Intelligent wall coating color matching system

Abstract

The present application relates to the technical field of image processing, specifically to an intelligent wall coating color matching system, which comprises an image capture module, a color feature analysis module, a color trend prediction module and a color matching optimization module. In the present application, the color histogram and texture analysis are comprehensively utilized to conduct in-depth analysis on the color shift and texture resolution information in the image, thereby improving the prediction accuracy of the color change of the wall surface. Time series analysis is used to calculate the color change trend based on historical data, thereby enhancing the prediction ability of the future color change of the wall surface. This deep analysis and prediction method can not only effectively predict the degradation trend of the color of the wall surface, but also predict the color shift caused by external factors such as light changes, thereby optimizing the selection and use of the coating. The adoption of this method improves the aesthetic durability of the wall surface and reduces the need for frequent maintenance, and is suitable for commercial and residential fields, which helps users or designers to make more accurate color matching and long-term maintenance planning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to an intelligent wall coating color matching system. BACKGROUND

[0002] The technical field of image processing involves the process of analyzing and modifying images to improve or extract information, and the technical field mainly relies on digital image algorithms to perform tasks such as image enhancement, restoration, compression and feature recognition. Specific applications include medical imaging, satellite data processing, machine vision, automated monitoring systems, through image processing, useful features can be extracted from the original image, the quality of the image can be improved, or the image format can be converted to meet different processing needs, this technology not only includes basic pixel operations, but also includes more advanced mathematical and statistical methods to identify patterns and make decisions.

[0003] Among them, the intelligent wall coating color matching system is a system that applies image processing technology to automatically recommend or select wall coating colors, the system analyzes the input image (such as an indoor environment photo) to determine the color matching scheme, and then recommends the coating color that is coordinated with the existing environment or the user's preference. Widely used, can be applied to interior design, renovation and commercial display fields, helping users or designers to quickly make aesthetic color selection, improve decision-making efficiency and satisfaction. This system is particularly suitable for situations that require frequent or large-scale color selection, such as decoration companies, designers or DIY projects.

[0004] The existing intelligent wall coating color matching system mainly relies on basic pixel operations and simple statistical methods in color analysis and prediction, which limits the in-depth understanding of complex color change trends. For example, in the field of interior design and renovation, existing technologies cannot accurately predict color changes caused by environmental factors, such as the long-term impact of light and humidity changes on wall color. This technical limitation prevents designers and users from effectively predicting the performance of colors in different environmental conditions, resulting in the selection of unsuitable coating colors, which ultimately affects the aesthetics of the overall design and the service life of the materials. The existing technology fails to provide a solution that can dynamically optimize color matching by combining user preferences and environmental factors, which limits the selection range and flexibility of design schemes in practical applications. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art, and an intelligent wall coating color matching system is proposed.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: an intelligent wall coating color matching system comprises:

[0007] The image capturing module automatically adjusts the shooting angle and the light intensity based on the continuous monitoring of the wall surface, captures a continuous image sequence of the wall surface, generates a wall surface image dataset, corrects color deviation and brightness variation using the wall surface image dataset, and generates a corrected image dataset;

[0008] The color feature analysis module obtains color deviation and texture resolution information in each image based on the corrected image dataset through color histogram and texture analysis, generates a color feature set, performs dynamic analysis of color deviation using the color feature set, and generates a color characteristic evaluation result;

[0009] The color trend prediction module calculates a color change trend prediction based on historical data using time series analysis based on the color characteristic evaluation result, generates a trend analysis result, predicts the fading and change of the wall surface color in the future time period using the trend analysis result, and generates a dynamic deviation prediction analysis result;

[0010] The color matching optimization module analyzes user preferences and environmental factors based on the dynamic deviation prediction analysis result, performs color matching tests, generates matching test results, selects colors based on the matching test results, adjusts paint color matching using environmental simulation, and generates an optimal color matching scheme.

[0011] As a further scheme of the present application, the corrected image dataset acquisition step is specifically:

[0012] Based on the continuous monitoring of the wall surface, the color values of each image are extracted from the original wall surface image sequence, the color deviation between each image and the standard color value is calculated, and the formula is:

[0013]

[0014] A color deviation list is generated, wherein respectively represent the red, green, and blue channel values of the image , is the corresponding channel of the standard color value, represents the color deviation of each image, represents the color value of the th image, represents the color deviation list;

[0015] Using the color deviation list, the color deviation of the image sequence is calculated, and the deviation values are weighted and averaged, using the formula:

[0016]

[0017] A weighted average color deviation is generated, wherein, represents the adaptive weight, representing the weighted average color deviation, representing the color deviation of each image;

[0018] correcting each image using the weighted average color deviation, applying the formula:

[0019]

[0020] adjusting the color of each image and reducing the deviation, generating a corrected image dataset, wherein, is a correction parameter, is a weighted average color deviation, is a color deviation of each image, represents a corrected image value, represents an uncorrected image value.

[0021] As a further scheme of the present application, the color feature set acquisition step is specifically:

[0022] Based on the corrected image dataset, color distribution analysis is performed on each image in the image dataset, using the formula:

[0023]

[0024] The color class distinction is enhanced by a logical function, and color histogram data is generated, wherein, represents the color histogram of the th image, is the color of each pixel point in the image, is the color class, is a color difference sensitivity control parameter, represents the color distance;

[0025] Based on the color histogram data, a texture analysis technique is applied, and the formula

[0026]

[0027] Texture information data is generated, wherein, represents the texture information of the th image, is the number of windows, represents the color histogram of the th image, is a smoothing parameter;

[0028] The results of combining color histogram data and texture information data are combined using a fusion formula:

[0029]

[0030] generating a color feature set, wherein, is the first color feature of the image, is a fusion coefficient, represents the first texture information of the image, represents the first color histogram of the image.

[0031] As a further scheme of the present application, the color characteristic evaluation result obtaining step is specifically:

[0032] Based on the color feature set, the color offset of each image to the reference image is calculated, and the formula is:

[0033]

[0034] generating a normalized color offset, wherein, is the first feature value of the image in the first dimension, is the corresponding feature value of the reference image, is a normalization factor, is the normalized color offset;

[0035] Based on the normalized color offset, the statistical trend of the color offset is calculated, and the exponential weighted moving average formula is used:

[0036]

[0037] generating a color offset dynamic trend, wherein, is a smoothing parameter, is a new parameter for adjusting historical influence, is the color offset at the current time point, is the color offset dynamic trend;

[0038] Using the color offset dynamic trend, abnormality detection is performed and it is judged whether the color change exceeds the threshold value, by the judging formula:

[0039]

[0040] obtaining a color characteristic evaluation result, wherein, is a set abnormality detection threshold value, is the color characteristic evaluation result, is the color offset dynamic trend.

[0041] As a further scheme of the present application, the trend analysis result obtaining step is specifically:

[0042] Extracting historical color data from the color characteristic evaluation results, using time series analysis method to predict color change trend in future time period, applying formula:

[0043]

[0044] Generating trend prediction results, wherein, is the predicted color value, is the coefficient of time series model, represents the order of regression, is the decay factor, is the model error term;

[0045] Seasonal and periodic adjustment is made to the trend prediction results, using formula:

[0046]

[0047] Generating adjusted prediction values, wherein, is the seasonal adjusted prediction value, adjusts the strength of periodic influence, is the length of the cycle, is the phase adjustment, adjusts the decay effect of time;

[0048] Accuracy verification is made to the adjusted prediction values, using error analysis formula:

[0049]

[0050] Generating trend analysis results, wherein, represents the trend analysis results, is the prediction value, is the corresponding observation value, is the number of observations in the evaluation period.

[0051] As a further scheme of the present application, the obtaining step of the drift dynamic prediction analysis results is specifically:

[0052] Using the data obtained from the trend analysis results, initializing the prediction model of drift dynamics, using weighted autoregressive model for preliminary prediction, through formula:

[0053]

[0054] Generating preliminary drift prediction results, wherein, is the historical color data, controls the strength of historical data influence, represents the preliminary drift prediction results, represents the adjustment decay rate, a time interval representing historical data;

[0055] seasonally adjusting the preliminary offset prediction result, referring to the influence of periodic change, using the formula:

[0056]

[0057] generating an adjusted offset prediction result, wherein, adjusting the amplitude of the periodic influence, determining the length of the period, for the phase offset, representing the adjusted offset prediction result.

[0058] combining the adjusted offset prediction result and the preliminary offset prediction result, performing error analysis and prediction output, using the formula:

[0059]

[0060] generating a dynamic offset prediction analysis result, wherein, representing the dynamic offset prediction analysis result, is the total number of data points participating in the calculation, representing the real-time observation value, representing the adjusted prediction value.

[0061] As a further scheme of the present application, the step of obtaining the matching test result is specifically:

[0062] based on the dynamic offset prediction analysis result, analyzing the influence of environmental factors, using the formula:

[0063]

[0064] controlling the influence weight of environmental brightness by adjusting the coefficient, generating an environmental adjustment color value, wherein, representing the adjustment proportion of environmental factors on color prediction, representing the environmental brightness value, representing the environmental adjustment color value, initial color prediction, is the environmental brightness index;

[0065] combining the environmental adjustment color value with user preferences, using the formula:

[0066]

[0067] balancing the color result of user preferences and environmental adjustment by adjusting the coefficient, generating an adjusted color matching value, wherein, is the adjusted color matching value, adjusting the relative importance of user preference and environmental impact, for a user preference intensity index, representing an environmental adjustment color value;

[0068] performing a color matching effect evaluation on the adjusted color matching value, introducing an evaluation formula:

[0069]

[0070] generating a matching test result, wherein, representing the sum of deviations between predicted and real-time matching results, is the adjusted color matching value, is the observed real-time matching result, and is a newly introduced adjustment parameter, is the number of samples participating in the evaluation.

[0071] As a further scheme of the present application, the step of obtaining the optimal color matching scheme is specifically:

[0072] based on the matching test result, analyzing the environmental and user preference factors, and applying a formula:

[0073]

[0074] generating an environment-aware color value, wherein, represents an adjusted color matching preliminary result, is the matching test result, represents an environmental factor, adjusts the weight of the matching test result, adjusts the weight of the environmental factor;

[0075] According to the generated environment-aware color value, color selection and adjustment are performed in a simulated environment, using a formula:

[0076]

[0077] generating a simulated adjusted color matching scheme, wherein, represents a simulated adjusted color matching result, represents a simulation coefficient based on user preference, represents an adjusted color matching preliminary result, represents the influence weight of color test results in the simulation result, represents the influence weight of adjusting user preference;

[0078] using the simulated adjusted color matching scheme for color matching optimization, using a formula:

[0079]

[0080] generating an optimal color matching scheme, wherein for the optimal color matching scheme, representing a color stability parameter, highlighting a color depth impact parameter, adjusting a color stability impact parameter.

[0081] Compared with the prior art, the advantages and positive effects of the present application are:

[0082] In the present application, by comprehensively utilizing color histogram and texture analysis to deeply analyze the color shift and texture resolution information in the image, the prediction accuracy of the color change of the wall surface is improved, time series analysis is used to calculate the color change trend according to historical data, and the prediction ability of the future color change of the wall surface is enhanced. This kind of deep analysis and prediction method can not only effectively predict the degradation trend of the color of the wall surface, but also predict the color shift caused by external factors such as light change, so as to optimize the selection and use of paint. The use of this method improves the aesthetic durability of the wall surface and reduces the need for frequent maintenance, and is suitable for commercial and residential fields, which helps users or designers to make more accurate color matching and long-term maintenance planning. BRIEF DESCRIPTION OF DRAWINGS

[0083] Figure 1 is a system flowchart of the present application;

[0084] Figure 2 is a flowchart of the corrected image data set in the present application;

[0085] Figure 3 is a flowchart of the color feature set in the present application;

[0086] Figure 4 is a flowchart of the color characteristic evaluation result in the present application;

[0087] Figure 5 is a flowchart of the trend analysis result in the present application;

[0088] Figure 6 is a flowchart of the color shift dynamic prediction analysis result in the present application;

[0089] Figure 7 is a flowchart of the matching test result in the present application;

[0090] Figure 8 is a flowchart of the optimal color matching scheme in the present application. DETAILED DESCRIPTION

[0091] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0092] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0093] Embodiment one

[0094] Please refer to Figure 1 An intelligent wall coating color matching system comprises:

[0095] The image capture module automatically adjusts the shooting angle and light intensity based on continuous monitoring of the wall surface, captures a continuous image sequence of the wall surface, generates a wall surface image dataset, corrects color deviation and brightness variation using the wall surface image dataset, and generates a corrected image dataset;

[0096] The color feature analysis module obtains color offset and texture resolution information in each image based on the corrected image dataset through color histogram and texture analysis, generates a color feature set, performs dynamic analysis of color deviation using the color feature set, and generates a color characteristic evaluation result;

[0097] The color trend prediction module calculates color change trend prediction based on historical data using time series analysis based on the color characteristic evaluation result, generates a trend analysis result, predicts the fading and change of the wall color in the future time period using the trend analysis result, and generates a dynamic deviation prediction analysis result;

[0098] The color matching optimization module analyzes user preferences and environmental factors based on the dynamic deviation prediction analysis result, performs color matching test, generates a matching test result, selects color according to the matching test result, adjusts the color of the paint using environmental simulation, and generates an optimal color matching scheme.

[0099] The wall surface image dataset includes hue data, brightness level, image resolution, the corrected image dataset includes color balance degree, light consistency, contrast adjustment, the color feature set includes dominant hue distribution, detail texture identification, color depth index, the color characteristic evaluation result includes color offset, texture resolution, color saturation, the trend analysis result includes fading trend index, saturation change prediction, brightness attenuation prediction, the movement prediction analysis result includes long-term color offset prediction, short-term change amount, color difference range prediction, the matching test result includes toning test score, contrast test result, environment matching evaluation, and the optimal color matching scheme includes color matching scheme, visual impact evaluation, and user satisfaction prediction.

[0100] Please refer to Figure 2 The step of obtaining the corrected image dataset is specifically:

[0101] Based on the continuous monitoring of the wall surface, the color values of each image are extracted from the original wall surface image sequence, and the color deviation between each image and the standard color value is calculated. The formula is:

[0102]

[0103] A color deviation list is generated, wherein respectively represent the red, green, and blue channel values of the image , is the corresponding channel of the standard color value, represents the color deviation of each image, represents the color value of the th image, represents the color deviation list;

[0104] Using the color deviation list, the color deviation of the image sequence is calculated, and the deviation values are weighted and averaged. The formula is:

[0105]

[0106] A weighted average color deviation is generated, wherein represents the adaptive weight, represents the weighted average color deviation, represents the color deviation of each image;

[0107] Each image is corrected using the weighted average color deviation, and the formula is:

[0108]

[0109] The color of each image is adjusted and the deviation is reduced, and a corrected image dataset is generated, wherein is the correction parameter, is the weighted average color deviation, is the color deviation of each image, represents the corrected image value, represents the uncorrected image value.

[0110] Color deviation formula for each image:

[0111]

[0112] : the red, green, blue channel values of the first image;

[0113] : the red, green, blue channel values of the standard color;

[0114] Suppose an image has RGB values of (100, 120, 130);

[0115] Calculate the square of the red channel difference:

[0116]

[0117] Calculate the square of the green channel difference:

[0118]

[0119] Calculate the square of the blue channel difference:

[0120]

[0121] Sum and take the square root:

[0122]

[0123] Color deviation , which represents the absolute difference in color between the image and the standard color.

[0124] Weighted average color deviation formula:

[0125]

[0126] : adaptive weight based on deviation size;

[0127] Suppose there are three image color deviation values ;

[0128] Calculate each weight:

[0129] ​

[0130]

[0131]

[0132] The weighted average is calculated as:

[0133]

[0134] The weighted average color deviation represents the average color deviation in the image sequence considering the weights.

[0135] The corrected image value formula is:

[0136]

[0137] : correction adjustment parameter;

[0138] : exponential factor adjusting sensitivity;

[0139] Assume , use the previous and , ;

[0140] Calculate the correction term:

[0141]

[0142] Apply the correction:

[0143]

[0144] The corrected image value represents the new value of the image channel after correction.

[0145] Please refer to Figure 3 , the color feature set acquisition step is specifically:

[0146] Based on the corrected image dataset, perform color distribution analysis on each image in the image dataset, using the formula:

[0147]

[0148] Enhance the distinction between color categories through a logical function to generate color histogram data, where represents the color histogram of the th image, is the color of each pixel point in the image, is the color category, is the color difference sensitivity control parameter, Indicates color distance;

[0149] Based on color histogram data, texture analysis techniques are applied, using formulas...

[0150]

[0151] Generate texture information data, where, Representing the Texture information of the image, It is the number of windows. Representing the The color histogram of the image. It is a smoothing parameter;

[0152] The fusion formula is used by combining the results of color histogram data and texture information data:

[0153]

[0154] Generate a color feature set, where, It is the first The color characteristics of the image It is the fusion coefficient. Representing the Texture information of the image, Representing the The color histogram of the image.

[0155] No. Formula for the color histogram of an image:

[0156]

[0157] : No. The image is for color categories Color histogram values;

[0158] Pixels in an image The color value;

[0159] : Preset color categories;

[0160] Color distance, calculation and The Euclidean distance between them;

[0161] The standard deviation parameter controls the sensitivity and is used to adjust the effect of color difference on reaction intensity.

[0162] Suppose we are analyzing a simple single-channel grayscale image, where The value is 120, and the selected color category 100, standard deviation It is 15;

[0163] Calculate color distance :

[0164]

[0165] Calculate the color histogram value using the formula:

[0166]

[0167] This value represents the intensity of the image's color response at color category 100.

[0168] No. The formula for texture information of an image:

[0169]

[0170] : No. Texture resolution information of the image;

[0171] The number of analysis windows;

[0172] : No. The image in the first The color histogram of each window;

[0173] : Smoothing parameters to prevent division by zero;

[0174] Assume the image has 3 windows, each window The values ​​are as follows:

[0175] Window 1: ;

[0176] Window 2: ;

[0177] Window 3: ;

[0178] set up ;

[0179] Calculate the texture analysis value for each window:

[0180] Window 1:

[0181] Window 2: Similarly, the calculation yields...

[0182] Window 3: Similarly, the texture value is calculated as

[0183] Sum up the texture values of all windows:

[0184]

[0185] This value represents the overall texture information intensity of the image.

[0186] The first image color feature formula:

[0187]

[0188] The first image comprehensive color feature value;

[0189] Fusion coefficient, assumed to be 1.5;

[0190] Assuming the values of and are known, and is set to 1.5;

[0191] Calculate the color feature set:

[0192]

[0193]

[0194] This value represents the overall color and texture feature intensity of the image.

[0195] Please refer to Figure 4 , the color characteristic evaluation result acquisition step is specifically:

[0196] Based on the color feature set, calculate the color offset of each image to the reference image, using the formula:

[0197]

[0198] Generate normalized color offset, where is the feature value of the first image in the first dimension, is the corresponding feature value of the reference image, is the normalization factor, is the normalized color offset;

[0199] Based on the normalized color offset, calculate the statistical trend of the color offset using the exponential weighted moving average formula:

[0200]

[0201] a color shift dynamic trend is generated, wherein, is a smoothing parameter, is a new parameter for adjusting the historical impact, is the color shift amount at the current time point, is the color shift dynamic trend;

[0202] Using the color shift dynamic trend, abnormality detection is performed and it is determined whether the color change exceeds a threshold value, by the determination formula:

[0203]

[0204] a color characteristic evaluation result is obtained, wherein, is a set abnormality detection threshold value, is the color characteristic evaluation result, is the color shift dynamic trend.

[0205] The normalized color shift amount formula is:

[0206]

[0207] and represent the feature values of the first image and the reference image in the first dimension of the color feature vector, respectively;

[0208] is the normalization factor of the first dimension, used to adjust the proportion of different dimensions;

[0209] is the dimension of the feature vector;

[0210] Suppose there is a three-dimensional color feature vector , and the specific feature values are as follows:

[0211] For the image : ;

[0212] The reference image : ;

[0213] The normalization factor is: ;

[0214] Calculate :

[0215] Calculate the sum of squared differences and normalize:

[0216] denominator ;

[0217] sum of squared differences ;

[0218] normalized difference ;

[0219] calculate :

[0220]

[0221] this value (5.88) represents the euclidean distance in the normalized color feature space between the image and the reference image, lower values indicate close color characteristics to the reference, higher values indicate greater deviation.

[0222] color deviation dynamic trend formula:

[0223]

[0224] : smoothing coefficient, used to adjust the weight of new information;

[0225] : parameter to adjust the influence of historical trend;

[0226] : color deviation dynamic trend;

[0227] : color deviation dynamic at the previous time point;

[0228] Assume: , , (from step 1), ;

[0229] calculate :

[0230]

[0231] this value (4.19) represents a higher level considering historical data, indicating a sustained change in image color.

[0232] color characteristic evaluation result formula:

[0233]

[0234] : threshold value, used to distinguish normal and abnormal color change;

[0235] Assume: ;

[0236] Calculate :

[0237]

[0238]

[0239] This result (abnormal) indicates that the current color shift exceeds the set threshold, indicating an unusual color change that requires further investigation or intervention.

[0240] See Figure 5 , the trend analysis result obtaining step is specifically:

[0241] Extract historical color data from the color characteristic evaluation result, use time series analysis method to predict color change trend in future time period, apply formula:

[0242]

[0243] Generate trend prediction result, wherein, is the predicted color value, is the coefficient of time series model, represents the order of regression, is the decay factor, is the model error term;

[0244] Adjust the trend prediction result for seasonality and periodicity, using formula:

[0245]

[0246] Generate adjusted prediction value, wherein, is the seasonally adjusted prediction value, adjusts the strength of periodic influence, is the length of the cycle, is the phase adjustment, adjusts the decay effect of time;

[0247] Verify the accuracy of the adjusted prediction value using error analysis formula:

[0248]

[0249] Generate trend analysis result, wherein, represents the trend analysis result, is the prediction value, is the corresponding observation value, is the number of observations in the evaluation period.

[0250] The formula for the predicted color value:

[0251]

[0252] : time The predicted color value at time

[0253] : coefficients of the time series model, obtained by fitting to historical data;

[0254] : number of historical data points, representing the historical period considered by the model.

[0255] : decay factor, used to reduce the impact of old data;

[0256] : model error term, assumed to follow a normal distribution.

[0257] Suppose there are historical color values , and we use a third-order model with coefficients , a decay factor , and an error term as a simplification.

[0258] The value of is calculated as:

[0259]

[0260]

[0261]

[0262]

[0263] This means that the predicted color value at time is approximately , taking into account that the model has no error term.

[0264] The formula for the seasonally adjusted predicted value:

[0265]

[0266] : the seasonally adjusted predicted value;

[0267] : the periodic adjustment coefficient;

[0268] : Period length, e.g. 12 (months) for annual cycle;

[0269] : Phase adjustment, set according to season start month;

[0270] : Time decay factor, reduces influence of far-out data;

[0271] Using , assuming a period of months, current month (June, say) phase , decay factor ;

[0272] Calculating :

[0273]

[0274]

[0275]

[0276]

[0277] This indicates that the seasonally adjusted forecast is about , showing a seasonal growth.

[0278] Trend analysis result formula:

[0279]

[0280] : Square root of average relative error;

[0281] : Forecast value;

[0282] : Actual observation;

[0283] : Number of data points in evaluation period;

[0284] Assuming an actual observation , there is only one data point ;

[0285] Calculating :

[0286]

[0287]

[0288]

[0289]

[0290] This means that the prediction error is about 5%, showing the accuracy of the model.

[0291] See Figure 6 The step of obtaining the result of the drift dynamic prediction analysis is specifically:

[0292] Using the data obtained from the trend analysis result, initialize the prediction model of the drift dynamics, and use the weighted autoregressive model for preliminary prediction. Through the formula:

[0293]

[0294] Generate a preliminary drift prediction result, where, is the historical color data, controls the strength of the influence of historical data, represents the preliminary drift prediction result, represents the adjustment decay rate, represents the time interval of the historical data;

[0295] Seasonally adjust the preliminary drift prediction result, refer to the influence of periodic changes, and use the formula:

[0296]

[0297] Generate an adjusted drift prediction result, where, adjusts the amplitude of the periodic influence, determines the length of the cycle, is the phase shift, represents the adjusted drift prediction result.

[0298] Combine the adjusted drift prediction result and the preliminary drift prediction result, perform error analysis and prediction output, and use the formula:

[0299]

[0300] Generate a drift dynamic prediction analysis result, where, represents the drift dynamic prediction analysis result, is the total number of data points participating in the calculation, represents the real-time observation value, represents the adjusted prediction value.

[0301] Preliminary drift prediction result formula:

[0302]

[0303] Weighting coefficients influence the contribution of historical data to prediction.

[0304] : The color value at any given moment.

[0305] Time decay coefficient: Reduces the impact of old data.

[0306] : The time interval of historical data;

[0307] Assumption: , , , , ,

[0308] calculate :

[0309]

[0310]

[0311]

[0312]

[0313] Here The predicted offset is 82.712, indicating the color shift at the new moment predicted based on historical data.

[0314] The adjusted formula for the offset prediction result is as follows:

[0315]

[0316] : moderating the intensity of periodic effects;

[0317] Period length;

[0318] Phase adjustment parameters;

[0319] Assumption: , , , (30 degrees);

[0320] calculate :

[0321]

[0322]

[0323]

[0324] Here The seasonal adjusted predicted value is 55.141, showing the influence of periodic changes.

[0325] The formula for the moving dynamic prediction analysis result is:

[0326]

[0327] : Number of data points;

[0328] : Adjusted predicted value;

[0329] : Actual observed value;

[0330] Assume: , , , , ;

[0331] Calculate

[0332]

[0333]

[0334]

[0335] Here The root mean square of the prediction error is about 1.414, providing a quantitative evaluation of the accuracy of the model.

[0336] See Figure 7 , the steps for obtaining the test results are as follows:

[0337] Based on the moving dynamic prediction analysis result, the influence of environmental factors is analyzed, and the formula is:

[0338]

[0339] The influence weight of environmental brightness is controlled by adjusting the coefficient, and the environmental adjustment color value is generated, where represents the adjustment proportion of environmental factors on color prediction, represents the environmental brightness value, an environmental adjustment color value, an initial color prediction, an environmental brightness index;

[0340] combining the environmental adjustment color value with user preferences using the formula:

[0341]

[0342] balancing user preferences with environmental adjustment color results through an adjustment coefficient, generating an adjusted color palette value, wherein, is the adjusted color palette value, adjusting the relative importance of user preferences and environmental influences, is the user preference intensity index, an environmental adjustment color value;

[0343] performing a palette effectiveness evaluation on the adjusted color palette value, introducing an evaluation formula:

[0344]

[0345] generating a palette test result, wherein, is the sum of deviations between predicted and real-time palette results, is the adjusted color palette value, is the observed real-time palette result, and is a newly introduced adjustment parameter, is the number of samples participating in the evaluation.

[0346] environmental adjustment color value formula:

[0347]

[0348] : initial color prediction value, obtained from bias dynamic prediction analysis results, for example ;

[0349] : environmental brightness index, assuming ;

[0350] : adjustment coefficient, reflecting the weight of environmental factors, determined through experimental data fitting, assuming ;

[0351] determining the initial color prediction value and the environmental brightness index ;

[0352] determining the adjustment coefficient through experimental data fitting or experience;

[0353] Substitute , and into the formula and calculate, substituting specific numbers:

[0354]

[0355]

[0356]

[0357] The calculation result represents the color value adjusted by the environmental impact under the given environmental brightness index and initial color prediction.

[0358] Adjusted color matching value formula:

[0359]

[0360] : Environmentally adjusted color prediction value, from Step 1, for example ;

[0361] : User preference data, assuming ;

[0362] : Adjustment coefficient, reflecting the weight of user preferences, determined by user feedback data fitting, assuming ;

[0363] Obtain the calculation result of from the previous step;

[0364] Determine the user preference data ;

[0365] Determine the adjustment coefficient by user feedback data fitting or experience;

[0366] Substitute , and into the formula and calculate, substituting specific numbers:

[0367]

[0368]

[0369]

[0370] The calculation result Final color matching value after combining user preference data and environment adjustment results.

[0371] Deviation sum formula between prediction and real-time matching results:

[0372]

[0373] : Predicted matching result, e.g. ;

[0374] : Actual observed matching result, assuming ;

[0375] : Adjustment parameter to quantify the importance of each matching deviation, assuming ;

[0376] : Baseline deviation adjustment, assuming ;

[0377] : Sample size, assuming ;

[0378] Obtain the calculation result of from the previous step;

[0379] Determine the actual observed matching result ;

[0380] Determine the adjustment parameter and the baseline deviation adjustment through experiments or experience.

[0381] Substitute , , and into the formula for calculation;

[0382] Substitute specific numbers for calculation:

[0383]

[0384]

[0385]

[0386] The calculation result represents the deviation between prediction and real-time matching results, which is used to evaluate the accuracy of the matching effect.

[0387] Please refer toFigure 8 The obtaining step of the optimal color matching scheme is specifically:

[0388] Based on the results of the matching test, analyze the environmental and user preference factors, and apply the formula:

[0389]

[0390] Generate environment-aware color values, where, represents the adjusted color matching preliminary result, is the matching test result, represents the environmental factor, adjusts the weight of the matching test result, adjusts the weight of the environmental factor;

[0391] According to the generated environment-aware color values, perform color selection and adjustment in the simulated environment, using the formula:

[0392]

[0393] Generate simulated adjusted color matching schemes, where, represents the simulated adjusted color matching result, represents the simulation coefficient based on user preferences, represents the adjusted color matching preliminary result, represents the control simulation result color test result impact weight, represents the adjustment user preference impact weight;

[0394] Use the simulated adjusted color matching scheme for color matching optimization, using the formula:

[0395]

[0396] Generate the optimal color matching scheme, where is the optimal color matching scheme, represents the color stability parameter, emphasizes the impact parameter of color depth, adjusts the impact parameter of color stability.

[0397] Adjusted color matching preliminary result formula:

[0398]

[0399] Parameter explanation:

[0400] and : This is the adjustment weight, used to balance the impact of matching test results and environmental factors;

[0401] : numerical value of test result, representing color preference score from user feedback;

[0402] : numerical value of environmental factors, including light, surrounding color, and other environmental parameters;

[0403] Assumed parameter values: (assuming test result is 80 points) (assuming environmental factor score is 70);

[0404] Calculation :

[0405]

[0406] Here represents the color score after environmental and user preference weighting.

[0407] Formula for simulated adjusted color matching result:

[0408]

[0409] and : adjust the influence of result and simulation coefficient during color selection;

[0410] : simulation coefficient based on user preference adjustment, assuming environmental compatibility factor;

[0411] Assumed parameter values: , (assuming user preference simulation score)

[0412] Calculation :

[0413]

[0414] Here represents the color matching score after further simulation adjustment.

[0415] Formula for optimized color matching scheme:

[0416]

[0417] and : weight used to adjust color depth and stability;

[0418] : color stability parameter, which can be color persistence or resistance to light changes;

[0419] Assumed parameter values: (assumed color stability score);

[0420] Calculation :

[0421]

[0422] Here represents the comprehensive score of the optimal color matching scheme after the user preference, environmental factors and color stability are comprehensively considered.

[0423] The above merely describes the preferred embodiments of the present application, but is not intended to limit the present application in other forms. Any person skilled in the art can make changes or modifications to the above disclosed technical contents to obtain equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made to the above embodiments without departing from the technical solution of the present application, according to the technical essence of the present application, still belongs to the protection scope of the technical solution of the present application.

Claims

1. An intelligent wall paint color matching system, characterized in that, The system includes an image capture module, a color feature analysis module, a color trend prediction module, and a color matching optimization module. The image capture module is based on continuous monitoring of the wall, automatically adjusts the shooting angle and light intensity, captures a continuous image sequence of the wall, generates a wall image dataset, and uses the wall image dataset to correct color deviation and brightness changes, generating a corrected image dataset. Based on the corrected image dataset, the color feature analysis module obtains color shift and texture resolution information in each image through color histogram and texture analysis, generates a color feature set, and uses the color feature set to perform dynamic color shift analysis to generate color characteristic evaluation results. The specific steps for obtaining the color characteristic evaluation results are as follows: Based on the color feature set, calculate the color shift of each image relative to the reference image using the formula: ； Generate normalized color offsets, where... It is the first The image in the first eigenvalues ​​of dimension These are the corresponding feature values ​​of the reference image. It is a normalization factor. It is the normalized color offset; Based on the normalized color shift, the statistical trend of the color shift is calculated using the exponentially weighted moving average formula: ； Generate a dynamic trend for color shift, where... It is a smoothing parameter. To adjust the new parameters for historical impact, It is the color offset at the current time point. It is a dynamic trend of color shift; Using the aforementioned color shift dynamic trend, anomaly detection is performed to determine whether the color change exceeds a threshold, based on the following formula: ； The color characteristic evaluation results were obtained, among which, It is the set anomaly detection threshold. This is the result of color characteristic evaluation. It is a dynamic trend of color shift; Based on the color characteristic evaluation results, the color trend prediction module uses time series analysis to calculate color change trend predictions based on historical data, generates trend analysis results, and uses the trend analysis results to predict the fading and changes of wall color in the future time period, generating offset dynamic prediction analysis results. Based on the offset dynamic prediction analysis results, the color matching optimization module analyzes user preferences and environmental factors, conducts color matching tests, generates matching test results, selects colors according to the matching test results, adjusts paint color matching using environmental simulation, and generates the optimal color matching scheme.

2. The intelligent wall coating color matching system according to claim 1, characterized in that, The specific steps for obtaining the corrected image dataset are as follows: Based on continuous wall monitoring, the color value of each image is extracted from the original wall image sequence. The color deviation between each image and the standard color value is calculated using the following formula: ； Generate a list of color deviations, in which Representing images respectively The red, green, and blue channel values, These are the corresponding channels for standard color values. This represents the color deviation of each image. Representing the The color values ​​of the image, List of color deviations; Using the aforementioned color deviation list, the color deviation of the image sequence is calculated, and the deviation values ​​are weighted and averaged using the formula: ； Generate a weighted average color deviation, where, Represents adaptive weights. This represents the weighted average color deviation. Represents the color deviation of each image; Each image is corrected using the weighted average color deviation correction, applying the formula: ； The colors of each image are adjusted and biases are reduced to generate a corrected image dataset, in which... These are calibration parameters. It is a weighted average color deviation. It's the color deviation of each image. This represents the corrected image value. This indicates uncorrected image values.

3. The intelligent wall coating color matching system according to claim 2, characterized in that, The specific steps for obtaining the color feature set are as follows: Based on the corrected image dataset, a color distribution analysis is performed on each image in the dataset using the following formula: ； Color histogram data is generated by enhancing the distinguishability between color categories using logistic functions. Representing the The color histogram of the image. It is the color of each pixel in the image. It is a color category. It is a color difference sensitivity control parameter. Indicates color distance; Based on the color histogram data, texture analysis techniques are applied using the formula... ； Generate texture information data, where, Representing the Texture information of the image, It is the number of windows. Representing the The color histogram of the image. It is a smoothing parameter; The fusion formula is used by combining the results of color histogram data and texture information data: ； Generate a color feature set, where, It is the first The color characteristics of the image It is the fusion coefficient. Representing the Texture information of the image, Representing the The color histogram of the image.

4. The intelligent wall coating color matching system according to claim 1, characterized in that, The specific steps for obtaining the trend analysis results are as follows: Historical color data is extracted from the color characteristic evaluation results, and time series analysis is used to predict color change trends in future time periods, applying the formula: ； Generate trend prediction results, where, It is the predicted color value. These are the coefficients of the time series model. Indicates the order of regression. It is the attenuation factor. It is the model error term; The trend forecast results are adjusted for seasonality and cyclicality using the following formula: ； Generate adjusted predicted values, where, This is a seasonally adjusted forecast. To modulate the intensity of periodic effects, It is the length of the period. It's phase adjustment. Adjust for the decaying effect of time; The accuracy of the adjusted predicted values ​​was verified using the error analysis formula: ； Generate trend analysis results, among which, This indicates the results of the trend analysis. It is a predicted value. These are the corresponding observed values. This refers to the number of observations during the evaluation period.

5. The intelligent wall coating color matching system according to claim 4, characterized in that, The specific steps for obtaining the offset dynamic prediction analysis results are as follows: Using the data obtained from the trend analysis results, an offset dynamic prediction model is initialized, and a weighted autoregressive model is used for preliminary prediction, through the formula: ； Generate preliminary migration prediction results, in which, It is historical color data. Controlling the intensity of the influence of historical data, This indicates the preliminary offset prediction results. This indicates an adjustment to the decay rate. Indicates the time interval for historical data; The preliminary offset prediction results are seasonally adjusted, taking into account the impact of periodic changes, using the following formula: ； Generate adjusted offset prediction results, where, Adjusting the magnitude of cyclical effects, Determine the length of the period. For phase shift, This indicates the adjusted offset prediction result; Combining the adjusted offset prediction results and the preliminary offset prediction results, error analysis and prediction output are performed, using the formula: ； Generate offset dynamic prediction analysis results, in which, This indicates the results of the dynamic prediction analysis of the offset. The total number of data points involved in the calculation. Represents real-time observations. This represents the adjusted forecast value.

6. The intelligent wall coating color matching system according to claim 5, characterized in that, The specific steps for obtaining the matching test results are as follows: Based on the aforementioned offset dynamic prediction analysis results, the influence of environmental factors is analyzed using the following formula: ； By adjusting the coefficients to control the influence weight of ambient brightness, ambient color values ​​are generated. This indicates the adjustment ratio of environmental factors to color prediction. Indicates the ambient brightness value. This indicates the ambient color adjustment value. Initial color prediction, Ambient brightness index; Combine the environmental color adjustment values ​​with user preferences using the formula: ； By adjusting coefficients to balance user preferences and environmentally adjusted color results, adjusted color matching values ​​are generated. The adjusted color matching values. Adjusting the relative importance of user preferences and environmental influences, For user preference intensity index, Indicates the ambient color adjustment value; The adjusted color matching values ​​are evaluated to assess their effectiveness, and an evaluation formula is introduced: ； Generate matching test results, among which, This represents the sum of the deviations between the predicted and real-time matching results. The adjusted color matching values. For the observed real-time matching results, and These are newly introduced adjustment parameters. The number of samples participating in the evaluation.

7. The intelligent wall coating color matching system according to claim 6, characterized in that, The specific steps for obtaining the optimal color scheme are as follows: Based on the results of the matching test, environmental and user preference factors were analyzed, and the following formula was applied: ； Generate environment-aware color values, where, This indicates the preliminary results of the adjusted color scheme. These are the results of the matching test. Indicates environmental factors, Adjust the weighting of the matching test results. Adjust the weighting of environmental factors; Based on the generated environment-aware color values, color selection and adjustment are performed in the simulated environment using the formula: ； Generate a simulated and adjusted color scheme, in which, This indicates the simulated color scheme result. This represents the simulation coefficients based on user preferences. This indicates the preliminary results of the adjusted color scheme. This indicates the influence weight of the color test results in the control simulation results. This indicates the weight of the influence on moderating user preferences; The color scheme is optimized using the simulated and adjusted color scheme, employing the following formula: ； Generate the optimal color scheme, where For an optimized color scheme, This represents a color stability parameter. Emphasizing the influence of color depth parameters, Adjusting parameters that affect color stability.

Images