Feature extraction method and system of huxiang pattern in display evaluation framework

Abstract

The application relates to the technical field of image processing, and particularly discloses a feature extraction method and system of lake Xiang pattern in a display evaluation architecture, which comprises the following steps: obtaining a geometric fidelity coefficient through a geometric fidelity model based on image scanning resolution, lens focal length and surface curvature of a photographed object; obtaining color fidelity through a color fidelity model based on color restoration accuracy, file color depth and spectral resolution; obtaining a processing feasibility coefficient through a processing feasibility model based on background complexity, digital acquisition accuracy and image compression rate; and obtaining a signal-to-noise contrast adaptation degree through a signal-to-noise contrast adaptation model based on the signal-to-noise ratio of an image under the conditions of geometric fidelity and color fidelity and the self contrast of a pattern image. The application establishes a systematic feature extraction evaluation system by constructing multi-dimensional quantitative models of geometric fidelity, color fidelity and processing feasibility, and improves the objectivity and comprehensiveness of evaluation.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically a method and system for feature extraction of Hunan-style patterns in a display and evaluation framework. Background Technology

[0002] As an important regional cultural heritage, the digital display and evaluation of Hunan patterns are of great significance for cultural inheritance and innovative application. Within the display and evaluation framework, the scientific and efficient extraction of key features of the patterns is fundamental to achieving high-quality digital presentation and subsequent analysis.

[0003] Currently, existing technologies for digital feature extraction of traditional patterns mostly focus on optimizing single image acquisition parameters or using general image processing algorithms. For example, they may perform simple enhancements by adjusting resolution or color space, or use fixed filtering and segmentation methods to extract pattern outlines.

[0004] However, existing technologies have significant drawbacks: First, they fail to establish a unified quantitative evaluation model covering multiple dimensions such as geometric fidelity, color fidelity, and processing feasibility, leading to subjective and one-sided evaluation of feature extraction quality. Second, they lack dynamic analysis of image signal-to-noise ratio, pattern contrast, and pattern feature adaptability, making it difficult to optimize extraction strategies based on specific pattern characteristics. Finally, they do not consider the synergistic optimization of imaging lighting conditions and the aforementioned evaluation indicators, resulting in poor robustness of feature extraction and poor display effects under changing environments. Therefore, how to provide a systematic and quantifiable feature extraction method to improve the data comprehensiveness in the display evaluation process is the technical problem that this invention aims to solve. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for feature extraction of Hunan patterns in a display evaluation framework, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for feature extraction of Hunan-style patterns in an evaluation framework, the method comprising: Based on image scanning resolution, lens focal length, and the surface curvature of the subject, geometric fidelity coefficients are obtained through a geometric fidelity model. Based on color reproduction accuracy, file color depth, and spectral resolution, color fidelity is obtained through a color fidelity model. Based on background complexity, digital acquisition accuracy, and image compression rate, processing feasibility coefficients are obtained through a processing feasibility model. Based on the image signal-to-noise ratio under geometric fidelity and color fidelity, as well as the contrast of the pattern itself, the signal-to-noise contrast adaptation degree is obtained through a signal-to-noise contrast adaptation model. Based on the signal-to-noise ratio, processing feasibility coefficient, and the standard deviation of the overall grayscale value of the image under the current illumination conditions, the standard deviation of the overall grayscale value of the image under the target illumination conditions is obtained through an illumination optimization model.

[0007] As a further aspect of the present invention: the step of obtaining the geometric fidelity coefficients based on the image scanning resolution, lens focal length, and surface curvature of the subject through a geometric fidelity model includes: Acquire image scanning resolution, lens focal length, and the surface curvature of the subject; The resolution deviation index is obtained by comparing the absolute difference between the image scanning resolution and the reference resolution with the reference resolution. The focal length deviation index is obtained by comparing the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length. The curvature index is obtained by performing maximum-min normalization on the surface curvature of the photographed object. The resolution deviation index, focal length deviation index, and curvature index are imported into the geometric fidelity model to obtain geometric fidelity. Among them, the larger the geometric fidelity value, the better the fidelity of the geometric features.

[0008] As a further aspect of the present invention: the step of obtaining color fidelity based on color reproduction accuracy, file color depth, and spectral resolution through a color fidelity model includes: Obtain color reproduction accuracy, file color depth, and spectral resolution; The color accuracy, file color depth, and spectral resolution are subjected to maximum-min normalization to obtain the color accuracy index, file color depth index, and spectral resolution index. The color accuracy index, file color depth index, and spectral resolution index are imported into the color fidelity model to obtain the color fidelity. The color fidelity model calculates color fidelity using a weighted fusion function; the higher the color fidelity value, the better the color reproduction fidelity.

[0009] As a further aspect of the present invention: the step of obtaining the processing feasibility coefficient through a processing feasibility model based on background complexity, digital acquisition accuracy, and image compression rate includes: Obtain background complexity, digital acquisition accuracy, and image compression rate; The background complexity, digital acquisition accuracy, and image compression rate are subjected to maximum-minimum normalization to obtain the background complexity index, acquisition accuracy index, and image compression rate index. The background complexity index, acquisition accuracy index, and image compression rate index are imported into the processing feasibility model to obtain the processing feasibility coefficient. The higher the processing feasibility coefficient value, the higher the feasibility of image processing.

[0010] As a further aspect of the present invention: the step of obtaining the signal-to-noise contrast adaptation degree based on the image signal-to-noise ratio under geometric fidelity and color fidelity, and the contrast of the pattern itself, through a signal-to-noise contrast adaptation model includes: The image signal-to-noise ratio and the contrast of the pattern itself are subjected to maximum-min normalization to obtain the signal-to-noise ratio index and the contrast index of the pattern. The signal-to-noise ratio influence weight is obtained by processing the ratio of geometric fidelity to the sum of geometric fidelity and color fidelity. The weight of the influence of pattern contrast is obtained by processing the ratio of color fidelity to the sum of geometric fidelity and color fidelity. The signal-to-noise ratio (SNR) index, pattern contrast index, SNR influence weight, and pattern contrast influence weight are imported into the SNR contrast fitting model to obtain the SNR contrast fitting degree. The SNR contrast fitting model is expressed as follows: ; in, Indicates signal-to-noise ratio (SNR) adaptation. This represents the signal-to-noise ratio index. Indicates the contrast index of the pattern. This indicates that the signal-to-noise ratio affects the weights. The weighting of pattern contrast is indicated by the following. Furthermore, the higher the value, the better the signal-to-noise ratio and contrast are adapted to the current pattern features.

[0011] As a further aspect of the present invention: the illumination optimization model is expressed as follows: in, This represents the standard deviation of the overall grayscale values ​​of the image under the target lighting conditions. This represents the standard deviation of the overall grayscale values ​​of the image under the current lighting conditions. Indicates the target signal-to-noise ratio fit. Indicates signal-to-noise ratio (SNR) adaptation. This represents the feasibility coefficient for the treatment.

[0012] The present invention also provides a feature extraction system for Hunan-style patterns in a display and evaluation framework, the system comprising: The geometric analysis module is used to obtain geometric fidelity coefficients based on image scanning resolution, lens focal length, and the surface curvature of the subject through a geometric fidelity model. The color analysis module is used to obtain color fidelity based on color reproduction accuracy, file color depth, and spectral resolution through a color fidelity model. The processing and analysis module is used to obtain the processing feasibility coefficient based on the background complexity, digital acquisition accuracy, and image compression rate through a processing feasibility model. The adaptation degree acquisition module is used to obtain the signal-to-noise contrast adaptation degree based on the image signal-to-noise ratio and the contrast of the pattern itself under geometric fidelity and color fidelity. The standard deviation extraction module is used to obtain the standard deviation of the overall gray value of the image under the target lighting condition based on the signal-to-noise contrast adaptation, processing feasibility coefficient, and the standard deviation of the overall gray value of the image under the current lighting condition, through the lighting optimization model.

[0013] As a further aspect of the present invention: the geometric analysis module includes: The information acquisition unit is used to acquire image scanning resolution, lens focal length, and surface curvature of the subject. The resolution deviation analysis unit is used to process the ratio of the absolute difference between the image scanning resolution and the reference resolution to the reference resolution to obtain the resolution deviation index. The focal length deviation analysis unit is used to compare the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length to obtain the focal length deviation index. The first normalization processing unit is used to perform maximum-minimum normalization processing on the surface curvature of the photographed object to obtain the curvature index. The fidelity output unit is used to import the resolution deviation index, focal length deviation index and curvature index into the geometric fidelity model to obtain the geometric fidelity. Among them, the larger the geometric fidelity value, the better the fidelity of the geometric features.

[0014] As a further aspect of the present invention: the color analysis module includes: The color information acquisition unit is used to acquire color reproduction accuracy, file color depth, and spectral resolution. The data conversion unit is used to perform maximum-min normalization processing on color reproduction accuracy, file color depth, and spectral resolution to obtain color accuracy index, file color depth index, and spectral resolution index. The second normalization processing unit is used to import the color accuracy index, file color depth index and spectral resolution index into the color fidelity model to obtain color fidelity. The color fidelity model calculates color fidelity using a weighted fusion function; the higher the color fidelity value, the better the color reproduction fidelity.

[0015] As a further aspect of the present invention: the processing and analysis module includes: The background information acquisition unit is used to acquire background complexity, digital acquisition accuracy, and image compression rate. The third normalization processing unit is used to perform maximum-minimum normalization processing on the background complexity, digital acquisition accuracy and image compression rate to obtain the background complexity index, acquisition accuracy index and image compression rate index. The feasibility output unit is used to import the background complexity index, acquisition accuracy index, and image compression rate index into the processing feasibility model to obtain the processing feasibility coefficient. The higher the processing feasibility coefficient value, the higher the feasibility of image processing.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: By constructing multi-dimensional quantitative models such as geometric fidelity, color fidelity, and processing feasibility, this invention establishes a systematic feature extraction evaluation system, improving the objectivity and comprehensiveness of the evaluation; by introducing a signal-to-noise contrast adaptation model, it dynamically analyzes the matching relationship between image quality indicators and specific pattern features, achieving targeted optimization of feature extraction strategies and improving the accuracy and adaptability of feature extraction; through an illumination optimization model, it collaboratively considers image quality evaluation results and illumination conditions, providing optimization targets for the imaging environment, effectively improving the robustness of feature extraction under different illuminations and the final display effect. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0018] Figure 1 This is a flowchart illustrating the feature extraction method for Hunan patterns within the evaluation framework.

[0019] Figure 2 This is a structural diagram of the feature extraction system for Hunan patterns in the evaluation framework. Detailed Implementation

[0020] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0021] Figure 1This is a flowchart illustrating a method for feature extraction of Hunan-style patterns within a display and evaluation framework. In one embodiment of the present invention, a method for feature extraction of Hunan-style patterns within a display and evaluation framework is provided, the method comprising: Step S100: Based on the image scanning resolution, lens focal length, and surface curvature of the subject, obtain the geometric fidelity coefficient through the geometric fidelity model; Step S200: Based on color reproduction accuracy (average color difference compared with the standard color chart), file color depth (bit depth), and spectral resolution, obtain color fidelity through a color fidelity model; Step S300: Based on background complexity (absolute difference in contrast between foreground and background), digital acquisition accuracy (point cloud density), and image compression rate (compression ratio), obtain the processing feasibility coefficient through the processing feasibility model; Step S400: Based on the image signal-to-noise ratio under geometric fidelity and color fidelity, and the contrast of the pattern itself, obtain the signal-to-noise contrast adaptation degree through the signal-to-noise contrast adaptation model; Step S500: Based on the signal-to-noise contrast adaptation, processing feasibility coefficient, and the standard deviation of the overall grayscale value of the image under the current illumination conditions, obtain the standard deviation of the overall grayscale value of the image under the target illumination conditions through the illumination optimization model.

[0022] The feature extraction method for Hunan-style patterns proposed in this invention, within the evaluation framework, systematically and quantitatively evaluates the feature extraction process through the collaborative work of multi-dimensional models. Compared to traditional methods that rely solely on single-parameter optimization or general image processing algorithms, this invention first objectively and quantitatively evaluates the geometric shape and color material of the digitized patterns using geometric fidelity and color fidelity models, thus solving the problem of subjective and one-sided evaluation in existing technologies. For example, in the digitization process of the aforementioned wood carving patterns, traditional methods may only focus on resolution or color difference, while this invention comprehensively considers multiple key parameters such as scanning resolution, lens focal length, surface curvature, color reproduction accuracy, file color depth, and spectral resolution, providing a more comprehensive fidelity evaluation.

[0023] Preferably, the step of obtaining the geometric fidelity coefficients through a geometric fidelity model based on image scanning resolution, lens focal length, and surface curvature of the subject is as follows: Acquire image scanning resolution, lens focal length, and the surface curvature of the subject; The resolution deviation index is obtained by comparing the absolute difference between the image scanning resolution and the reference resolution with the reference resolution. The focal length deviation index is obtained by comparing the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length. The curvature index is obtained by performing maximum-min normalization on the surface curvature of the photographed object. The resolution deviation index, focal length deviation index, and curvature index are imported into the geometric fidelity model to obtain the geometric fidelity. The geometric fidelity model is constructed such that the larger the geometric fidelity value, the better the geometric feature fidelity. The geometric fidelity model is expressed as follows: in, Indicates geometric fidelity. This indicates the resolution deviation index. Indicates the focal length deviation index. Represents the curvature index, the Furthermore, the larger the value, the better the fidelity of the geometric features.

[0024] The steps of obtaining image scanning resolution, lens focal length, and object surface curvature aim to collect raw physical parameters affecting image geometric fidelity. Image scanning resolution refers to the level of detail in the image during digitization, directly related to the ability to capture texture details; lens focal length determines the image's angle of view and perspective distortion, crucial for geometric shape reproduction; object surface curvature reflects the geometric complexity of the object's surface, with higher curvature increasing the risk of geometric distortion during digitization. These parameters can be obtained in various ways. For example, image scanning resolution can be directly read from the scanning device or camera settings, or estimated through image processing algorithms; lens focal length can be extracted from camera EXIF ​​information or obtained through calibration experiments; object surface curvature can be analyzed after acquiring point cloud data from a 3D scanner, or estimated through methods such as structured light projection and multi-view image reconstruction. The step of obtaining the resolution deviation index by comparing the absolute difference between the image scanning resolution and the reference resolution with the reference resolution is used to quantify the degree of deviation between the actual scanning resolution and the ideal reference resolution. The reference resolution can be preset according to the detail of the pattern, display requirements, or industry standards. Resolution Deviation Index The larger the value, the greater the deviation between the actual resolution and the reference resolution, and the lower the fidelity of the image's geometric details may be. This ratio ensures that the deviations at different reference resolutions are comparable. The focal length deviation index is obtained by comparing the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length. This step aims to assess the impact of the lens focal length on geometric fidelity during actual shooting. The optimal focal length typically refers to the focal length that minimizes perspective distortion and maximizes geometric fidelity at a specific shooting distance and scene. The allowable deviation from the optimal focal length sets an acceptable range of focal length error. Focal Length Deviation Index The larger the value, the greater the deviation between the actual focal length and the optimal focal length, and the more severe the geometric distortion of the image may be. For example, the optimal focal length can be obtained through experimental calibration or set based on empirical values; the allowable deviation from the optimal focal length can be determined based on lens characteristics, shooting distance, and requirements for geometric accuracy. The step of performing maximum-minimum normalization on the surface curvature of the subject to obtain the curvature index is used to unify surface curvature data of different ranges and units into a standardized interval for subsequent model calculations. Geometric Fidelity Model It adopts the form of an exponential function, characterized by the fact that as any deviation from the exponent increases, the geometric fidelity decreases. The error rate decreases exponentially, which aligns with the nonlinear impact of geometric distortion on fidelity in real-world scenarios. This model comprehensively reflects the combined effects of resolution, focal length, and object curvature on geometric feature reconstruction during image acquisition.

[0025] The following example illustrates the process of digitally acquiring a Hunan-style pattern. First, the image scanning resolution, the current lens focal length, and the surface curvature data of the subject obtained through a 3D scanner can be acquired using an image acquisition device. For instance, assuming the image scanning resolution is 600 dpi, the preset baseline resolution is 1200 dpi; the lens focal length is 50mm, the optimal focal length is 60mm, and the allowable deviation from the optimal focal length is 10mm; the processed surface curvature data shows a maximum value of 0.5 and a minimum value of 0, with an average curvature of 0.2 for the current pattern area. Based on this, the resolution deviation index can be calculated. The absolute difference between the image scanning resolution and the reference resolution is |600-1200| = 600 dpi. Ratioing this to the reference resolution of 1200 dpi yields... =600 / 1200=0.5. Next, calculate the focal length deviation index. The absolute difference between the lens focal length and the optimal focal length is |50-60|=10mm. This difference is then compared to the allowable deviation from the optimal focal length of 10mm to obtain the result. =10 / 10=1. Then, calculate the curvature index. The average curvature of the object's surface (0.2) is then subjected to max-min normalization. Assuming the normalization range is [0,1], then... =(0.2-0) / (0.5-0)=0.4. Finally, the calculated resolution deviation index is... =0.5, Focal Length Deviation Index =1 and curvature index =0.4 Import the geometric fidelity model. Through calculation, a specific geometric fidelity can be obtained. The value is approximately 0.149. The value is the geometric fidelity coefficient of the Hunan pattern under the current acquisition conditions. The closer the value is to 1, the better the geometric feature fidelity.

[0026] Through the above technical solution, this invention provides a more comprehensive, objective, and quantitative method for evaluating geometric fidelity. This method comprehensively considers three key physical parameters—image scanning resolution, lens focal length, and the surface curvature of the subject—and transforms them into a calculable deviation index. This index is then integrated through an exponential model, overcoming the problems of one-sided and inaccurate geometric feature evaluation in existing technologies. Specifically, this solution avoids evaluation bias caused by ignoring key acquisition parameters, resulting in more accurate geometric fidelity acquisition. This accurate geometric fidelity evaluation, as input for subsequent signal-to-noise ratio (SNR) influence weight calculation, significantly improves the accuracy of SNR fit, thereby optimizing the performance of the entire Hunan pattern feature extraction method. Ultimately, this contributes to achieving higher-quality digital presentation and more reliable subsequent analysis of Hunan patterns, providing stronger technical support for the digital protection and inheritance of cultural heritage.

[0027] Preferably, the steps for obtaining color fidelity using a color fidelity model, based on color reproduction accuracy (average color difference compared to a standard color chart), file color depth (bit depth), and spectral resolution, are as follows: Obtain color reproduction accuracy (average color difference compared to a standard color chart), file color depth (bit depth), and spectral resolution; The color accuracy, file color depth, and spectral resolution are subjected to maximum-min normalization to obtain the color accuracy index, file color depth index, and spectral resolution index. The color accuracy index, file color depth index, and spectral resolution index are imported into the color fidelity model to obtain the color fidelity. The color fidelity model calculates the color fidelity through a weighted fusion function, wherein the weighted fusion function is constructed such that the larger the color fidelity value, the better the color reproduction fidelity. The color fidelity model is expressed as follows: in, Indicates color fidelity. Indicates color accuracy index. Indicates the color depth index of the file. Indicates the spectral resolution index. Represents the weight coefficient and The Furthermore, the higher the value, the better the color fidelity.

[0028] Color accuracy (average color difference compared to a standard color chart) refers to the degree to which colors in a digital image match the colors of real objects. It is typically quantified by measuring the color difference (e.g., CIE Delta E value) between standard color patches in the image and the actual standard color chart. A smaller average color difference indicates higher color accuracy. Implementation methods can include calibrating the acquisition device using color calibration equipment (such as a spectrophotometer or colorimeter); or, after image acquisition, calculating the average Delta E value between a preset standard color patch area in the image and the known color values ​​of the standard color chart using image processing software, for example, using the CIE Delta E2000 or CIE Delta E76 standard. File color depth (bit depth) refers to the number of bits used to represent color information per pixel in an image file. A higher bit depth allows for a greater number of colors to be represented, smoother color transitions, and richer details. For example, an 8-bit bit depth can represent 256 colors, while a 16-bit bit depth can represent 65,536 colors. Implementation methods can include: selecting a high bit-depth output mode, such as 10-bit, 12-bit, or 16-bit, in the settings of the image acquisition device (such as a digital camera or scanner); or ensuring that the file format supports and maintains high bit-depth data, such as TIFF or RAW formats, during image processing and storage. Spectral resolution refers to the ability of an imaging system or sensor to distinguish different wavelengths of light. High spectral resolution means the system can capture finer spectral information, which is crucial for accurately reproducing the color and material properties of objects, especially when it is necessary to distinguish subtle color differences or analyze material composition. Implementation methods can include: using multispectral or hyperspectral imaging equipment for data acquisition, which is typically equipped with narrowband filter arrays or spectrometers capable of acquiring image data at smaller wavelength intervals (e.g., 5 nm, 10 nm); or, combining multiple broadband filter images and using spectral reconstruction algorithms to estimate finer spectral information. A color fidelity model is a mathematical model used to quantify the color fidelity of digital images of Hunan patterns. Its function is to comprehensively consider multiple indicators related to color quality and output a unified fidelity value in the form of a weighted sum. This model provides an objective and quantifiable evaluation standard to reflect the realism and accuracy of image colors. Weighting coefficients. , , The relative importance of each indicator to the final color fidelity can be flexibly adjusted according to actual application needs or expert experience, and the sum of them is 1 to ensure a reasonable allocation of weights.

[0029] As a specific implementation method, when digitally acquiring a Hunan-style pattern, its color-related parameters can be obtained first. For example, by photographing using the X-Rite ColorChecker Classic standard color chart and using image processing software (such as Capture One or Lightroom) to calculate the average CIE Delta E2000 value between the color chart area in the image and the color value of the standard color chart, a color reproduction accuracy of 5.2 can be obtained. Simultaneously, the acquired image file can be stored using a 12-bit color depth. Furthermore, if a multispectral camera is used for acquisition, its spectral resolution can be set to 10 nanometers. Next, these parameters are subjected to maximum-minimum normalization. Assuming the theoretical range of the preset color reproduction accuracy (Delta E) is 0 to 20, the file color depth range is 8 to 16 bits, and the spectral resolution range is 5 nanometers to 50 nanometers. The color accuracy index can be calculated as (20-5.2) / (20-0)=0.74; the file color depth index can be calculated as (12-8) / (16-8)=0.5; and the spectral resolution index can be calculated as (50-10) / (50-5)=0.88. These normalized indices are then imported into the color fidelity model. Assuming weighting coefficients... , , Let's set the values ​​to 0.4, 0.3, and 0.3 respectively. Then, the color fidelity C can be calculated as: The final color fidelity C value is 0.518. This value will be used as input for the subsequent signal-to-noise contrast adaptation model and illumination optimization model to guide the further extraction and display optimization of Hunan pattern features.

[0030] Through the above technical solution, this invention effectively solves the problem of inaccurate evaluation caused by the lack of comprehensive processing and standardization of color reproduction accuracy, file color depth, and spectral resolution in traditional methods for assessing the color fidelity of digital images of Hunan patterns. This solution achieves accurate, objective, and quantitative evaluation of the color fidelity of digital images of Hunan patterns by comprehensively acquiring and standardizing these three key color parameters and importing them into a weighted color fidelity model. This accurate color fidelity evaluation result, as an important input in the subsequent feature extraction process, can significantly improve the signal-to-noise contrast ratio and the accuracy of the illumination optimization model, thereby ensuring that the feature extraction results of Hunan patterns more realistically reflect the color characteristics of the patterns themselves, laying a solid foundation for high-quality digital display and evaluation.

[0031] Preferably, the steps for obtaining the processing feasibility coefficient through a processing feasibility model based on background complexity (absolute difference in contrast between foreground and background), digital acquisition accuracy (point cloud density), and image compression rate (compression ratio) are as follows: Obtain background complexity (absolute difference in contrast between foreground and background), digitization acquisition accuracy (point cloud density), and image compression rate (compression ratio); The background complexity (absolute difference in contrast between foreground and background), digital acquisition accuracy (point cloud density), and image compression rate (compression ratio) are subjected to maximum-minimum normalization to obtain the background complexity index, acquisition accuracy index, and image compression rate index. The background complexity index, acquisition accuracy index, and image compression rate index are imported into the processing feasibility model to obtain the processing feasibility coefficient. The processing feasibility model is constructed such that the larger the processing feasibility coefficient value, the higher the feasibility of subsequent image processing. The feasibility model for the process is expressed as follows: in, This represents the feasibility coefficient for the treatment. This represents the background complexity index. Indicates the data acquisition accuracy index. Indicates the image compression ratio index. Represents the weight coefficient and The Furthermore, the larger the value, the better the feasibility of processing.

[0032] Background complexity (absolute difference in contrast between foreground and background) refers to the degree of difference in brightness or color between foreground patterns and background areas in an image. Its function is to quantify the degree of interference from the background on pattern feature extraction. High background complexity may lead to difficulties in accurately identifying pattern features. This value can be obtained by calculating the absolute difference between the average brightness or color value of pixels in the foreground area and the average color value of pixels in the background area, or by calculating the histogram difference or texture feature difference between the foreground and background areas after image segmentation. Digital acquisition accuracy (point cloud density) refers to the fineness of capturing surface details of an object during the digitization process. Especially for 3D scanning, point cloud density directly reflects the spatial sampling rate. Its function is to evaluate the ability of the original data to retain the geometric details of the pattern. The higher the point cloud density, the more complete the pattern details are preserved. This value can be directly obtained by the number of points acquired by the 3D scanning device per unit area, or indirectly reflected by evaluating the average distance between adjacent points in the point cloud data. Image compression ratio (or compression rate) refers to the degree to which image data is compressed during storage or transmission, typically expressed as the ratio of the original file size to the compressed file size. Its purpose is to quantify the amount of information that may be lost during image data compression. A high compression ratio may lead to the loss of texture details, affecting the accuracy of feature extraction. This value can be obtained by calculating the ratio of the original image file size to the compressed image file size, or indirectly reflected by evaluating the quality factor or bit rate of image compression algorithms (such as JPEG, WebP, etc.). Weighting coefficients are used to adjust the relative importance of different factors in the feasibility assessment of processing. For example, they can be set through expert experience or historical data analysis, or through machine learning methods such as regression analysis or optimization algorithms, trained and optimized based on actual feature extraction performance data.

[0033] As a specific implementation method, a feasibility assessment of feature extraction can be performed on an image of a Hunan-style pattern. First, the background complexity of the image is obtained; for example, by calculating the absolute difference in average brightness between the foreground pattern and the background area, a value of 0.65 is obtained. Simultaneously, it is assumed that the pattern was acquired through 3D scanning, with a point cloud data acquisition accuracy (point cloud density) of 1000 points / cm². Furthermore, the image is JPEG compressed, with a compression ratio of 1:10, meaning the original file size is 10 times the compressed file size. Next, these original parameters are subjected to max-min normalization. Assuming the normalization range of the background complexity is [0.1, 0.9], the background complexity index is... Assuming the normalized range of digital acquisition accuracy is [500, 2000] points per square centimeter, then the acquisition accuracy index is... =(1000-500) / (2000-500)=0.3333. Assuming the normalized range of the image compression ratio exponent is [0.05, 0.5] (where 0.05 represents high compression ratio and 0.5 represents low compression ratio, i.e., the higher the exponent, the better the compression quality), for a compression ratio of 1:10, the corresponding image compression ratio exponent is... This can be set to 0.3. Finally, these indices are imported into the feasibility model. Assume weighting coefficients... =0.4, =0.3, =0.3. Therefore, the feasibility coefficient is... The calculation results indicate that, under the current conditions of background complexity, digital acquisition accuracy, and image compression rate, the feasibility of extracting features from Hunan patterns is at a moderate level.

[0034] Through the above technical solution, this invention provides a systematic and quantifiable method to evaluate the feasibility of feature extraction for Hunan patterns. This method overcomes the limitations of traditional methods, which often suffer from inaccurate or incomplete evaluations, by comprehensively considering key factors such as background complexity, digital acquisition accuracy, and image compression rate. Therefore, it provides objective and dynamic quantitative indicators even under adverse conditions such as complex backgrounds, low acquisition accuracy, or high compression rates, significantly enhancing the reliability and robustness of feature extraction. This ensures high-quality pattern feature data can be obtained in various environments, laying a solid foundation for the digital display and evaluation of Hunan patterns.

[0035] Preferably, the step of obtaining the signal-to-noise contrast fitting degree through a signal-to-noise contrast fitting model based on the image signal-to-noise ratio under geometric fidelity and color fidelity, as well as the contrast of the pattern itself, is as follows: The image signal-to-noise ratio and the contrast of the pattern itself are subjected to maximum-min normalization to obtain the signal-to-noise ratio index and the contrast index of the pattern. The signal-to-noise ratio influence weight is obtained by processing the ratio of geometric fidelity to the sum of geometric fidelity and color fidelity. The weight of the influence of pattern contrast is obtained by processing the ratio of color fidelity to the sum of geometric fidelity and color fidelity. The signal-to-noise ratio (SNR) index, pattern contrast index, SNR influence weight, and pattern contrast influence weight are imported into the SNR contrast fitting model to obtain the SNR contrast fitting degree. The SNR contrast fitting model is expressed as follows: in, Indicates signal-to-noise ratio (SNR) adaptation. This represents the signal-to-noise ratio index. Indicates the contrast index of the pattern. This indicates that the signal-to-noise ratio affects the weights. The weighting of pattern contrast is indicated by the following. Furthermore, the higher the value, the better the signal-to-noise ratio and contrast are adapted to the current pattern features.

[0036] Image signal-to-noise ratio (SNR) is an important indicator of image quality, representing the relative intensity of image signal and noise. Its function is to quantify the ratio between effective and interfering information in an image; a high SNR usually indicates a clear image with rich details. Image SNR can be obtained in various ways. For example, it can be obtained by calculating the ratio of the average power of the image signal to the average power of the noise; or by using image processing software to estimate noise in the image and combining it with the image's brightness information to calculate the SNR. Pattern contrast itself refers to the degree of brightness or color difference between different regions of a fingerprint sample. Its function is to reflect the visual clarity and recognizability of the pattern; high contrast helps to highlight pattern details. Pattern contrast itself can be obtained in various ways. For example, it can be obtained by calculating the difference between the maximum and minimum brightness values ​​in the pattern and then dividing by their sum; or by using local contrast enhancement algorithms, such as histogram equalization or adaptive histogram equalization, to quantify and adjust the contrast. Max-min normalization is a data preprocessing technique used to linearly transform raw data to a specified range. Its function is to eliminate differences between data of different dimensions, making them comparable. This can be achieved by subtracting the minimum value of the dataset from each data point and then dividing by the difference between the maximum and minimum values; or by using Z-score standardization to convert the data into a distribution with a mean of 0 and a standard deviation of 1, followed by linear mapping to a specified range. The signal-to-noise ratio (SNR) index is the image SNR value after max-min normalization, typically between 0 and 1, used to provide a standardized SNR quantization value that can be compared with other metrics. The pattern contrast index is the contrast value of the pattern itself after max-min normalization, typically between 0 and 1, used to provide a standardized pattern contrast quantization value that can be compared with other metrics.

[0037] Geometric fidelity measures the consistency of a digital image with the original physical object's pattern in terms of geometric form. Its function is to evaluate the accuracy of the image in terms of shape, size, and proportion. Color fidelity measures the consistency of a digital image with the original physical object's pattern in terms of color representation. Its function is to evaluate the accuracy, saturation, and hue of the image's colors. Ratio processing is a mathematical operation used to calculate the proportional relationship between two values. Its function is to quantify the relative contribution or importance of a part to the whole. Signal-to-noise ratio (SNR) influence weight is a coefficient used to adjust the influence of the SNR index in the SNR contrast fitting model. Its function is to dynamically allocate the weight of the SNR in the fitting calculation based on the relative importance of geometric fidelity in the overall fidelity. Pattern contrast influence weight is a coefficient used to adjust the influence of the pattern contrast index in the SNR contrast fitting model. Its function is to dynamically allocate the weight of pattern contrast in the fitting calculation based on the relative importance of color fidelity in the overall fidelity. The signal-to-noise ratio (SNR) contrast ratio (SNR) index, pattern contrast ratio (BCR) index, and their respective weights are used to calculate the SNR contrast ratio. Its purpose is to provide a quantitative indicator to evaluate the degree of matching between the image's SNR and contrast and the current pattern features.

[0038] As a specific implementation method, when performing digital feature extraction on Hunan-style patterns, the original signal-to-noise ratio (SNR) and the contrast of the pattern itself in the image to be processed can be obtained first. For example, the peak signal-to-noise ratio (PSNR) of the image can be calculated using image processing software, and the contrast of the pattern itself can be quantified by calculating the ratio of the standard deviation of the gray levels of the pattern region to the average gray value. Next, these original values ​​are subjected to max-min normalization processing; for example, the PSNR value is mapped to the [0,1] interval, and the contrast value is also mapped to the [0,1] interval, thus obtaining the SNR index and the pattern contrast index. Simultaneously, the geometric fidelity and color fidelity of the image need to be obtained. For example, geometric fidelity can be calculated using a preset model based on the image scanning resolution, lens focal length, and surface curvature of the subject to obtain a value between 0 and 1; color fidelity can be calculated using another preset model based on color reproduction accuracy, file color depth, and spectral resolution to obtain a value between 0 and 1. Subsequently, these fidelity values ​​are used to calculate weights: the signal-to-noise ratio (SNR) influence weight is obtained by dividing the geometric fidelity value by the sum of the geometric fidelity value and the color fidelity value; the pattern contrast influence weight is obtained by dividing the color fidelity value by the sum of the geometric fidelity value and the color fidelity value. Finally, the calculated SNR index, pattern contrast index, SNR influence weight, and pattern contrast influence weight are substituted into the SNR contrast fitting model to calculate the SNR contrast fitting degree of the current image. For example, if the signal-to-noise ratio index is 0.8, the pattern contrast index is 0.7, the influence weight of the signal-to-noise ratio is 0.6, and the influence weight of the pattern contrast is 0.4, then the fit is... This calculated signal-to-noise ratio (SNR) is suitable for the desired fit. This can then be used as a parameter input into the illumination optimization model to guide the optimization of the overall grayscale standard deviation of the image under the target illumination condition, ensuring that the final feature extraction effect can better adapt to the specific characteristics of the Hunan pattern.

[0039] Through the above technical solution, this invention can dynamically evaluate the adaptability of the image's signal-to-noise ratio and pattern contrast to the current pattern features. Specifically, by normalizing the image's signal-to-noise ratio and the pattern's own contrast, and dynamically calculating the influence weights based on geometric fidelity and color fidelity, the calculation of signal-to-noise contrast adaptability becomes more comprehensive and accurate. This dynamic adaptability mechanism solves the problem of traditional methods having a single lighting optimization strategy that cannot effectively adapt to different pattern features. By introducing this adaptability into the lighting optimization model, the target lighting conditions can be adaptively adjusted according to the actual image quality and pattern characteristics, thereby significantly improving the accuracy, robustness, and display effect of Hunan pattern feature extraction, ensuring high-quality digital presentation in various complex environments.

[0040] Preferably, the illumination optimization model is expressed as: in, This represents the standard deviation of the overall grayscale values ​​of the image under the target lighting conditions. This represents the standard deviation of the overall grayscale values ​​of the image under the current lighting conditions. Indicates the target signal-to-noise ratio fit. Indicates signal-to-noise ratio (SNR) adaptation. This represents the feasibility coefficient for the treatment.

[0041] in, The parameter represents the degree of dispersion of all pixel grayscale values ​​in the image under optimized ideal lighting conditions, reflecting the overall brightness distribution and contrast range of the image under optimal lighting conditions, serving as the ultimate goal of lighting adjustment. The parameter represents the degree of dispersion of all pixel grayscale values ​​in the image during actual acquisition or processing, reflecting the influence of the current ambient light on the image brightness distribution and contrast, and serving as a benchmark for adjusting the lighting optimization model. The parameter represents the optimal match between the image signal-to-noise ratio and the contrast of the pattern itself under ideal conditions. It is a preset, desired performance metric used to guide lighting optimization. The parameter represents the actual matching degree between the signal-to-noise ratio of the current image and the contrast of the pattern itself. It quantifies the comprehensive performance of image quality and pattern feature clarity and is an important input for the dynamic adjustment of the illumination optimization model. The coefficient is used to measure the ease or efficiency of image digitization, storage and processing. It takes into account factors such as background complexity, digitization accuracy and image compression rate, and plays an adjustment role in illumination optimization to ensure that the illumination adjustment scheme is feasible in actual operation.

[0042] The model uses the standard deviation of the gray values ​​of the current image. Based on this, through the exponential function To reflect the current signal-to-noise contrast ratio Compatibility with target signal-to-noise ratio The difference between them necessitates adjustments to the lighting. This is based on the current signal-to-noise ratio. Off-target signal-to-noise contrast fit At this time, the exponential term will correspondingly amplify or reduce the magnitude of the illumination adjustment, thereby allowing the illumination conditions to better adapt to the needs of pattern feature extraction. For example, if the signal-to-noise contrast ratio of the current image... A low value indicates poor image quality or texture clarity. The model will guide the lighting adjustment in a direction that is more conducive to improving these indicators through the exponential term. Simultaneously, the feasibility coefficient is processed. The introduction of this model allows the lighting optimization process to consider both the ease of practical operation. If the image processing feasibility is low, the model will appropriately limit the aggressiveness of lighting adjustments to avoid introducing new processing challenges or reducing efficiency. In this way, the lighting optimization model can not only dynamically adjust lighting conditions based on the image's own quality and the adaptability of pattern features, but also consider the constraints of actual processing, thereby ensuring the robustness of the feature extraction process and the optimization of the final display effect. This model enables the acquisition of Hunan pattern features to comprehensively consider multiple dimensions such as geometric fidelity, color fidelity, signal-to-noise contrast adaptability, and processing feasibility, forming an adaptive lighting adjustment mechanism. This effectively solves the problems of unstable feature extraction and poor display effect caused by changes in lighting conditions in traditional methods.

[0043] The following is a concrete example. When digitally acquiring a Hunan-style pattern, the aforementioned lighting optimization model can be used to guide the adjustment of lighting conditions. For example, firstly, an image of the Hunan-style pattern under the current lighting conditions is acquired using an image sensor, and then the standard deviation of the overall grayscale value of the image under the current lighting conditions is calculated. Simultaneously, based on the complexity of the pattern and the expected display effect, a target signal-to-noise contrast ratio is set. For example, it can be set to 0.8. Next, by analyzing the current image, the current signal-to-noise contrast ratio is calculated. and processing feasibility coefficient Signal-to-noise ratio adaptation This can be obtained by evaluating the signal-to-noise ratio of the image and the contrast of the pattern, while the processing feasibility coefficient... This can be calculated based on factors such as background complexity, digital acquisition accuracy, and image compression rate. Assuming the currently calculated... It is 50. It is 0.6. The value is 0.9. Substitute these values ​​into the lighting optimization model. Calculated from Approximately 36.8. This... This value indicates the target level to which the overall grayscale standard deviation of the image should be adjusted to achieve better feature extraction results. Subsequently, based on this target value... This involves adjusting the actual shooting lighting environment. For example, this can be done by increasing or decreasing the intensity of the light source, adjusting the angle of the light source, or using soft lighting equipment, so that the overall grayscale standard deviation of the captured image gradually approaches a certain value. During the adjustment process, the standard deviation of the image's grayscale values ​​can be monitored in real time and compared with... The comparisons are made until the preset error range is reached. Through this iterative adjustment, it is possible to ensure that the feature extraction of Hunan patterns is carried out under optimal lighting conditions, thereby obtaining high-quality digital pattern data.

[0044] Through the above technical solution, the illumination optimization model of the present invention can dynamically adjust the target illumination conditions according to the signal-to-noise ratio and processing feasibility of the image, thereby effectively solving the problems of insufficient robustness of feature extraction and poor display effect caused by changes in illumination conditions in traditional methods.

[0045] As a preferred embodiment of the technical solution of the present invention, a feature extraction system 10 for Hunan patterns in a display and evaluation framework is also provided. The feature extraction system 10 for Hunan patterns in a display and evaluation framework includes: The geometric analysis module 11 is used to obtain geometric fidelity coefficients based on the image scanning resolution, lens focal length, and surface curvature of the subject through a geometric fidelity model. Color analysis module 12 is used to obtain color fidelity based on color reproduction accuracy, file color depth and spectral resolution through a color fidelity model; The processing and analysis module 13 is used to obtain the processing feasibility coefficient based on the background complexity, digital acquisition accuracy and image compression rate through a processing feasibility model. The adaptation degree acquisition module 14 is used to obtain the signal-to-noise contrast adaptation degree based on the image signal-to-noise ratio and the contrast of the pattern itself under geometric fidelity and color fidelity. The standard deviation extraction module 15 is used to obtain the standard deviation of the overall gray value of the image under the target illumination condition through the illumination optimization model based on the signal-to-noise contrast adaptation, processing feasibility coefficient and the standard deviation of the overall gray value of the image under the current illumination condition.

[0046] Furthermore, the geometric analysis module 11 includes: The information acquisition unit is used to acquire image scanning resolution, lens focal length, and surface curvature of the subject. The resolution deviation analysis unit is used to process the ratio of the absolute difference between the image scanning resolution and the reference resolution to the reference resolution to obtain the resolution deviation index. The focal length deviation analysis unit is used to compare the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length to obtain the focal length deviation index. The first normalization processing unit is used to perform maximum-minimum normalization processing on the surface curvature of the photographed object to obtain the curvature index. The fidelity output unit is used to import the resolution deviation index, focal length deviation index and curvature index into the geometric fidelity model to obtain the geometric fidelity. Among them, the larger the geometric fidelity value, the better the fidelity of the geometric features.

[0047] Specifically, the color analysis module 12 includes: The color information acquisition unit is used to acquire color reproduction accuracy, file color depth, and spectral resolution. The data conversion unit is used to perform maximum-min normalization processing on color reproduction accuracy, file color depth, and spectral resolution to obtain color accuracy index, file color depth index, and spectral resolution index. The second normalization processing unit is used to import the color accuracy index, file color depth index and spectral resolution index into the color fidelity model to obtain color fidelity. The color fidelity model calculates color fidelity using a weighted fusion function; the higher the color fidelity value, the better the color reproduction fidelity.

[0048] Furthermore, the processing and analysis module 13 includes: The background information acquisition unit is used to acquire background complexity, digital acquisition accuracy, and image compression rate. The third normalization processing unit is used to perform maximum-minimum normalization processing on the background complexity, digital acquisition accuracy and image compression rate to obtain the background complexity index, acquisition accuracy index and image compression rate index. The feasibility output unit is used to import the background complexity index, acquisition accuracy index, and image compression rate index into the processing feasibility model to obtain the processing feasibility coefficient. The higher the processing feasibility coefficient value, the higher the feasibility of image processing.

[0049] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for feature extraction of Hunan-style patterns in a display evaluation framework, characterized in that, The method includes: Based on image scanning resolution, lens focal length, and the surface curvature of the subject, geometric fidelity coefficients are obtained through a geometric fidelity model. Based on color reproduction accuracy, file color depth, and spectral resolution, color fidelity is obtained through a color fidelity model. Based on background complexity, digital acquisition accuracy, and image compression rate, processing feasibility coefficients are obtained through a processing feasibility model. Based on the image signal-to-noise ratio under geometric fidelity and color fidelity, as well as the contrast of the pattern itself, the signal-to-noise contrast adaptation degree is obtained through a signal-to-noise contrast adaptation model. Based on the signal-to-noise ratio, processing feasibility coefficient, and the standard deviation of the overall grayscale value of the image under the current illumination conditions, the standard deviation of the overall grayscale value of the image under the target illumination conditions is obtained through an illumination optimization model.

2. The feature extraction method for Hunan-style patterns in the display and evaluation framework according to claim 1, characterized in that, The step of obtaining the geometric fidelity coefficients based on the image scanning resolution, lens focal length, and surface curvature of the subject through a geometric fidelity model includes: Acquire image scanning resolution, lens focal length, and the surface curvature of the subject; The resolution deviation index is obtained by comparing the absolute difference between the image scanning resolution and the reference resolution with the reference resolution. The focal length deviation index is obtained by comparing the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length. The curvature index is obtained by performing maximum-min normalization on the surface curvature of the photographed object. The resolution deviation index, focal length deviation index, and curvature index are imported into the geometric fidelity model to obtain geometric fidelity. Among them, the larger the geometric fidelity value, the better the fidelity of the geometric features.

3. The feature extraction method for Hunan-style patterns in the display and evaluation framework according to claim 2, characterized in that, The steps for obtaining color fidelity based on color reproduction accuracy, file color depth, and spectral resolution using a color fidelity model include: Obtain color reproduction accuracy, file color depth, and spectral resolution; The color accuracy, file color depth, and spectral resolution are subjected to maximum-min normalization to obtain the color accuracy index, file color depth index, and spectral resolution index. The color accuracy index, file color depth index, and spectral resolution index are imported into the color fidelity model to obtain the color fidelity. The color fidelity model calculates color fidelity using a weighted fusion function; the higher the color fidelity value, the better the color reproduction fidelity.

4. The feature extraction method for Hunan-style patterns in the display and evaluation framework according to claim 3, characterized in that, The step of obtaining the processing feasibility coefficient based on background complexity, digital acquisition accuracy, and image compression rate through a processing feasibility model includes: Obtain background complexity, digital acquisition accuracy, and image compression rate; The background complexity, digital acquisition accuracy, and image compression rate are subjected to maximum-minimum normalization to obtain the background complexity index, acquisition accuracy index, and image compression rate index. The background complexity index, acquisition accuracy index, and image compression rate index are imported into the processing feasibility model to obtain the processing feasibility coefficient. The higher the processing feasibility coefficient value, the higher the feasibility of image processing.

5. The feature extraction method for Hunan-style patterns in the display and evaluation framework according to claim 4, characterized in that, The step of obtaining the signal-to-noise contrast fitting degree through a signal-to-noise contrast fitting model based on the image signal-to-noise ratio under geometric fidelity and color fidelity, and the contrast of the pattern itself, includes: The image signal-to-noise ratio and the contrast of the pattern itself are subjected to maximum-min normalization to obtain the signal-to-noise ratio index and the contrast index of the pattern. The signal-to-noise ratio influence weight is obtained by processing the ratio of geometric fidelity to the sum of geometric fidelity and color fidelity. The weight of the influence of pattern contrast is obtained by processing the ratio of color fidelity to the sum of geometric fidelity and color fidelity. The signal-to-noise ratio (SNR) index, pattern contrast index, SNR influence weight, and pattern contrast influence weight are imported into the SNR contrast fitting model to obtain the SNR contrast fitting degree. The SNR contrast fitting model is expressed as follows: ； in, Indicates signal-to-noise ratio (SNR) adaptation. This represents the signal-to-noise ratio index. Indicates the contrast index of the pattern. This indicates that the signal-to-noise ratio affects the weights. The weighting of pattern contrast is indicated by the following. Furthermore, the higher the value, the better the signal-to-noise ratio and contrast are adapted to the current pattern features.

6. The feature extraction method for Hunan-style patterns in a display evaluation framework according to any one of claims 1 to 5, characterized in that, The illumination optimization model is expressed as follows: in, This represents the standard deviation of the overall grayscale values ​​of the image under the target lighting conditions. This represents the standard deviation of the overall grayscale values ​​of the image under the current lighting conditions. Indicates the target signal-to-noise ratio fit. Indicates signal-to-noise ratio (SNR) adaptation. This represents the feasibility coefficient for the treatment.

7. A feature extraction system for Hunan-style patterns in a display and evaluation framework, characterized in that, The system includes: The geometric analysis module is used to obtain geometric fidelity coefficients based on image scanning resolution, lens focal length, and the surface curvature of the subject through a geometric fidelity model. The color analysis module is used to obtain color fidelity based on color reproduction accuracy, file color depth, and spectral resolution through a color fidelity model. The processing and analysis module is used to obtain the processing feasibility coefficient based on the background complexity, digital acquisition accuracy, and image compression rate through a processing feasibility model. The adaptation degree acquisition module is used to obtain the signal-to-noise contrast adaptation degree based on the image signal-to-noise ratio and the contrast of the pattern itself under geometric fidelity and color fidelity. The standard deviation extraction module is used to obtain the standard deviation of the overall gray value of the image under the target lighting condition based on the signal-to-noise contrast adaptation, processing feasibility coefficient, and the standard deviation of the overall gray value of the image under the current lighting condition, through the lighting optimization model.

8. The feature extraction system for Hunan-style patterns in the display and evaluation framework according to claim 7, characterized in that, The geometric analysis module includes: The information acquisition unit is used to acquire image scanning resolution, lens focal length, and surface curvature of the subject. The resolution deviation analysis unit is used to process the ratio of the absolute difference between the image scanning resolution and the reference resolution to the reference resolution to obtain the resolution deviation index. The focal length deviation analysis unit is used to compare the absolute difference between the lens focal length and the optimal focal length with the allowable deviation from the optimal focal length to obtain the focal length deviation index. The first normalization processing unit is used to perform maximum-minimum normalization processing on the surface curvature of the photographed object to obtain the curvature index. The fidelity output unit is used to import the resolution deviation index, focal length deviation index and curvature index into the geometric fidelity model to obtain the geometric fidelity. Among them, the larger the geometric fidelity value, the better the fidelity of the geometric features.

9. The feature extraction system for Hunan-style patterns in the display and evaluation framework according to claim 8, characterized in that, The color analysis module includes: The color information acquisition unit is used to acquire color reproduction accuracy, file color depth, and spectral resolution. The data conversion unit is used to perform maximum-min normalization processing on color reproduction accuracy, file color depth, and spectral resolution to obtain color accuracy index, file color depth index, and spectral resolution index. The second normalization processing unit is used to import the color accuracy index, file color depth index and spectral resolution index into the color fidelity model to obtain color fidelity. The color fidelity model calculates color fidelity using a weighted fusion function; the higher the color fidelity value, the better the color reproduction fidelity.

10. The feature extraction system for Hunan-style patterns in the display and evaluation framework according to claim 9, characterized in that, The processing and analysis module includes: The background information acquisition unit is used to acquire background complexity, digital acquisition accuracy, and image compression rate. The third normalization processing unit is used to perform maximum-minimum normalization processing on the background complexity, digital acquisition accuracy and image compression rate to obtain the background complexity index, acquisition accuracy index and image compression rate index. The feasibility output unit is used to import the background complexity index, acquisition accuracy index, and image compression rate index into the processing feasibility model to obtain the processing feasibility coefficient. The higher the processing feasibility coefficient value, the higher the feasibility of image processing.

Images