Robust color formula determination method and system based on visual perception

Abstract

The invention discloses a robust color formula determination method and system based on visual perception. The method comprises the following steps: receiving target color standard data; judging whether the target color falls into a preset sensitive color system area or not; if the target color belongs to the sensitive color system, determining a reference color different from the target color in a local color difference permissible area meeting a preset color difference permissible threshold condition; respectively determining candidate color formulas corresponding to the target color and the reference color; and evaluating each candidate formula based on a preset production stability evaluation index, and selecting the formula with the optimal index as a production formula. According to the method and the device, the formula solving range is expanded from a single point to a region by utilizing the human eye vision latitude based on the characteristic of visual perception dynamic change of human eyes on colors, and the formula with higher production fluctuation tolerance and higher robustness is screened out on the premise of ensuring the vision consistency, so that the color difference risk caused by the production fluctuation is effectively reduced, and the product quality is improved. And the product qualification rate is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial color matching, and in particular to a robust color formula determination method and system based on visual perception. BACKGROUND

[0002] In the industrial production of textile dyeing, printing, coating, plastics, etc., the commonly used color matching process is usually as follows: the spectral data or colorimetric data of the target color (standard sample) provided by the customer is measured, and then a computer algorithm is used to search or calculate the formula that can match the target color in the database.

[0003] However, the inventors found that the existing color matching method has some problems in practical application when implementing the present application. At present, the design end often sets only a single fixed color difference tolerance (for example, a fixed numerical value) as the judgment basis for proofing and mass production when providing the target color. However, in practice, the visual perception of the human eye to color difference is not a static constant, but shows dynamic changes with different color regions, lightness, chroma, and hue.

[0004] Some colors, the human eye can be very sensitive to extremely small hue deviations. If the dynamic characteristics of the human eye are ignored, and only a single fixed standard or the center coordinates of the target color are used as the basis for calculating the formula, the problem of "production instability" often arises. That is, due to small fluctuations in the production process, such as small errors in dyeing pigment weighing, slight changes in temperature, small cumulative errors are easily magnified in the actual mass production process, resulting in the color of the final product deviating from the target color. And because the subtle differences in such colors are easily noticeable by the naked eye, even if they mathematically meet the fixed color difference tolerance, they are not acceptable in terms of vision, resulting in high product rejection rates and high color correction costs. SUMMARY

[0005] The purpose of the present application is to provide a robust color formula determination method and system based on visual perception for sensitive colors, which actively seeks a production fault tolerance higher formula within the visual perception range of the human eye.

[0006] To achieve the above-mentioned purpose of the application, an embodiment of the present application provides a robust color formula determination method based on visual perception, comprising the following steps: receiving standard data of a target color; determining whether the target color falls into a preset sensitive color system region according to the standard data, wherein the sensitive color system region is determined based on the visual recognition threshold characteristics of the color space, and the region has a higher sensitivity to color difference changes than a preset level; if the determination result is yes, determining at least one reference color different from the target color in a local color difference tolerance region, wherein the local color difference tolerance region is a region centered on the target color and having a color difference value from the target color satisfying a preset color difference tolerance threshold condition; determining candidate color formulas corresponding to the target color and the reference color; obtaining a preset production stability evaluation index, and evaluating each candidate color formula; selecting a candidate color formula with the optimal production stability evaluation index as the production formula of the target color.

[0007] As a further improvement of the present application, the determining whether the target color falls into a preset sensitive color system region according to the standard data comprises: calculating an area of a local color difference tolerance region corresponding to the target color; determining whether the area is less than or equal to a preset area threshold; if yes, determining that the target color falls into the preset sensitive color system region; if no, determining that the target color does not fall into the preset sensitive color system region.

[0008] As a further improvement of the present application, the determining whether the target color falls into a preset sensitive color system region according to the standard data comprises: determining a chroma value, an L value, an a value and a b value of the target color in a CIELAB color space according to the standard data; determining whether the chroma value, the L value, the a value and the b value all fall into a preset numerical interval corresponding to each parameter; if yes, determining that the target color falls into the preset sensitive color system region; if no, determining that the target color does not fall into the preset sensitive color system region.

[0009] As a further improvement of the present application, the local color difference tolerance region is a MacAdam circle.

[0010] As a further improvement of the present application, the determining at least one reference color different from the target color comprises: determining a long-axis endpoint coordinate and a short-axis endpoint coordinate of the MacAdam circle; determining at least one of the long-axis endpoint coordinate and the short-axis endpoint coordinate as the reference color.

[0011] As a further improvement of the present application, the production stability evaluation index comprises a process deviation degree index. The evaluating each of the candidate color formulas comprises: determining a theoretical coloring proportion of each colorant in each of the candidate color formulas; obtaining a statistical process reference proportion of each of the colorants by the production equipment; calculating an absolute value of a difference between the theoretical coloring proportion and the statistical process reference proportion, wherein the smaller the absolute value of the difference, the higher the score of the process deviation degree index.

[0012] As a further improvement of the present application, the production stability evaluation index comprises a raw material compatibility index. The evaluating each of the candidate color formulas comprises: detecting whether a preset repelling colorant combination exists in the candidate color formula; if yes, reducing the score of the raw material compatibility index corresponding to the candidate color formula.

[0013] As a further improvement of the present application, the production stability evaluation index comprises a tolerance performance index, wherein the tolerance performance index comprises at least one of heat resistance, light resistance, weather resistance, or migration resistance. The evaluating each of the candidate color formulas comprises: obtaining a tolerance performance parameter corresponding to the tolerance performance index for each of the candidate color formulas; The evaluating each of the candidate color formulas further comprises: calculating an evaluation total score corresponding to each of the candidate color formulas, the evaluation total score being a weighted sum result of the process deviation degree index, the raw material compatibility index, and the tolerance performance index respectively weighted by respective weights, wherein the higher the evaluation total score, the better the production stability evaluation index.

[0014] As a further improvement of the present application, further comprising the steps of: if the target color does not fall into the preset sensitive color system region, making a production formula corresponding to the target color.

[0015] To achieve one of the above-mentioned purposes, an embodiment of the present application provides a robust color formula determination system based on visual perception, comprising: a data receiving module configured to receive standard data of a target color; a sensitivity determination module configured to determine, according to the standard data, whether the target color falls into a preset sensitive color system region, wherein the sensitive color system region is a region in which the sensitivity of human eyes to color difference changes is higher than a preset level, and is determined based on the visual recognition threshold characteristics of a color space. The reference color determination module is configured to determine at least one reference color different from the target color in a local color difference tolerance region when it is determined that the target color falls into the sensitive color system region, wherein the local color difference tolerance region is a region centered on the target color and having a color difference value from the target color satisfying a preset color difference tolerance threshold condition; The formula calculation module is configured to determine candidate color formulas corresponding to the target color and the reference color. The formula optimization module is configured to obtain a preset production stability evaluation index, evaluate each candidate color formula, and select a candidate color formula with the optimal production stability evaluation index as the production formula of the target color.

[0016] Compared with the conventional technology, the robust color formula determination method and system based on visual perception have the following beneficial effects: the robust color formula determination method and system based on visual perception break through the limitation of a single fixed color difference tolerance value based on the dynamic change characteristics of visual perception of the human eye on color. For the target color determined as a sensitive color, the formula corresponding to different reference points in the region is calculated and evaluated, the visual tolerance of the human eye is utilized, the solution range is expanded from a single "point" to a "region", and the formula corresponding to different reference points in the region is calculated and evaluated, and those formulas that are more stable in production and have stronger robustness in the visual tolerance range are found, so that the finally selected formula not only meets the color difference standard, but also has stable process. Under the premise of ensuring visual consistency, the tolerance of the formula to production fluctuation is simultaneously improved, the problem that the color difference caused by process fluctuation is detected by the human eye in actual production is significantly reduced, the product qualification rate is improved, and the production efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a flowchart of a robust color formula determination method based on visual perception according to an embodiment of the present application; Figure 2 is a schematic diagram of a color space according to an embodiment of the present application; Figure 3 is a flowchart of one of the embodiments of the method for determining whether the target color falls into the preset sensitive color system region according to an embodiment of the present application; Figure 4 is a flowchart of another embodiment of the method for determining whether the target color falls into the preset sensitive color system region according to an embodiment of the present application; Figure 5 is a schematic diagram of the framework of a robust color formula determination system based on visual perception according to an embodiment of the present application. DETAILED DESCRIPTION

[0018] The present application will now be described in detail with reference to the specific embodiments shown in the accompanying drawings. However, these embodiments do not limit the present application, and any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the scope of protection of this application.

[0019] One embodiment of this application provides a method and system for determining robust color formulations based on visual perception that actively seeks out colors within the range permissible by human vision and with a higher production error tolerance for sensitive colors.

[0020] The following sections will describe the robust color formulation determination method and system based on visual perception.

[0021] Example 1: A robust color formulation determination method based on visual perception The robust color formulation determination method based on visual perception in this embodiment aims to solve the problems of poor production stability and high defect rate caused by relying solely on a single standard color to calculate formulations for visually sensitive colors in existing color matching technologies. This method introduces a visual perception model to proactively seek formulations with higher production tolerance within a range of color differences imperceptible to the human eye.

[0022] The following is combined Figures 1-4 This application describes a method for determining a robust color formula based on visual perception, provided in one embodiment. Although this application provides method operation steps as shown in the following embodiments or flowcharts, the execution order of these steps is not limited to the execution order provided in the embodiments of this application, as they are steps that do not have a necessary causal relationship in logic and are based on conventional or non-creative labor.

[0023] This embodiment presents a robust color formulation determination method based on visual perception, such as... Figure 1 As shown, it includes the following steps: Step S10: Receive standard data for the target color.

[0024] Step S20: Based on the standard data, determine whether the target color falls into a preset sensitive color system area, wherein the sensitive color system area is determined based on the visual recognition threshold characteristics of the color space, and the human eye is more sensitive to color difference changes than a preset level.

[0025] Step S30: If the determination result is yes, within the local color difference tolerance area, determine at least one reference color that is different from the target color, wherein the local color difference tolerance area is the area centered on the target color and the color difference value between the color and the target color satisfies the preset color difference tolerance threshold condition.

[0026] Step S40: Determine the candidate color formulas corresponding to the target color and the reference color.

[0027] Step S50: Obtain preset production stability evaluation indicators and evaluate each of the candidate color formulations.

[0028] Step S60: Select the candidate color formula with the best production stability evaluation index as the production formula for the target color.

[0029] In step S10, the digitized standard data of the target color specified by the client or designer is first acquired. This standard data is typically obtained by a spectrophotometer under specific light sources (such as D65, TL84, etc.) and viewing angles (such as 10° or 2°). The standard data can be in the form of spectral reflectance data or converted chromaticity coordinate data, such as CIE Lab values, CIE XYZ values, or CIE LCh values.

[0030] Since not all colors require complex tolerance analysis, step S20 sets up a filtering mechanism - a preset "sensitive color area", which is determined based on the visual recognition threshold characteristics in the color space, such as according to the McAdam ellipse theory or experimental data on human visual sensitivity.

[0031] like Figure 2 The diagram illustrates the CIE 1931 XYZ color space and the distribution of MacAdam ellipses within it. The horseshoe-shaped boundary in the diagram represents the trajectory of the visible spectrum, and the black elliptical areas in the diagram are the MacAdam ellipses. Each ellipse represents a color range where the human eye cannot visually distinguish the difference when the target color is at the center of that ellipse; this is the local color difference tolerance area described in this application. A larger ellipse area indicates a higher tolerance (low sensitivity) for color changes in that area by the human eye, while a smaller ellipse means that even extremely small coordinate changes (color differences) within that area can be perceived by the human eye. The areas containing these small ellipses correspond to the sensitive color system areas described in this application.

[0032] Based on the physical fact of these characteristics of human vision, this application proposes to prioritize determining whether the target color falls within these small elliptical (sensitive color) regions before determining the formula. For colors that fall within this region, the optimization algorithm based on the local color difference tolerance region proposed in this application must be used to avoid the risk of the color exceeding the range of this narrow ellipse due to production fluctuations.

[0033] Colors within the sensitive color range are typically low-saturation neutral colors such as gray, beige, and pastels. The human eye is extremely sensitive to minute changes in hue or saturation, meaning it is highly sensitive to color differences. As described in the background section, due to production instability—that is, minute fluctuations in the production process, such as slight errors in pigment weighing or minor temperature variations—the final product's color may deviate from the target color. In such cases, for sensitive colors, the difference is more likely to be perceived as excessive by the user.

[0034] Step S20 implements targeted tiered processing for different colors. For non-sensitive colors, the conventional process can be used to improve efficiency; while for "high-risk" colors that fall into the sensitive color range, the subsequent optimization process is initiated. This ensures special consideration for "difficult-to-process colors" due to visual characteristics, while avoiding the waste of computing power caused by a one-size-fits-all approach to all colors.

[0035] When the target color in step S20 is determined to be a sensitive color, step S30, based on a visual perception model, determines a "local color difference tolerance region" centered on the target color. Mathematically, the local color difference tolerance region represents a set where the color difference (ΔE) between any color within the set and the central target color satisfies a preset color difference tolerance threshold condition (e.g., ΔE < 0.8 or ΔE < 1.0). This means that within this region, the human eye can hardly distinguish the color difference. Step S30 selects one or more coordinate points as "reference colors" within or on the boundary of the local color difference tolerance region. These reference colors are mathematically distinct from the target color (e.g., located at the edge of the tolerance region), but are considered acceptable in visual perception, or in other words, the difference is imperceptible to the naked eye.

[0036] Traditional methods only pursue mathematical "zero color difference," often resulting in formulations being at process-sensitive points. Step S30 trades space for stability by defining a local color difference tolerance area, expanding the solution space from a "point" to a "domain," selecting a reference color different from the center, and exploring whether there exists a better-produced color coordinate in a place slightly off-center (but within the range imperceptible to the human eye).

[0037] Step S40 uses computer color matching algorithms, such as the Kubelka-Munk theory, to calculate multiple sets of candidate color formulas. That is, the candidate color formulas not only include color formulas that theoretically perfectly match the target color, but also formulas that match those located within the local color difference tolerance area.

[0038] Step S40 greatly expands the range of color formulation options. Typically, to get closer to the center point, it may be necessary to use some unstable combinations of dyes or extreme concentration ratios. However, for the formulation of the reference color, a completely different combination of dyes or a more reasonable concentration ratio may be used, thus providing a basis for finding a "more robust color formulation" in the future.

[0039] Step S50 introduces a second evaluation dimension independent of "color difference"—"production stability," which pre-defines a set of indicators (production stability evaluation indicators) to measure the robustness of the formulation in actual industrial production. Step S50 performs simulation analysis or data comparison on each candidate formulation generated in step S40 to calculate its stability score. This evaluation process does not focus on whether the formulation is "accurate" (because step S30 has already ensured that it is within the visually acceptable range), but rather on whether the formulation is "stable" and "easy to produce."

[0040] By introducing a stability assessment, step S50 can identify formulations with minimal color difference but high process risk, such as those that are extremely sensitive to temperature or have poor dye compatibility, and deduct points from them; at the same time, it can identify formulations with slightly larger color difference (but acceptable) but extremely stable in production and add points to them.

[0041] Based on the evaluation results of step S50, step S60 sorts or filters all candidate formulations and finally selects the formulation with the best stability index to output to the production end.

[0042] The final selected formula strikes a balance between "visual indistinguishability" and "maximum production stability." It may correspond to a reference color rather than the target color center, but due to its optimal production stability, this formula can resist interference from equipment fluctuations and human error in actual mass production, maintaining color consistency and significantly reducing the rate of color repetition and defective products. This represents a leap from "theoretical color matching" to "engineering color matching."

[0043] Given that once the spinning process is complete, it is extremely difficult to correct the color of spun yarn products (i.e., irreversible), the method of "predicting stability and avoiding risks before production" proposed in this application has higher application value and economic benefits for the spun yarn industry than traditional dyeing and printing.

[0044] The following describes two implementation methods to illustrate different technical approaches for determining the "sensitive color system" in step S20: In one implementation, based on geometric features (area method), the judgment is made by quantifying the "magnitude of human eye's resolving power" from the geometric properties of colorimetry. Specifically, such as... Figure 3 As shown, step S20 includes: Step S21: Calculate the area of ​​the local color difference tolerance region corresponding to the target color.

[0045] Step S22: Determine whether the area of ​​the region is less than or equal to a preset area threshold.

[0046] Step S221: If yes, then determine that the target color falls into the preset sensitive color system area.

[0047] Step S222: If not, it is determined that the target color does not fall into the preset sensitive color system area.

[0048] Step S21, based on a preset color difference formula, such as CIE2000 or CMC2:1, constructs a local color difference tolerance region (typically represented as a McAdam ellipse or its approximate geometry) centered on the received target color. The local color difference tolerance region represents the set of all colors whose visual color difference from the center point is less than a specific unit (e.g., 1.0 JND). Subsequently, the area of ​​this geometric region is calculated mathematically.

[0049] For example, if the local color difference tolerance area is approximately an ellipse, the system will use differential calculation or eigenvalue decomposition of the color difference formula to find the major semi-axis (a) and minor semi-axis (b) of the ellipse, and then use the formula Area=π×a×b to calculate the precise area value of the region.

[0050] The system internally stores a preset area threshold, which represents the boundary between "high sensitivity" and "normal sensitivity". Step S22 compares the area of ​​the region calculated in step S21 with this preset area threshold.

[0051] If the area of ​​the region is less than or equal to a preset area threshold, it means that the "indistinguishable range" around the target color is very narrow. In other words, the color is in an area where human vision is extremely sensitive, and even a very slight color shift in the formula will exceed this narrow elliptical range and be perceived by the naked eye. Therefore, the system determines that the target color falls into a preset sensitive color area.

[0052] If the area of ​​the target color is larger than a preset area threshold, it means that the "indistinguishable range" around the target color is relatively wide, and the human eye has a high tolerance for that color. Therefore, the system determines that the target color does not fall within the preset sensitive color system area.

[0053] Steps S21-S22 use "region area" as the judgment criterion, which is highly scientific and universally applicable. It directly utilizes the core physical meaning of the MacAdam ellipse theory—the smaller the ellipse, the more sensitive the human eye is. This method does not rely on the specific hue of the color (red, yellow, blue, etc.), but is purely based on geometric quantification indicators of visual perception. In this way, this method can accurately identify those "hidden killer" colors that are easily overlooked in traditional production but actually have extremely low error tolerance, ensuring that subsequent formula optimization processes are only initiated for the truly needed sensitive colors, achieving precise allocation of computing resources.

[0054] In another implementation, based on coordinate parameters (numerical range method), and also on industrial big data and the experience of color experts, a rapid judgment is made by directly checking whether the color coordinates fall into a specific high-risk area. Specifically, such as... Figure 4 As shown, step S20 includes: Step S23: Determine the chroma value, L value, a value, and b value of the target color in the CIELAB color space based on the standard data.

[0055] Step S24: Determine whether the chroma value, the L value, the a value, and the b value all fall within the preset value range corresponding to each parameter.

[0056] Step S241: If yes, then determine that the target color falls into the preset sensitive color system area.

[0057] Step S242: If not, it is determined that the target color does not fall into the preset sensitive color system area.

[0058] Step S23 parses the received target color data. If the original data is spectral reflectance, the system converts it into coordinate values ​​in the CIELAB color space using standard colorimetry calculations: luminance (L), red-green axis (a), and yellow-blue axis (b). Simultaneously, according to the formula C = ... Calculate the chroma value of this color.

[0059] The system pre-defines a set of parameter boundaries for "sensitive color regions." These boundaries are determined based on a large amount of historical color matching data and visual evaluation experiments.

[0060] Step S24 checks the chroma value, for example, whether the chroma value is less than or equal to 12; and checks whether the L, a, and b values ​​are within the preset value range, for example, whether the L value is between 30 and 90, the a value is between 0 and 11.4, and the b value is between 2.3 and 11.6.

[0061] If (all conditions are met): that is, all parameters of the color meet all the above conditions, it means that the color belongs to a typical low-saturation neutral color (such as gray, beige, etc.) and is in the range where the human eye is extremely sensitive to color deviation. Step S241 determines that the target color falls into the preset sensitive color system area.

[0062] If not (any parameter does not match): for example, if the chroma is very high (vibrant color) or the L value is extremely low (extreme black), it means that the color is not in the high-risk area, and step S242 determines that the target color does not fall into the preset sensitive color system area.

[0063] The judgment method based on the "parameter range" in steps S23-S24 has extremely high computational efficiency and engineering practical value. Unlike steps S21-S22, this implementation method can complete the judgment through simple numerical comparison, enabling faster processing of large amounts of order data. Furthermore, this method transforms the fuzzy empirical knowledge of "difficulty in dyeing gray" and "prone to color deviation in neutral colors" in the textile dyeing and printing industries into precise digital rules. It specifically targets the low-saturation areas (gray and earth tones) with the highest defect rates in industrial production, providing focused defense and demonstrating strong practical relevance.

[0064] Other technical approaches for determining "sensitive color systems" in step S20 can be based on "color difference change rate (gradient)" or other color spaces / datasets. For example, steps S23-S24 above are based on Pantone's empirical data in the CIELAB color space, and can also be extended to "based on DIN99 color space" or "based on CAM02-UCS color appearance model" for low saturation areas.

[0065] In one embodiment, the local color difference tolerance area is a McAdam circle.

[0066] The MacAdam circle is typically elliptical, hence it can also be called a MacAdam ellipse. However, in a uniform color space, the MacAdam circle can also be a perfect circle. The following explanation of the permissible area for local color difference will use the MacAdam ellipse as an example.

[0067] When constructing the local color difference tolerance area centered on the target color, this method does not simply use a geometric circle (assuming uniform color difference in all directions) or a rectangle (assuming independent components), but instead constructs a McAdam ellipse.

[0068] Specifically, this method utilizes a pre-defined color difference formula (preferably the CIE2000 color difference formula or the CMC2:1 formula, which is closest to human visual perception) for inverse solving or numerical fitting. On a two-dimensional chromaticity plane or a three-dimensional color space section, the boundary of the McAdam ellipse represents the critical line where the human eye can no longer distinguish the difference near the target color. The shape, size, and tilt angle of this ellipse dynamically change depending on the position of the target color in the color space: in some color areas, it may exhibit tolerance for saturation changes through its major axis, while in other areas it may exhibit tolerance for hue changes.

[0069] Because human color perception is non-uniform, the sensitivity and direction of color changes vary drastically at different locations in the color space. Traditional box or circle discrepancies often include colors whose differences are actually perceptible to the human eye (leading to misjudgments) or exclude colors whose differences are not actually perceptible to the human eye (resulting in wasted production windows). This embodiment, by employing McAdam ellipses, can most accurately simulate the visual "blind spot" of the human eye, ensuring that the reference formula found within this area is visually absolutely safe, maximizing the "visually indistinguishable" solution space, and providing the most accurate geometric boundary basis for subsequently finding more stable production formulas.

[0070] In one embodiment, step S30 includes: Step S31: Determine the coordinates of the major axis endpoint and the minor axis endpoint of the McAdam loop.

[0071] Step S32: Determine at least one of the major axis endpoint coordinates and the minor axis endpoint coordinates as the reference color.

[0072] Step S31 analyzes the mathematical parameters describing the McAdam ellipse, which involves eigenvalue decomposition of the local color difference metric matrix. Specifically, the second derivative matrix (Hessian matrix or metric tensor) of the color difference function at the target color is calculated. Through eigenvalue decomposition, two eigenvalues ​​and corresponding eigenvectors are obtained. The direction of the eigenvector corresponding to the larger eigenvalue is the direction of the major axis of the ellipse (representing the direction with the largest visual tolerance), and the direction corresponding to the smaller eigenvalue is the direction of the minor axis (representing the direction with the smallest visual tolerance).

[0073] Step S31 combines the preset color difference threshold to calculate the semi-major axis length (a) and semi-minor axis length (b), and based on the center coordinates (x0, y0) of the target color, extends a and b along the major axis and minor axis directions respectively to calculate the coordinates of four extreme points on the ellipse boundary: the coordinates of the two major axis endpoints and the coordinates of the two minor axis endpoints.

[0074] Step S32 inputs the four endpoint coordinates (or one, two, or three of the four endpoint coordinates, or only the two at the ends of the major axis) obtained above as "reference colors" into the recipe calculation module. These points represent the maximum limit positions at which the color can deviate from the center in the hue, saturation, or brightness direction within the visually permissible range.

[0075] Selecting the endpoints of the major and minor axes as reference colors offers two advantages. First, it provides strong representativeness: the major and minor axes represent the two extreme directions of color visual perception within the local area (usually corresponding to extreme changes in hue or saturation). If the formula in these two extreme directions meets production stability requirements, it usually means that other directions within the area can also achieve good balance. Second, it offers high computational efficiency: compared to performing massive formula calculations by randomly sampling multiple points on the elliptical boundary, selecting only four feature endpoints (or even just two major axis endpoints) can reduce the computational load by an order of magnitude while covering the maximum span of the tolerance region. This significantly improves the system's computational speed while ensuring the discovery of robust formulas.

[0076] In one embodiment, the production stability evaluation index includes at least one of the following: process deviation index, raw material compatibility index, and tolerance performance index. The process deviation index, raw material compatibility index, and tolerance performance index correspond to the three dimensions of process matching degree, raw material compatibility, and physicochemical tolerance, respectively, thereby quantifying the production stability of the color formula.

[0077] The process deviation index assesses the ease of color formulation from the perspective of "equipment compatibility." Step S50 includes: Step S51: Determine the theoretical coloring ratio of each pigment in each candidate color formula.

[0078] Step S52: Obtain the statistical process baseline ratio of each pigment for the production equipment.

[0079] Step S53: Calculate the absolute value of the difference between the theoretical coloring ratio and the statistical process baseline ratio, wherein the smaller the absolute value of the difference, the higher the score of the process deviation index.

[0080] For each candidate formulation calculated in step S40, step S51 analyzes its formulation composition, clarifying the theoretical addition amount or concentration percentage of each colorant (such as dye, pigment, or masterbatch) required in the formulation. For example, a formulation requires the addition ratio of "red masterbatch A" to be 0.53%.

[0081] This system connects to the factory's MES (Manufacturing Execution System) or SPC (Statistical Process Control) database, which records the actual output data of specific production equipment (such as injection molding machines, spinning machines, or dyeing vats) for various pigments over a long historical production period. Step S52 extracts the Statistical Process Baseline (SPC) ratio for the aforementioned "red masterbatch A". This baseline ratio represents the habitual operating ratio of the equipment under the most stable condition or when the yield is highest. For example, historical data shows that the equipment operates most stably and has the highest control precision when the addition amount is 0.50%.

[0082] In step S53, the absolute value of the difference = |theoretical ratio - SPC benchmark ratio|. The system assigns a score to the process deviation index based on this absolute value of the difference. A smaller difference (closer to 0) indicates that the formula's requirements fall precisely within the equipment's "comfort zone," requiring no extreme adjustments to equipment parameters during production, resulting in extremely low process fluctuation risk. Therefore, the higher the score for the candidate formula's process deviation index. A larger difference indicates that the formula's requirements deviate significantly from the equipment's optimal operating conditions, making inaccurate feeding or uneven mixing more likely during production. Therefore, the lower the score.

[0083] By introducing a process deviation index, this embodiment deeply integrates "formula design" with "production capacity." Traditional color matching only focuses on "accuracy" without considering "feasibility." This method, however, prioritizes formulas that match the factory's equipment production capacity, reducing batch-to-batch color differences caused by equipment precision limitations or improper process parameter adjustments from the outset.

[0084] Raw material compatibility indicators mitigate potential formulation risks from the perspective of "physicochemical interactions." Step S50 includes: Step S54: Detect whether there is a preset incompatible colorant combination in the candidate color formula.

[0085] Step S541: If so, reduce the score of the raw material compatibility index corresponding to the candidate color formulation.

[0086] Step S542: If not, the raw material compatibility index corresponding to the candidate color formula is scored as full marks.

[0087] The system has a built-in "incompatibility rule base" or "compatibility database". Step S54 iterates through all the ingredients in each candidate formulation and checks for the following typical high-risk combinations: Adsorption differences: For example, check if the formulation contains both "carbon black pigments" and "low-adsorption colorants". Because carbon black has an extremely high specific surface area and adsorption capacity, it easily adsorbs and coats other trace color powders during melting or mixing, causing other colorants to "fail to develop color" or "be overtaken", resulting in a dark or unstable final color.

[0088] Spinning pressure risk combination: For chemical fiber spinning scenarios, check whether it contains a combination of "phthalocyanine pigments" (such as phthalocyanine blue PB15, phthalocyanine green PG7) and "organic yellow / orange / red pigments". Experience data shows that this combination is very prone to agglomeration under high temperature and pressure, which can cause the filter screen of the spinning component to become clogged, resulting in an abnormal increase in the pressure of the spinning component, causing yarn breakage or color fluctuation.

[0089] Density difference combination: Check if the formula contains "inorganic pigments" (high density) and "organic pigments" (low density) with huge density differences. Such combinations are very prone to separation during processing due to centrifugal force or gravity, resulting in uneven color.

[0090] If the above high-risk combinations exist, step S541 will trigger a penalty mechanism, significantly reducing the compatibility score of the recipe, or in extreme cases, directly marking the recipe as "unusable" and removing it.

[0091] Many formulations show no problems in the laboratory sample stage, but fail when implemented on a large production line due to physicochemical reasons such as adsorption, agglomeration, and stratification. The raw material compatibility index, through a pre-set compatibility testing mechanism, is equivalent to a "virtual trial and error" during the calculation stage, eliminating high-risk formulations in advance and greatly avoiding expensive trial costs and waste generation.

[0092] The resistance performance index includes at least one of heat resistance, light resistance, weather resistance, or migration resistance, and step S50 includes: Step S55: Obtain the tolerance performance parameters corresponding to the tolerance performance index for each of the candidate color formulations.

[0093] Step S56: Calculate the total evaluation score for each of the candidate color formulations. The total evaluation score is the weighted sum of the process deviation index, the raw material compatibility index, and the tolerance performance index with their respective weights. The higher the total evaluation score, the better the production stability evaluation index.

[0094] Step S55 involves obtaining the physicochemical property data of each component in the candidate formulation based on the product's final application scenario (e.g., outdoor use, high-temperature injection molding). This data includes, but is not limited to: Heat resistance: Whether the pigment changes color at the processing temperature.

[0095] Lightfastness / weatherfastness: The degree of fading of pigments under ultraviolet light.

[0096] Migration resistance: Whether the pigment will migrate and contaminate the object it comes into contact with.

[0097] Step S55 determines whether these parameters meet the customer's minimum threshold requirements; the more they meet the requirements, the higher the score.

[0098] Step S56 uses a multi-dimensional weighted scoring method to calculate the total evaluation score S. The calculation formula is as follows: S=w1×P 工艺 +w2×P 相容 +w3×P 耐受 Where: P 工艺 This is a normalized score calculated based on the process deviation index; P 相容 The score is calculated based on the raw material compatibility index (full marks for no contraindications, deductions for contraindications); P 耐受 The score is calculated based on the tolerance performance index parameters; w1, w2, w3 are the preset weight coefficients corresponding to each index (for example, for mass production orders, the process has a large weight; for outdoor products, the weather resistance has a large weight).

[0099] Because the physical dimensions and value ranges of process deviation indicators (expressed as numerical differences, such as percentages), raw material compatibility indicators (expressed as presence or absence, such as 0 or 1), and tolerance indicators (expressed as grades, such as 1-8) are different and their evaluation directions are inconsistent—for example, a smaller process deviation value indicates greater stability, while a higher tolerance grade usually indicates greater stability—the system will normalize and align the parameters of each indicator before substituting them into the above weighted summation formula to calculate the total evaluation score.

[0100] The specific processing method is as follows: For the process deviation index, the system uses reverse mapping or linear interpolation to convert it into a positive score P. 工艺 For example, a perfect score (e.g., 100 points) and a maximum allowable deviation threshold (e.g., 0.05) can be set. When the absolute value of the difference is 0, P... 工艺 The score is 100; as the absolute value of the difference increases, P... 工艺 It decreases linearly proportionally; when the absolute value of the difference exceeds the threshold, P... 工艺 The value is set to zero or negative. After this processing, the smaller the difference, the higher the corresponding normalized score P. 工艺 The higher the score, the more consistent it will be with the overall evaluation score.

[0101] For raw material compatibility and tolerance indicators, the system employs a piecewise mapping method. For example, non-repulsive combinations are mapped to P. 相容 =100, and there is a repulsive combination mapping to P. 相容 =0 or 60; directly map the tolerance level to the corresponding standard score (e.g., level 8 corresponds to 100 points, level 1 corresponds to 20 points).

[0102] Through the above normalization process, each sub-indicator is transformed into a dimensionless value of the same magnitude and in the same direction (all are better the higher they are), thus ensuring the mathematical meaning and evaluation accuracy of the weighted summation result.

[0103] In addition, any one of the three indicators—process deviation index, raw material compatibility index, and tolerance index—can be selected, i.e., one of w1, w2, and w3 is 1, and the others are 0. Alternatively, any two of the three indicators can be selected, for example, two of them have a weight of 1:1, and the remaining one has a weight of 0.

[0104] Through a weighted summation-based comprehensive evaluation mechanism, this invention achieves a multi-objective global optimization. It does not merely seek the formulation with the smallest color difference, nor simply the "easiest" formulation to produce, but rather finds the optimal balance between process feasibility (easy to produce), chemical stability (non-reactive), and physical durability (durable). This decision-making mechanism ensures that the final selected formulation is a robust product that performs excellently throughout its entire life cycle (production, storage, and use).

[0105] In one embodiment, the visual perception-based robust color formulation determination method further includes the step of: Step S70: If the target color does not fall within the preset sensitive color system area, then the color formula corresponding to the target color is used as the production formula.

[0106] When the judgment result of step S20 is "no", for example, if the target color is a highly saturated bright red or dark blue, its McAdam ellipse area is large, or its chroma value is greater than the preset threshold, the system will directly use the original "target color" data provided by the customer as the sole basis and call the color matching algorithm to calculate the candidate formula.

[0107] For non-sensitive colors (regular colors), the corresponding color difference tolerance area (McAdam ellipse) is geometrically large, meaning that the "visual safety zone" for this color is wide. For these colors, the formula calculated directly from the target color is itself located in the center of this large safety zone. Even if certain process fluctuations occur in actual production, such as dye weighing errors or temperature fluctuations, because the safety boundary is far from the center, the color coordinates of the final product will most likely still fall within this large ellipse range, without causing a color difference visible to the human eye. Therefore, for these colors, there is no need to deliberately find a reference color; the color formula for the target color already possesses sufficient statistical robustness.

[0108] Step S70 further demonstrates that, based on the judgment in step S20, high-level optimization can be initiated only for sensitive colors with extremely low error tolerance and high production risks, while the pigment can be directly determined for non-sensitive colors. This design solves the problem of inaccurate application of sensitive colors while retaining the high reliability of non-sensitive colors, significantly improving the overall operating efficiency of the color matching system.

[0109] Compared with commonly used technologies, this embodiment has the following advantages: This robust color formulation determination method and system based on visual perception breaks through the limitations of traditional single fixed color difference tolerance values ​​by leveraging the dynamic changes in human color perception. For target colors identified as sensitive colors, the system calculates and evaluates formulations corresponding to different reference points within the region. By utilizing the visual tolerance of the human eye, the solution scope is expanded from a single "point" to a "region." By calculating and evaluating formulations corresponding to different reference points within this region, formulations that are within the visual tolerance range and are more stable and robust in production are identified. This ensures that the final selected formulation not only meets color difference standards but also maintains process stability. While ensuring visual consistency, the system simultaneously improves the formulation's tolerance to production fluctuations, significantly reducing the problem of color differences caused by process fluctuations being perceived by the human eye in actual production, thereby increasing product qualification rates and improving production efficiency.

[0110] Example 2: A robust color formulation determination system based on visual perception In one embodiment, this application provides a robust color formulation determination system based on visual perception, such as... Figure 5 As shown, it includes: The data receiving module is used to receive standard data for the target color.

[0111] The sensitivity determination module is used to determine whether the target color falls into a preset sensitive color system area based on the standard data. The sensitive color system area is determined based on the visual recognition threshold characteristics of the color space, and is an area where the human eye is more sensitive to color difference changes than a preset level.

[0112] The reference color determination module is used to determine at least one reference color that is different from the target color when the target color is determined to fall into a sensitive color system area, within a local color difference tolerance area. The local color difference tolerance area is an area centered on the target color and where the color difference value between the reference color and the target color meets a preset color difference tolerance threshold condition.

[0113] The formula calculation module is used to determine the candidate color formulas corresponding to the target color and the reference color.

[0114] The formulation optimization module is used to obtain preset production stability evaluation indicators, evaluate each of the candidate color formulations, and select the candidate color formulation with the best production stability evaluation indicators as the production formulation of the target color.

[0115] It should be noted that for details not disclosed in the visual perception-based robust color formulation determination system of this application embodiments, please refer to the details disclosed in the visual perception-based robust color formulation determination method of this application embodiments.

[0116] It should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0117] The detailed descriptions listed above are merely specific descriptions of feasible implementation methods of this application, and are not intended to limit the scope of protection of this application. All equivalent implementation methods or modifications made without departing from the specific spirit of this application should be included within the scope of protection of this application.

Claims

1. A robust color formulation determination method based on visual perception, characterized in that, Includes the following steps: Receive standard data for the target color; Based on the standard data, it is determined whether the target color falls into a preset sensitive color system area, wherein the sensitive color system area is determined based on the visual recognition threshold characteristics of the color space, and is an area where the human eye is more sensitive to color difference changes than a preset level. If the judgment result is yes, within the local color difference tolerance area, at least one reference color different from the target color is determined, wherein the local color difference tolerance area is the area centered on the target color and the color difference value between the color and the target color satisfies the preset color difference tolerance threshold condition; Determine the candidate color formulas corresponding to the target color and the reference color; Obtain preset production stability evaluation indicators and evaluate each of the candidate color formulations; The candidate color formulation with the best production stability evaluation index is selected as the production formulation for the target color.

2. The robust color formulation determination method based on visual perception according to claim 1, characterized in that, The step of determining whether the target color falls within a preset sensitive color range based on the standard data includes: Calculate the area of ​​the local color difference tolerance region corresponding to the target color; Determine whether the area of ​​the region is less than or equal to a preset area threshold; If so, the target color is determined to fall within a preset sensitive color area; If not, it is determined that the target color does not fall within the preset sensitive color system area.

3. The robust color formulation determination method based on visual perception according to claim 1, characterized in that, The step of determining whether the target color falls within a preset sensitive color range based on the standard data includes: Based on the standard data, determine the chroma value, L value, a value, and b value of the target color in the CIELAB color space; Determine whether the chroma value, the L value, the a value, and the b value all fall within the preset value range corresponding to each parameter; If so, the target color is determined to fall within a preset sensitive color area; If not, it is determined that the target color does not fall within the preset sensitive color system area.

4. The robust color formulation determination method based on visual perception according to claim 1, characterized in that, The permissible area for local color difference is the McAdam circle.

5. The robust color formulation determination method based on visual perception according to claim 4, characterized in that, Determining at least one reference color distinct from the target color includes: Determine the coordinates of the major axis endpoint and the minor axis endpoint of the McAdam loop; At least one of the major axis endpoint coordinates and the minor axis endpoint coordinates is determined as the reference color.

6. The robust color formulation determination method based on visual perception according to claim 1, characterized in that, The production stability evaluation index includes the process deviation index; The evaluation of each of the candidate color formulations includes: Determine the theoretical coloring ratio of each pigment in each candidate color formula; Obtain the statistical process baseline ratio of the production equipment for each of the pigments; Calculate the absolute value of the difference between the theoretical coloring ratio and the statistical process baseline ratio, wherein the smaller the absolute value of the difference, the higher the score of the process deviation index.

7. The robust color formulation determination method based on visual perception according to claim 6, characterized in that, The production stability evaluation indicators include raw material compatibility indicators; The evaluation of each of the candidate color formulations includes: Detect whether there is a preset combination of incompatible colorants in the candidate color formula; If so, the score of the raw material compatibility index corresponding to the candidate color formulation shall be reduced.

8. The robust color formulation determination method based on visual perception according to claim 7, characterized in that, The production stability evaluation index includes a tolerance performance index, wherein the tolerance performance index includes at least one of heat resistance, light resistance, weather resistance, or migration resistance. The evaluation of each of the candidate color formulations includes: Obtain the tolerance performance parameters corresponding to the tolerance performance index for each of the candidate color formulations; The evaluation of each of the candidate color formulations further includes: Calculate the total evaluation score for each of the candidate color formulations. The total evaluation score is the weighted sum of the process deviation index, the raw material compatibility index, and the tolerance performance index with their respective weights. The higher the total evaluation score, the better the production stability evaluation index.

9. The robust color formulation determination method based on visual perception according to claim 1, characterized in that, It also includes the following steps: If the target color does not fall within the preset sensitive color system area, then the color formula corresponding to the target color will be used as the production formula.

10. A robust color formulation determination system based on visual perception, characterized in that, include: The data receiving module is used to receive standard data for the target color; The sensitivity determination module is used to determine whether the target color falls into a preset sensitive color system area based on the standard data. The sensitive color system area is determined based on the visual recognition threshold characteristics of the color space and is an area where the human eye is more sensitive to color difference changes than a preset level. The reference color determination module is used to determine at least one reference color that is different from the target color within a local color difference tolerance area when it is determined that the target color falls into a sensitive color system area. The local color difference tolerance area is an area centered on the target color and where the color difference value between the color and the target color meets a preset color difference tolerance threshold condition. A formula calculation module is used to determine candidate color formulas corresponding to the target color and the reference color; The formulation optimization module is used to obtain preset production stability evaluation indicators, evaluate each of the candidate color formulations, and select the candidate color formulation with the best production stability evaluation indicators as the production formulation of the target color.

Images