] The composition is preferably a solder resist. Conventional computer color matching has been widely used in applications where coloring, such as printing, is the primary purpose. In the printing field, basic data such as the absorption coefficient and scattering coefficient of colorants are readily available. However, for industrial products such as solder resist, the primary purpose is to ensure the performance of the industrial product, and basic data such as the absorption coefficient and scattering coefficient are not readily available under the conditions of use for the industrial product. Furthermore, because solder resist may contain multiple components, it is extremely difficult to obtain basic data for each component. Furthermore, as described above, if the basic data contains errors, the errors in the basic data for the multiple components are accumulated, potentially resulting in very large errors in the ultimately predicted absorption coefficient and scattering coefficient of the solder resist. However, even in the case of industrial products for which basic data is unavailable or compositions composed of multiple components, the present embodiment makes it possible to accurately predict the spectral reflectance of a coating film formed using the composition based on the composition's formulation.

[0060] The color prediction device of this embodiment may also include a suggestion unit 116 that proposes a composition formula that satisfies a desired color tone based on a color tone desired by a user. For example, the acquisition unit 113 generates a plurality of composition formulas at random, the prediction unit 114 predicts the absorption coefficients and scattering coefficients of the compositions based on the composition formulas using the prediction model, and the calculation unit 115 calculates the predicted spectral reflectance of a coating film made of the compositions. The suggestion unit 116 may then identify a composition formula in which the difference between the calculated predicted spectral reflectance and the spectral reflectance corresponding to the user's desired color tone is equal to or less than a certain value, and propose the composition formula. Furthermore, the acquisition unit 113 may use known algorithms such as a genetic algorithm or Bayesian optimization to generate a plurality of composition formulas at random.

[0061] 2. Color Tone Prediction Method The color tone prediction method of this embodiment includes an acquisition step of acquiring a composition formula, and a prediction step of predicting the absorption coefficient and scattering coefficient of the composition based on the composition formula.

[0062] Furthermore, in the color tone prediction method of the present embodiment, the obtaining step may further include a calculating step of obtaining a spectral reflectance of a substrate to which the composition is applied, and calculating a predicted spectral reflectance of a coating film made of the composition to be formed on the substrate based on the absorption coefficient and scattering coefficient predicted in the predicting step and the spectral reflectance of the substrate obtained in the obtaining step.

[0063] Furthermore, in the color prediction method of the present embodiment, the obtaining step may further obtain a thickness of a coating film formed using the composition, and the calculating step may calculate a predicted spectral reflectance of the coating film made of the composition at the thickness based on the absorption coefficient and scattering coefficient predicted in the predicting step and the thickness obtained in the obtaining step.

[0064] Note that the specific aspects of the method of this embodiment have been described above in the control process, so a detailed description thereof will be omitted here.

[0065] 3. Method for Producing Composition The method for producing a composition of the present embodiment includes a proposing step of proposing a composition formula based on the absorption coefficient and scattering coefficient predicted by the above-described color tone prediction method, and a preparation step of preparing a composition based on the proposed composition formula.

[0066] Specifically, the proposing step calculates predicted spectral reflectances of a coating film made of one or more compositions based on the absorption coefficients and scattering coefficients of the compositions having one or more formulations predicted by the color prediction method, and proposes a composition formulation having a spectral reflectance corresponding to the user's desired color tone. The preparing step prepares a composition by blending each component based on the composition formulation proposed in the proposing step. Furthermore, for example, the proposing step may repeat the obtaining step, the predicting step, and the calculating step while changing the composition formulation, identify a composition formulation in which the difference between the predicted spectral reflectance and the spectral reflectance corresponding to the user's desired color tone is equal to or less than a certain value, and propose the composition formulation.

[0067] Alternatively, the proposing step may be performed as follows. For example, in the proposing step, desired values ​​of absorption coefficients and scattering coefficients are received from a user. Then, the color prediction device predicts the absorption coefficients and scattering coefficients for any generated composition formula using the color prediction method. It then determines whether the difference between the predicted absorption coefficients and scattering coefficients and the absorption coefficients and scattering coefficients desired by the user is equal to or less than a certain value. If the difference is not equal to or less than the certain value, the composition formula is regenerated, and this process is repeated to search for a composition formula that matches the absorption coefficients and scattering coefficients desired by the user. The searched composition formula may then be proposed.

[0068] Note that, since the specific aspects of the manufacturing method of this embodiment have been described above in the control process, detailed description thereof will be omitted here.

[0069] 4. Color Tone Prediction Program The color tone prediction program of this embodiment causes the color tone prediction device to execute an acquisition step of acquiring a composition formula, and a prediction step of predicting the absorption coefficient and scattering coefficient of the composition based on the composition formula.

[0070] Furthermore, the color prediction program of this embodiment may cause the acquisition step to further acquire the spectral reflectance of a substrate to which the composition is applied, and cause the color prediction device to execute a calculation step to calculate a predicted spectral reflectance of a coating film made of the composition to be formed on the substrate, based on the absorption coefficient and scattering coefficient predicted in the prediction step and the spectral reflectance of the substrate acquired in the acquisition step.

[0071] In the color prediction program of this embodiment, the obtaining step may further obtain a thickness of a coating film formed using the composition, and the calculating step may calculate a predicted spectral reflectance at the thickness of the coating film made of the composition, based on the absorption coefficient and scattering coefficient predicted in the predicting step and the thickness obtained in the obtaining step.

[0072] The program may be recorded on a readable recording medium. Note that the specific aspects of the processing executed by the program of this embodiment have been described in the control processing section above, and therefore will not be described in detail here.

[0073] The present invention will be described in more detail below using examples.

[0074] 1. Example 1 A green solder resist composition of Formula A containing two coloring materials, a blue pigment and a yellow pigment, was prepared as a composition sample. A total of 25 composition samples of Formulas A to Y were prepared by changing only the amount of pigment used from Formula A. Furthermore, the composition of each formula was used to apply the composition sample of each formula to a buffed copper plate so that the film thickness was 20 μm, 30 μm, or 40 μm, and then dried, to prepare a total of 75 coating film samples.

[0075] The spectral reflectance Rm of these coating samples was then measured, and the absorption coefficient Km and scattering coefficient Sm of each composition sample of formulas A to Y were calculated. Of the absorption coefficient Km and scattering coefficient Sm of the composition sample for each formula obtained, 80% were assigned to training data and 20% to validation data. A prediction model was created from the training data, and the prediction accuracy of the validation data was compared. The prediction accuracy was determined based on the average color difference ΔE between the predicted color tone and the actual color tone, and the prediction accuracy was judged to be high or low.

[0076] On the other hand, for comparison, similar 75 coating samples were prepared, and data sets of their corresponding formulations, film thicknesses, and spectral reflectances were created. The data for formulations A to T, which correspond to 80%, were assigned as training data, and the data for formulations U to Y, which correspond to 20%, were assigned as validation data. A model was created to predict the spectral reflectance from the formulation using a linear regression method based on the training data, and the prediction accuracy of the validation data was compared. The results are shown below.

[0077]

[0078] 2. Example 2 A green solder resist composition containing two coloring materials, a blue pigment and a yellow pigment, was prepared. Specifically, seven formulations with different combinations of filler and epoxy resin were prepared, and for each of these, the blue pigment and yellow pigment contents were varied to high, medium, and low loadings, resulting in a total of 21 composition samples. Each composition sample was then applied to a copper plate to film thicknesses of 20 μm, 30 μm, and 40 μm, followed by drying, to produce a total of 63 coating film samples.

[0079] Using the prepared coating film samples, a prediction model was prepared and the prediction accuracy was compared in the same manner as in Example 1. In Example 2, of the absorption coefficients Km and scattering coefficients Sm of the composition samples obtained for each formulation, 42 were used as training data and 21 were used as validation data to prepare a prediction model. The results are shown below.

[0080]

[0081] The above-described Examples 1 and 2 demonstrated that the prediction model of this embodiment exhibits higher prediction accuracy with the same amount of training data than a linear regression model created using recipes and their corresponding spectral reflectances as training data. It was also found that the prediction model exhibits high prediction accuracy even when the composition other than the colorant is changed.

[0082] 1...Color prediction system, 100...Color prediction device, 100...Server, 110...Processor, 111...Estimation unit, 112...Learning unit, 113...Acquisition unit, 114...Prediction unit, 115...Calculation unit, 116...Proposal unit, 120...Communication interface, 130...Input / output interface, 140...Memory, 150...Storage, 151...Learning data, 160...Communication bus, 200...User device

Claims

1. A color tone prediction device comprising: an acquisition unit that acquires a formulation of a composition; and a prediction unit that predicts an absorption coefficient and a scattering coefficient of the composition based on the formulation of the composition.

2. The color tone prediction device according to claim 1, further comprising a calculation unit that: the acquisition unit further acquires a spectral reflectance of a substrate on which the composition is applied; and calculates a predicted spectral reflectance of a coating film formed of the composition on the substrate based on the absorption coefficient and the scattering coefficient predicted by the prediction unit and the spectral reflectance of the substrate acquired by the acquisition unit.

3. The color tone prediction device according to claim 1, further comprising a calculation unit that: the acquisition unit further acquires a film thickness of a coating film formed using the composition; and calculates a predicted spectral reflectance at the film thickness of the coating film formed of the composition based on the absorption coefficient and the scattering coefficient predicted by the prediction unit and the film thickness acquired by the acquisition unit.

4. An estimation unit that estimates an absorption coefficient and a scattering coefficient in a composition sample having a predetermined formulation based on spectral reflectances at a plurality of different film thicknesses of a coating film sample formed from the composition sample having the predetermined formulation and the film thicknesses thereof; and a learning unit that creates a prediction model for predicting an absorption coefficient and a scattering coefficient of a composition from a formulation of the composition based on the absorption coefficient and the scattering coefficient estimated by the estimation unit and the formulation for a plurality of composition samples having different formulations. The prediction unit predicts the absorption coefficient and the scattering coefficient of the composition based on the prediction model. The color tone prediction device according to claim 1.

5. The color tone prediction device according to claim 1, wherein the composition is a solder resist.

6. A color tone prediction method including: an acquisition step of acquiring a formulation of a composition; and a prediction step of predicting an absorption coefficient and a scattering coefficient of the composition based on the formulation of the composition.

7. An estimation step of estimating an absorption coefficient and a scattering coefficient in a composition sample having a predetermined formulation based on spectral reflectances at a plurality of different film thicknesses and the film thicknesses of a coating film sample formed from the composition sample having the predetermined formulation; and a learning step of creating a prediction model for predicting an absorption coefficient and a scattering coefficient of a composition from a formulation of the composition based on the absorption coefficient and the scattering coefficient estimated in the estimation step and the formulation of a plurality of composition samples having different formulations. The prediction step predicts the absorption coefficient and the scattering coefficient of the composition based on the prediction model. The color tone prediction method according to claim 6.

8. A proposal step of proposing a formulation of a composition based on the absorption coefficient and the scattering coefficient of the composition predicted by the color tone prediction device according to any one of claims 1 to 5 or the color tone prediction method according to claim 6 or 7; and a preparation step of preparing a composition based on the formulation of the composition proposed in the proposal step. A method for manufacturing a composition.

9. A color tone prediction program that causes a color tone prediction device to execute an acquisition step of acquiring a formulation of a composition and a prediction step of predicting an absorption coefficient and a scattering coefficient of the composition based on the formulation of the composition.

10. A color tone prediction program that further causes a color tone prediction device to execute an estimation step of estimating an absorption coefficient and a scattering coefficient in a composition sample having a predetermined formulation based on spectral reflectances at a plurality of different film thicknesses and the film thicknesses of a coating film sample formed from the composition sample having the predetermined formulation, and a learning step of creating a prediction model for predicting an absorption coefficient and a scattering coefficient of a composition from a formulation of the composition based on the absorption coefficient and the scattering coefficient estimated in the estimation step and the formulation of a plurality of composition samples having different formulations. The prediction step predicts the absorption coefficient and the scattering coefficient of the composition based on the prediction model. The color tone prediction program according to claim 9.

Images