A method and system for analyzing the color characteristics of a silk floss knitted short-process printed fabric

Abstract

The present invention relates to the technical field of image processing, and specifically provides a method and system for analyzing the color characteristics of a silk floss knitted short-process printed fabric, including: obtaining a first multi-spectral image and a second high-resolution RGB image of the silk floss knitted short-process printed fabric, and obtaining a multi-modal printed fabric image through cross-modal fusion; separating the low-frequency color base and high-frequency knitted texture noise through dual-tree complex wavelet transform combined with an attention mechanism and gradient perception filtering, extracting the image color feature matrix, and further obtaining multivariate image color feature coefficients, including color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient, and color saturation balance coefficient; finally, calculating the quality coefficient of the printed fabric based on these color feature coefficients and adjusting the printing process parameters. The present invention can effectively improve the quality and production efficiency of the silk floss knitted short-process printed fabric.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a method and system for analyzing the color characteristics of a silk floss knitted short-process printed fabric. Background Art

[0002] Image processing technology has been widely applied to the quality control of silk floss knitted short-process printed fabrics; therefore, it is of great significance to perform quality analysis and production control on silk floss knitted short-process printed fabrics through image analysis; however, the traditional methods have the following problems in fabric quality analysis and production control: The traditional methods do not fully consider the comprehensive analysis of multi-spectral images and high-resolution RGB images; the traditional image filtering methods do not fully consider the comprehensive analysis of dual-tree complex wavelet transform, attention mechanism, and gradient-aware filtering, which easily leads to insufficient efficiency in extracting color characteristics; the traditional methods often analyze through single-modal images and often analyze according to single-image color characteristic parameters, making it difficult to achieve a comprehensive and overall analysis of fabric quality and working parameters, thus resulting in limited fabric quality and production efficiency.

[0003] Therefore, a method and system for analyzing the color characteristics of a silk floss knitted short-process printed fabric are proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for analyzing the color characteristics of a silk floss knitted short-process printed fabric. By obtaining the first multi-spectral image and the second high-resolution RGB image of the silk floss knitted short-process printed fabric, and performing cross-modal fusion to obtain a multi-modal printed fabric image; separating the low-frequency color base and high-frequency knitted texture noise through dual-tree complex wavelet transform combined with attention mechanism and gradient-aware filtering, extracting the image color characteristic matrix, and further obtaining multi-element image color characteristic coefficients, including color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient, and color saturation balance coefficient; finally, calculating the quality coefficient of the printed fabric based on these color characteristic coefficients and adjusting the printing process parameters. The present invention can effectively improve the quality and production efficiency of silk floss knitted short-process printed fabrics.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A method for analyzing the color characteristics of a silk floss knitted short-process printed fabric, comprising:

[0007] S1. Obtain the first multi-spectral image and the second high-resolution RGB image of the silk floss knitted short-process printed fabric, and perform cross-modal fusion to obtain a multi-modal printed fabric image;

[0008] S2. Perform dual-tree complex wavelet transform on the multi-modal printed fabric image, combine the attention mechanism with gradient-aware filtering to separate the low-frequency color base and high-frequency knitted texture noise, and obtain the image color feature matrix of the printed fabric;

[0009] S3. Analyze the image color feature matrix to obtain multi-variate image color feature coefficients; the multi-variate image color feature coefficients include color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient, and color saturation balance coefficient;

[0010] S4. Obtain the first printed fabric quality coefficient according to the multi-variate image color feature coefficients and adjust the printing process parameters.

[0011] Preferably, the acquisition process of the multi-modal printed fabric image includes: using the pre-trained ResNet-18 to extract features from the first multi-spectral image to obtain the first multi-spectral feature map; using the pre-trained VGG-16 to extract features from the second high-resolution RGB image to obtain the second high-resolution RGB feature map, and performing cross-modal alignment, attention-weighted fusion, and upsampling on the first multi-spectral feature map and the second high-resolution RGB feature map based on the attention weights, and finally generating the multi-modal printed fabric image; the multi-modal printed fabric image is:

[0012] ;

[0013] Among them, is the multi-modal printed fabric image; is the multi-spectral image attention weight; is the first multi-spectral feature map; is the second high-resolution RGB feature map.

[0014] Preferably, the specific acquisition process of the image color feature matrix is:

[0015] Perform 3-layer decomposition on the multi-modal printed fabric image through dual-tree complex wavelet transform to obtain the low-frequency component and 6 high-frequency sub-bands in different directions; calculate the 5×5 neighborhood energy based on each high-frequency sub-band, and dynamically assign the attention weights of each sub-band based on the softmax function; calculate the gradient magnitude based on the low-frequency component; and correct each high-frequency sub-band according to the gradient magnitude and the sub-band attention weights; finally, perform inverse transform reconstruction based on the corrected high-frequency sub-bands and the low-frequency component to obtain the image color feature matrix.

[0016] Preferably, the color frequency domain complexity coefficient is obtained by performing a Fourier transform on the LAB color space of the image color feature matrix to obtain an amplitude spectrum, and performing circular frequency band energy statistics based on the amplitude spectrum to obtain the color frequency domain complexity coefficient; the color uniformity coefficient is obtained by dividing the image color feature matrix into multiple windows, performing multi-scale covariance analysis of the LAB color space for each window, and performing maximum eigenvalue mapping analysis; the color texture coordination coefficient is obtained by extracting a gray-level co-occurrence matrix from the L channel of the image color feature matrix and obtaining contrast and correlation analysis; the color stability coefficient is obtained based on the image color feature matrix to obtain a color difference field and an edge weight matrix, and analyzing according to the color difference field and the edge weight matrix; the color saturation balance coefficient is obtained by calculating the saturation, partitioning according to the saturation, and then analyzing according to the pixel ratio of each region.

[0017] Preferably, the first printed fabric quality coefficient is:

[0018] ;

[0019] Wherein, represents the first printed fabric quality coefficient; represents the th multi-modal image color feature coefficient; represents the exponential adjustment coefficient.

[0020] A color feature analysis system for silk-cotton knitted short-process printed fabrics, comprising:

[0021] A multi-modal image acquisition and fusion module, configured to acquire a first multi-spectral image and a second high-resolution RGB image of the silk-cotton knitted short-process printed fabric, and perform cross-modal fusion to obtain a multi-modal printed fabric image;

[0022] A noise reduction and image color feature extraction module, configured to perform a dual-tree complex wavelet transform on the multi-modal printed fabric image, combine an attention mechanism and gradient perception filtering to separate the low-frequency color base and high-frequency knitted texture noise, and obtain an image color feature matrix of the printed fabric;

[0023] An image color feature analysis module, configured to analyze the image color feature matrix to obtain multi-modal image color feature coefficients; the multi-modal image color feature coefficients include a color frequency domain complexity coefficient, a color uniformity coefficient, a color texture coordination coefficient, a color stability coefficient, and a color saturation balance coefficient;

[0024] A quality analysis and parameter adjustment module, configured to obtain a first printed fabric quality coefficient according to the multi-modal image color feature coefficients, and perform parameter adjustment during the printing process.

[0025] Compared with the prior art, the beneficial effects of the present invention are:

[0026] 1. The present invention obtains a multimodal printed fabric image through a first multispectral image and a second high-resolution RGB image, and performs cross-modal fusion. Through the multimodal printed fabric image, the color characteristics of the printed fabric can be comprehensively analyzed, and then the quality analysis of the silk-cotton knitted short-process printed fabric can be comprehensively carried out, thereby effectively improving the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0027] 2. The present invention performs dual-tree complex wavelet transform on the multimodal printed fabric image, combines the attention mechanism with gradient-aware filtering to separate the low-frequency color base and high-frequency knitted texture noise, and generates an image color feature matrix. Through the dual-tree complex wavelet transform, the attention mechanism and gradient-aware filtering, color features can be effectively extracted, providing effective image preprocessing for comprehensively analyzing the color characteristics of the printed fabric, and then the quality analysis of the silk-cotton knitted short-process printed fabric can be comprehensively carried out, thereby effectively improving the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0028] 3. The present invention analyzes based on the image color feature matrix, extracts multivariate image color feature coefficients, including color frequency domain complexity coefficients, color uniformity coefficients, color texture coordination coefficients, color stability coefficients, and color saturation balance coefficients, and obtains a first printed fabric quality coefficient according to the multivariate image color feature coefficients, which can comprehensively perform the quality analysis of the silk-cotton knitted short-process printed fabric, thereby effectively improving the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0029] 4. The present invention performs quality analysis based on the first printed fabric quality coefficient, extracts the correlation matrix between the printed fabric quality coefficient and the printing process parameters, and adjusts the printing process parameters based on the correlation matrix, the target quality coefficient, and the first printed fabric quality coefficient, which can effectively improve the quality and production efficiency of the silk-cotton knitted short-process printed fabric. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of a method for analyzing the color characteristics of a silk-cotton knitted short-process printed fabric provided by an embodiment of the present invention;

[0031] Figure 2 It is a schematic diagram of the acquisition process of a first printed fabric quality coefficient provided by an embodiment of the present invention;

[0032] Figure 3 It is a schematic diagram of the structure of a system for analyzing the color characteristics of a silk-cotton knitted short-process printed fabric provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0033] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0034] Embodiment 1

[0035] In order to effectively improve the quality and production efficiency of the silk knitted short-process printed fabric on production line A, a method for analyzing the color characteristics of the silk knitted short-process printed fabric is applied;

[0036] Reference Figure 1 , which is a schematic flowchart of a method for analyzing the color characteristics of a silk knitted short-process printed fabric provided by an embodiment of the present invention, includes:

[0037] A method for analyzing the color characteristics of a silk knitted short-process printed fabric includes:

[0038] S1. Obtain the first multispectral image and the second high-resolution RGB image of the silk knitted short-process printed fabric, and perform cross-modal fusion to obtain a multimodal printed fabric image; further, the first multispectral image is obtained by a multispectral camera; the second high-resolution RGB image is obtained by a high-resolution camera;

[0039] Further, the acquisition process of the multimodal printed fabric image includes: using the pre-trained ResNet-18 to extract features from the first multispectral image to obtain the first multispectral feature map; using the pre-trained VGG-16 to extract features from the second high-resolution RGB image to obtain the second high-resolution RGB feature map, and performing cross-modal alignment, attention-weighted fusion, and upsampling on the first multispectral feature map and the second high-resolution RGB feature map based on the attention weight to finally generate a multimodal printed fabric image; the multimodal printed fabric image is:

[0040] ;

[0041] Among them, is the multimodal printed fabric image; is the multispectral image attention weight; is the first multispectral feature map; is the second high-resolution RGB feature map.

[0042] S2. Perform dual-tree complex wavelet transform on the multimodal printed fabric image, and combine the attention mechanism and gradient perception filtering to separate the low-frequency color base and high-frequency knitted texture noise to obtain the image color feature matrix of the printed fabric;

[0043] In this embodiment, a multimodal printed fabric image is obtained by using a first multispectral image and a second high-resolution RGB image and performing cross-modal fusion. Through the multimodal printed fabric image, the color characteristics of the printed fabric can be comprehensively analyzed, and then the quality analysis of the silk-cotton knitted short-process printed fabric can be comprehensively carried out, thereby effectively improving the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0044] Further, the specific process for obtaining the image color feature matrix is as follows:

[0045] The multimodal printed fabric image is decomposed into 3 layers by using the dual-tree complex wavelet transform to obtain a low-frequency component and 6 high-frequency subbands in different directions. The 5×5 neighborhood energy is calculated based on each high-frequency subband, and the attention weights of each subband are dynamically allocated based on the softmax function. The gradient magnitude is calculated based on the low-frequency component. Each high-frequency subband is corrected according to the gradient magnitude and the subband attention weights. Finally, an inverse transform reconstruction is performed based on the corrected high-frequency subbands and the low-frequency component to obtain the image color feature matrix. 。

[0046] In this embodiment, by performing the dual-tree complex wavelet transform on the multimodal printed fabric image, combining the attention mechanism and the gradient-aware filtering, separating the low-frequency color base and the high-frequency knitted texture noise, and generating the image color feature matrix. Through the dual-tree complex wavelet transform, the attention mechanism and the gradient-aware filtering, the color features can be effectively extracted, providing effective image preprocessing for comprehensively analyzing the color characteristics of the printed fabric, and then the quality analysis of the silk-cotton knitted short-process printed fabric can be comprehensively carried out, thereby effectively improving the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0047] S3. Analyze the image color feature matrix to obtain multivariate image color feature coefficients. The multivariate image color feature coefficients include a color frequency domain complexity coefficient, a color uniformity coefficient, a color texture coordination coefficient, a color stability coefficient, and a color saturation balance coefficient.

[0048] Further, the color frequency domain complexity coefficient is obtained by performing a Fourier transform on the LAB color space of the image color feature matrix to obtain an amplitude spectrum, and performing circular frequency band energy statistics based on the amplitude spectrum to obtain the color frequency domain complexity coefficient.

[0049] Further, the LAB color space includes an L channel, an a channel, and a b channel.

[0050] Further, the amplitude spectrum is obtained by 2D-FFT. ,where represents the frequency spectrum, and the amplitude spectrum is divided into 5 concentric rings, and the energy ratio of the amplitude spectrum within each ring is calculated. ,and based on Obtain the color frequency domain complexity coefficient;

[0051] The color frequency domain complexity coefficient is:

[0052] ;

[0053] Wherein, represents the color frequency domain complexity coefficient; represents the exponential function with the constant as the base;

[0054] The color uniformity coefficient is obtained by performing multi-window partitioning on the image color feature matrix, performing multi-scale covariance analysis in the LAB color space for each window, and performing maximum eigenvalue mapping analysis;

[0055] Further, the image color feature matrix is divided into 16×16 windows. Calculate the covariance matrix for the a and b channels of each window ; represents the variance of the a channel; represents the variance of the b channel; represents the covariance between the a channel and the b channel; and perform maximum eigenvalue mapping to obtain , specifically:

[0056] ;

[0057] Further, obtain the color uniformity coefficient according to the maximum eigenvalue; the color uniformity coefficient is:

[0058] ;

[0059] Wherein, represents the color uniformity coefficient; represents the exponential function with the constant as the base; represents the maximum eigenvalue; represents the adjustment parameter;

[0060] The color texture coordination coefficient is obtained by extracting the gray-level co-occurrence matrix from the L channel of the image color feature matrix and obtaining the contrast and correlation analysis in each direction; each direction includes 0°, 45°, 90°, and 135°;

[0061] Further, extract the gray-level co-occurrence matrix from the L channel of the image color feature matrix ; calculate the contrast and correlation features based on the gray-level co-occurrence matrix; obtain the color texture coordination coefficient based on the contrast and correlation; the color texture coordination coefficient is:

[0062] ;

[0063] Among them, represents the color texture coordination coefficient; represents the correlation degree, reflecting the dependence degree between pixels; represents the contrast;

[0064] The color stability coefficient is obtained based on the image color feature matrix to obtain a color difference field and an edge weight matrix, and is analyzed according to the color difference field and the edge weight matrix;

[0065] Furthermore, based on the image color feature matrix and the reference template matrix to obtain a color difference field, and based on the color difference field and the edge weight matrix detected by the Sobel operator to obtain the color stability coefficient; the color stability coefficient is:

[0066] ;

[0067] Among them, represents the color stability coefficient; represents the color difference field matrix; represents the edge weight matrix;

[0068] The color saturation balance coefficient is obtained by calculating the saturation, partitioning according to the saturation, and then analyzing according to the pixel proportion of each region.

[0069] Furthermore, first calculate the saturation, and conduct interval distribution statistics according to the saturation, which are divided into a low saturation region, a medium saturation region, and a high saturation region; then obtain the pixel proportion of each region, and obtain the color saturation balance coefficient according to the pixel proportion of each region; the color saturation balance coefficient is:

[0070] ;

[0071] Among them, represents the color saturation balance coefficient; , and respectively represent the pixel proportions of the low saturation region, the medium saturation region, and the high saturation region;

[0072] This embodiment analyzes based on the image color feature matrix, extracts multi - element image color feature coefficients. The multi - element image color feature coefficients include the color frequency domain complexity coefficient, the color uniformity coefficient, the color texture coordination coefficient, the color stability coefficient, and the color saturation balance coefficient, and obtains the first printed fabric quality coefficient according to the multi - element image color feature coefficients, which can comprehensively conduct quality analysis on the silk - cotton knitted short - process printed fabric, thereby effectively improving the quality and production efficiency of the silk - cotton knitted short - process printed fabric.

[0073] S4. Obtain the first printed fabric quality coefficient according to the multi-image color feature coefficients, and adjust the printing process parameters.

[0074] Further, the first printed fabric quality coefficient is:

[0075] ;

[0076] Wherein, represents the first printed fabric quality coefficient; represents the th multi-image color feature coefficient; represents the exponential adjustment coefficient.

[0077] Further, referring to Figure 2 is a schematic diagram of the acquisition process of the first printed fabric quality coefficient provided by an embodiment of the present invention;

[0078] The adjustment method of the printing process parameters is as follows: Based on a machine learning model, perform a correlation analysis on the first printed fabric quality coefficient and the printing process parameters of the silk-cotton knitted short-process printed fabric to obtain a correlation matrix, and adjust the printing process parameters based on the correlation matrix and the target quality coefficient to obtain the adjusted printing process parameters; the adjusted printing process parameters are:

[0079] ;

[0080] Wherein, represents the th improved printing process parameter; represents the th correlation coefficient between the printing fabric process parameter and the first printed fabric quality coefficient; represents the correlation matrix; represents the number of adjusted printing process parameters; represents the target quality coefficient; represents the first printed fabric quality coefficient; represents the th initial printing process parameter.

[0081] Further, the printing process parameters include printing speed and printing pressure.

[0082] This embodiment performs quality analysis based on the first printed fabric quality coefficient, extracts the correlation matrix between the printed fabric quality coefficient and the printing process parameters, and adjusts the printing process parameters based on the correlation matrix, the target quality coefficient, and the first printed fabric quality coefficient, which can effectively improve the quality and production efficiency of the silk-cotton knitted short-process printed fabric.

[0083] In this embodiment, the first multi-spectral image and the second high-resolution RGB image of the silk floss knitted short-process printed fabric are obtained, and a multi-modal printed fabric image is obtained through cross-modal fusion; the low-frequency color base and the high-frequency knitted texture noise are separated by the dual-tree complex wavelet transform combined with the attention mechanism and the gradient perception filter, and the image color feature matrix is extracted, and then the multi-source image color feature coefficients are obtained, including the color frequency domain complexity coefficient, the color uniformity coefficient, the color texture coordination coefficient, the color stability coefficient, and the color saturation balance coefficient; finally, the quality coefficient of the printed fabric is calculated based on these color feature coefficients, and the process parameters are adjusted. The present invention can effectively improve the quality and production efficiency of the silk floss knitted short-process printed fabric.

[0084] To verify the effectiveness of a method for analyzing the color characteristics of a silk floss knitted short-process printed fabric provided in this embodiment, different methods are applied to the silk floss knitted short-process printed fabric on production line A, including Method 1, Method 2, Method 3, Method 4, and Method 5; the quality qualification rate of production line A under different method applications is verified through comparison; Method 1 is a method for analyzing the color characteristics of a silk floss knitted short-process printed fabric provided in this embodiment; Method 2 does not consider the multi-modal image on the basis of Method 1; Method 3 does not consider noise reduction and image color feature extraction and analysis on the basis of Method 1; Method 4 does not consider the analysis of the first printed fabric quality coefficient on the basis of Method 1; Method 5 does not consider the adjustment of the printing process parameters on the basis of Method 1; the specific results are shown in Table 1;

[0085] Table 1 Comparison of the quality qualification rates of the silk floss knitted short-process printed fabric on production line A under different methods

[0086]

[0087] As can be seen from Table 1, a method for analyzing the color characteristics of a silk floss knitted short-process printed fabric proposed in this embodiment has a certain effectiveness and can effectively improve the quality and production efficiency of the silk floss knitted short-process printed fabric.

[0088] Embodiment 2

[0089] In order to effectively improve the quality and production efficiency of the silk floss knitted short-process printed fabric on production line B, a method for analyzing the color characteristics of a silk floss knitted short-process printed fabric is applied;

[0090] Refer to Figure 1 , which is a flow schematic diagram of a method for analyzing the color characteristics of a silk floss knitted short-process printed fabric provided in an embodiment of the present invention, including:

[0091] A method for analyzing the color characteristics of a silk floss knitted short-process printed fabric includes:

[0092] A multi-modal image acquisition and fusion module, which is used to obtain the first multi-spectral image and the second high-resolution RGB image of the silk knitted short process printed fabric, and perform cross-modal fusion to obtain a multi-modal printed fabric image; further, the first multi-spectral image is obtained by a multi-spectral camera; the second high-resolution RGB image is obtained by a high-resolution camera;

[0093] Further, the acquisition process of the multi-modal printed fabric image includes: using the pre-trained ResNet-18 to extract features from the first multi-spectral image to obtain the first multi-spectral feature map; extracting features from the second high-resolution RGB image according to the pre-trained VGG-16 to obtain the second high-resolution RGB feature map, and performing cross-modal alignment, attention-weighted fusion and upsampling on the first multi-spectral feature map and the second high-resolution RGB feature map based on the attention weight, and finally generating a multi-modal printed fabric image; the multi-modal printed fabric image is:

[0094] ;

[0095] Among them, is the multi-modal printed fabric image; is the multi-spectral image attention weight; is the first multi-spectral feature map; is the second high-resolution RGB feature map.

[0096] A noise reduction and image color feature extraction module, which is used to perform dual-tree complex wavelet transform on the multi-modal printed fabric image, and combine the attention mechanism and gradient perception filtering to separate the low-frequency color base and high-frequency knitted texture noise, and obtain the image color feature matrix of the printed fabric;

[0097] Further, the specific acquisition process of the image color feature matrix is:

[0098] Performing 3-layer decomposition on the multi-modal printed fabric image through dual-tree complex wavelet transform to obtain low-frequency components and 6 high-frequency sub-bands in different directions; calculating the 5×5 neighborhood energy based on each high-frequency sub-band, and dynamically allocating the attention weight of each sub-band based on the softmax function; calculating the gradient magnitude based on the low-frequency component; and correcting each high-frequency sub-band according to the gradient magnitude and the sub-band attention weight; finally, performing inverse transform reconstruction based on the corrected high-frequency sub-bands and low-frequency components to obtain the image color feature matrix .

[0099] An image color feature analysis module, which is used to analyze the image color feature matrix to obtain multi-element image color feature coefficients; the multi-element image color feature coefficients include color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient and color saturation balance coefficient;

[0100] Further, the color frequency domain complexity coefficient is obtained by performing a Fourier transform on the LAB color space of the image color feature matrix to obtain an amplitude spectrum, and performing circular frequency band energy statistics based on the amplitude spectrum to obtain the color frequency domain complexity coefficient;

[0101] Further, the LAB color space includes an L channel, an a channel, and a b channel;

[0102] Further, the amplitude spectrum is obtained by 2D-FFT , where represents the frequency spectrum, and the amplitude spectrum is divided into 5 concentric rings, and the proportion of the amplitude spectrum energy within each ring is calculated , and based on the color frequency domain complexity coefficient is obtained;

[0103] The color frequency domain complexity coefficient is:

[0104] ;

[0105] where, represents the color frequency domain complexity coefficient; represents the exponential function with a constant as the base;

[0106] The color uniformity coefficient is obtained by performing multi-window partitioning on the image color feature matrix, performing multi-scale covariance analysis on each window in the LAB color space, and performing maximum eigenvalue mapping analysis ;

[0107] Further, the image color feature matrix is divided into 16×16 windows. The covariance matrix is calculated for the a and b channels of each window ; represents the variance of the a channel; represents the variance of the b channel; represents the covariance between the a channel and the b channel; and maximum eigenvalue mapping is performed to obtain , specifically:

[0108] ;

[0109] Further, the color uniformity coefficient is obtained according to the maximum eigenvalue; the color uniformity coefficient is:

[0110] ;

[0111] where, represents the color uniformity coefficient; represents the exponential function with a constant Exponential function with a base of Denote the maximum eigenvalue; Denote the adjustment parameter;

[0112] The color texture coordination coefficient is obtained by extracting the gray-level co-occurrence matrix from the L channel of the image color feature matrix and analyzing the contrast and correlation in each direction; each direction includes 0°, 45°, 90°, and 135°;

[0113] Furthermore, from the image color feature matrix Extract the gray-level co-occurrence matrix from the L channel; calculate the contrast and correlation features based on the gray-level co-occurrence matrix; obtain the color texture coordination coefficient based on the contrast and correlation; the color texture coordination coefficient is:

[0114] ;

[0115] Where Denote the color texture coordination coefficient; Denote the correlation, reflecting the degree of dependence between pixels; Denote the contrast;

[0116] The color stability coefficient is obtained based on the image color feature matrix to obtain the color difference field and the edge weight matrix, and analyzed according to the color difference field and the edge weight matrix;

[0117] Furthermore, based on the image color feature matrix And the reference template matrix to obtain the color difference field, and obtain the color stability coefficient based on the color difference field and the edge weight matrix detected by the Sobel operator; the color stability coefficient is:

[0118] ;

[0119] Where Denote the color stability coefficient; Denote the color difference field matrix; Denote the edge weight matrix;

[0120] The color saturation balance coefficient is obtained by calculating the saturation, partitioning according to the saturation, and then analyzing according to the pixel proportion in each region.

[0121] Furthermore, first calculate the saturation, and conduct interval distribution statistics according to the saturation, divided into low saturation region, medium saturation region, and high saturation region; then obtain the pixel proportion in each region, and obtain the color saturation balance coefficient according to the pixel proportion in each region; the color saturation balance coefficient is:

[0122] ;

[0123] Where represents the color saturation balance coefficient; , and respectively represent the pixel proportion of the low saturation area, the medium saturation area and the high saturation area;

[0124] The quality analysis and parameter adjustment module is used to obtain the first printing fabric quality coefficient according to the multi - element image color feature coefficient and adjust the printing process parameters.

[0125] Further, the first printing fabric quality coefficient is:

[0126] ;

[0127] wherein, represents the first printing fabric quality coefficient; represents the th multi - element image color feature coefficient; represents the exponential adjustment coefficient.

[0128] Further, referring to Figure 2 is a schematic diagram of the acquisition process of a first printing fabric quality coefficient provided by an embodiment of the present invention;

[0129] The adjustment method of the printing process parameters is as follows: based on a machine learning model, a correlation analysis is performed on the first printing fabric quality coefficient and the printing process parameters of the silk - cotton knitted short - process printing fabric to obtain a correlation matrix, and the printing process parameters are adjusted based on the correlation matrix and the target quality coefficient to obtain the adjusted printing process parameters; the adjusted printing process parameters are:

[0130] ;

[0131] wherein, represents the th improved printing process parameter; represents the correlation coefficient between the th printing fabric process parameter and the first printing fabric quality coefficient; represents the correlation matrix; represents the number of adjusted printing process parameters; represents the target quality coefficient; represents the first printing fabric quality coefficient; represents the th initial printing process parameter.

[0132] Furthermore, the printing process parameters include printing speed and printing pressure. To verify the effectiveness of a color feature analysis system for silk-cotton knitted short-process printed fabrics provided in this embodiment, different systems are applied to the silk-cotton knitted short-process printed fabrics on production line B, including System 1, System 2, System 3, System 4, and System 5; verification is carried out by comparing the quality qualification rates of production line B under the application of different systems; System 1 is a color feature analysis system for silk-cotton knitted short-process printed fabrics provided in this embodiment; System 2 does not consider multimodal images based on System 1; System 3 does not consider noise reduction and image color feature extraction and analysis based on System 1; System 4 does not consider the analysis of the first printed fabric quality coefficient based on System 1; System 5 does not consider the adjustment of printing process parameters based on System 1; the specific results are shown in Table 2;

[0133] Table 2 Comparison of the quality qualification rates of the silk-cotton knitted short-process printed fabrics on production line B for different systems

[0134]

[0135] As can be seen from Table 2, a color feature analysis system for silk-cotton knitted short-process printed fabrics proposed in this embodiment has a certain effectiveness and can effectively improve the quality and production efficiency of silk-cotton knitted short-process printed fabrics.

[0136] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing color characteristics of silk-cotton knitted short-process printed fabrics, characterized in that: include: S1. Obtaining a first multispectral image and a second high-resolution RGB image of a silk-cotton knitted short-process printed fabric, and performing cross-modal fusion to obtain a multimodal printed fabric image; S2. Performing a dual-tree complex wavelet transform on the multimodal printed fabric image, combining the attention mechanism with the gradient-aware filter to separate the low-frequency color base and the high-frequency knitting texture noise, and obtaining the image color feature matrix of the printed fabric; the specific acquisition process of the image color feature matrix is ​​as follows: The multimodal printed fabric image is decomposed into three layers by dual-tree complex wavelet transform to obtain low-frequency components and high-frequency sub-bands in six directions; 5×5 neighborhood energy is calculated based on each high-frequency sub-band, and attention weights of each sub-band are dynamically allocated based on a softmax function; The gradient amplitude is calculated based on the low-frequency component; and each high-frequency sub-band is corrected according to the gradient amplitude and the sub-band attention weight; finally, the image color feature matrix is ​​obtained by inverse transformation and reconstruction based on the corrected high-frequency sub-band and low-frequency component; S3. Analyze the image color feature matrix to obtain multivariate image color feature coefficients; The multivariate image color feature coefficients include color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient and color saturation balance coefficient; S4. Obtaining a first printed fabric quality coefficient according to the multivariate image color characteristic coefficient, and adjusting printing process parameters.

2. The method for analyzing color characteristics of silk-cotton knitted short-process printed fabrics according to claim 1, characterized in that: The acquisition process of the multimodal printed fabric image includes: using a pre-trained ResNet-18 to extract features from a first multispectral image to obtain a first multispectral feature map; using a pre-trained VGG-16 to extract features from a second high-resolution RGB image to obtain a second high-resolution RGB feature map, and based on the attention weight, performing cross-modal alignment, attention weighted fusion and upsampling on the first multispectral feature map and the second high-resolution RGB feature map to finally generate a multimodal printed fabric image.

3. The method for analyzing color characteristics of silk-cotton knitted short-process printed fabrics according to claim 2, characterized in that: The multimodal printed fabric image is: ； in, It is a multimodal printed fabric image; is the attention weight of the multispectral image; is the first multi-spectral characteristic graph; It is the second high-resolution RGB feature map.

4. The method for analyzing color characteristics of silk-cotton knitted short-process printed fabrics according to claim 3, characterized in that: The color frequency domain complexity coefficient is obtained by performing Fourier transform on the LAB color space of the image color feature matrix to obtain an amplitude spectrum, and performing ring band energy statistics based on the amplitude spectrum to obtain the color frequency domain complexity coefficient; the color uniformity coefficient is obtained by dividing the image color feature matrix into multiple windows, performing multi-scale covariance analysis of the LAB color space on each window, and performing maximum eigenvalue mapping analysis; The color texture coordination coefficient is obtained by extracting the gray level co-occurrence matrix from the L channel of the image color feature matrix, and obtaining the contrast and correlation analysis; The color stability coefficient is obtained by obtaining the color difference field and the edge weight matrix based on the image color feature matrix, and is analyzed according to the color difference field and the edge weight matrix; the color saturation balance coefficient is obtained by calculating the saturation, partitioning according to the saturation, and then analyzing according to the pixel proportion of each area.

5. The method for analyzing color characteristics of silk-cotton knitted short-process printed fabrics according to claim 4, characterized in that: The first printed fabric quality coefficient is: ； in, Indicates the quality coefficient of the first printed fabric; Indicates Multivariate image color feature coefficients; Represents the exponential adjustment coefficient.

6. A color feature analysis system for silk-cotton knitted short-process printed fabrics, characterized in that: include: The multimodal image acquisition and fusion module is used to obtain the first multispectral image and the second high-resolution RGB image of the silk-cotton knitted short-process printed fabric, and perform cross-modal fusion to obtain a multimodal printed fabric image; The denoising and image color feature extraction module is used to perform a dual-tree complex wavelet transform on the multimodal printed fabric image, and combine the attention mechanism with the gradient-aware filtering to separate the low-frequency color base and the high-frequency knitting texture noise, so as to obtain the image color feature matrix of the printed fabric; the specific acquisition process of the image color feature matrix is ​​as follows: The multimodal printed fabric image is decomposed into three layers by dual-tree complex wavelet transform to obtain low-frequency components and high-frequency sub-bands in six directions; 5×5 neighborhood energy is calculated based on each high-frequency sub-band, and attention weights of each sub-band are dynamically allocated based on a softmax function; The gradient amplitude is calculated based on the low-frequency component; and each high-frequency sub-band is corrected according to the gradient amplitude and the sub-band attention weight; finally, the image color feature matrix is ​​obtained by inverse transformation and reconstruction based on the corrected high-frequency sub-band and low-frequency component; An image color feature analysis module, used to analyze the image color feature matrix to obtain multivariate image color feature coefficients; The multivariate image color feature coefficients include color frequency domain complexity coefficient, color uniformity coefficient, color texture coordination coefficient, color stability coefficient and color saturation balance coefficient; The quality analysis and parameter adjustment module is used to obtain the first printing fabric quality coefficient according to the multivariate image color characteristic coefficient and adjust the printing process parameters.

Images