Food mixing uniformity detection method and system based on artificial intelligence

Abstract

The invention discloses a food mixing uniformity detection method and system based on artificial intelligence, and relates to the technical field of food processing.The method comprises the following steps that an image of the surface of a food mixture is obtained and preprocessed; extracting mixture features from the preprocessed image, analyzing the mixture features and evaluating the uniformity of the surface of the food mixture; acquiring the conductivity distribution of the food mixture by using a food mixture inversion algorithm; based on the obtained conductivity distribution of the food mixture, evaluating the internal uniformity of the food mixture; and comprehensively evaluating the uniformity of the food mixture by combining the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture. By acquiring the surface and conductivity distribution of the food mixture and analyzing the uniformity of the food mixture, the uniformity of the food mixture can be comprehensively detected, the quality control level in the food processing process can be improved, and the production efficiency and the product quality stability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of food processing, and in particular to a method and system for detecting food mixing uniformity based on artificial intelligence. Background Art

[0002] Food is an indispensable part of our daily life. It has a wide variety of nutritional values ​​and is our main source of energy and nutrition. From processed food to semi-finished products to unprocessed food, they exist in various forms, providing people with a rich variety of choices. With the rapid development of science and technology, people's requirements for food innovation are also constantly increasing. Mixing different ingredients, such as mixing solids with liquids, can not only make them blend together to create more kinds of food and meet people's diverse taste needs, but also inject new vitality into the sustainable development and progress of the food industry.

[0003] Food mixing uniformity testing is a necessary step to ensure that all ingredients in food are evenly distributed during the production process. This test can effectively avoid the situation where there are too many or too few local ingredients in the food, and ensure that the food is balanced and consistent in taste, color, aroma, etc.; in addition, uniform mixing also helps to ensure the balanced distribution of nutrients in the food, so that consumers can obtain comprehensive and balanced nutrition and meet their expectations for the nutritional value of food.

[0004] In the existing food mixing uniformity detection process, the commonly used method is to evaluate the uniformity based on the observation of the surface color of the food mixture. Although this method is easy to operate and intuitive, it has certain limitations. It can only reflect the color distribution on the surface of the food mixture, but cannot go deep into its interior to explore the true component distribution and uniformity, and thus cannot accurately reflect the mixing condition of the mixture, resulting in reduced detection efficiency.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes a food mixing uniformity detection method and system based on artificial intelligence to overcome the above-mentioned technical problems existing in the existing related technology.

[0007] To this end, the specific technical solution adopted by the present invention is as follows: According to one aspect of the present invention, there is provided a method for detecting food mixing uniformity based on artificial intelligence, the method comprising the following steps: S1, obtaining an image of the surface of the food mixture and performing preprocessing; S2, extracting mixture features from the preprocessed images, analyzing the mixture features and evaluating the uniformity of the food mixture surface; S3, using a food mixture inversion algorithm to obtain the conductivity distribution of the food mixture; S4. evaluating the homogeneity inside the food mixture based on the obtained conductivity distribution of the food mixture; S5. Combine the evaluation results of the surface uniformity of the food mixture with the internal uniformity of the food mixture to comprehensively evaluate the uniformity of the food mixture.

[0008] Further, extracting mixture features from the preprocessed image, analyzing the mixture features and evaluating the uniformity of the food mixture surface includes the following steps: S21, performing frequency domain analysis on the preprocessed image using Fourier transform, and extracting mixture features of the food mixture therefrom, wherein the mixture features include color features, texture features, and particle features; S22, analyzing color features, texture features, and particle features, and using a food uniformity algorithm to obtain a uniformity classification of the surface of the food mixture; S23. Based on the uniformity classification of the food mixture surface and in combination with a preset uniformity classification standard, the uniformity of the food mixture surface is evaluated.

[0009] Further, analyzing the color features, texture features and particle features, and using the food uniformity algorithm to obtain the uniformity classification of the food mixture surface includes the following steps: S221, quantifying the color features, texture features and particle features respectively according to a preset threshold value to obtain a binary sequence; S222, performing a convolution operation on the input binary sequence according to a preset food uniformity model to extract color features, texture features, and particle features of the local mixture; S223, using a kernel principal component analysis algorithm, mapping the local mixture color features, texture features, and particle features to a linear subspace in a high-dimensional space to obtain a color feature matrix, a texture feature matrix, and a particle feature matrix; S224, fusing the color feature matrix, the texture feature matrix, and the particle feature matrix, and combining them with a weighted algorithm to obtain an eigenvalue matrix; S225, dividing the eigenvalue matrix into a number of subsets, and performing histogram statistics on each subset; S226, combining the histograms of several subsets into a feature vector, and inputting the feature vector into a support vector machine classifier to obtain a uniformity classification of the surface of the food mixture.

[0010] Further, using the food mixture inversion algorithm, obtaining the conductivity distribution of the food mixture includes the following steps: S31, applying voltage in sequence using array electrodes pre-installed on the wall of the food mixing container to establish a uniform electric field; S32, obtaining voltage data of the uniform electric field, and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture.

[0011] Furthermore, obtaining voltage data of the uniform electric field and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture includes the following steps: S321, acquiring voltage data of the array electrodes based on the uniform electric field; S322. Establishing a mathematical model of voltage data and electrical conductivity according to the electrical properties of the food mixture; S323. Based on the mathematical model, the conductivity distribution inside the mixture is obtained by using the food mixing inversion algorithm.

[0012] Furthermore, based on the mathematical model and using the food mixing inversion algorithm, obtaining the conductivity distribution inside the mixture includes the following steps: S3231, using the food mixture as a sound source to obtain voltage data generated by the food mixture; S3232, based on the voltage data generated by the food mixture obtained, combined with the mathematical model, using the food mixture inversion algorithm to obtain the initial sound source distribution; S3232, dividing the sound source distribution according to the initial sound source distribution obtained by the food mixture inversion algorithm and combining the acoustic characteristic prior information of the food mixture; S3233, according to the acoustic parameters of each divided area, using the sound source distribution calculation formula, to obtain the sound source distribution; S3234. Based on the sound source distribution, the conductivity distribution is obtained using the eddy current vector source algorithm.

[0013] Furthermore, the sound source distribution calculation formula is: In the formula, ▽ represents the gradient operator; v Represented as voltage data of array electrodes; r It is expressed as the radius of the food mixing container; f(r,k) Sound waves expressed as food mixtures; θ(r) The sound source represented as a food mixture.

[0014] Further, based on the obtained conductivity distribution of the food mixture, evaluating the homogeneity inside the food mixture includes the following steps: S41, based on the obtained conductivity distribution, using wavelet transform to convert the conductivity distribution into a conductivity distribution image; S42, extracting the mean value of conductivity from the conductivity image, and calculating the conductivity variation coefficient; S43, analyzing the conductivity variation coefficient and identifying the conductivity change area; S44. Evaluate the internal uniformity of the food mixture based on the conductivity variation coefficient and the conductivity variation area.

[0015] Further, combining the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture, the comprehensive evaluation of the uniformity of the food mixture includes the following steps: S51. Standardizing the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture; S52. Obtain comprehensive evaluation criteria and comprehensive weights based on the type, characteristics and evaluation requirements of the food mixture; S52. Based on the obtained comprehensive weight, the evaluation results of the surface uniformity and the internal uniformity of the food mixture are weighted averaged to obtain a numerical result of the comprehensive evaluation; S53. Analyze and comprehensively evaluate the numerical results, identify the difference between the surface and internal uniformity of the food mixture, and visualize the difference between the surface and internal uniformity of the food mixture through graphs; S54. Comprehensively evaluate the uniformity of food mixtures based on visual display results and comprehensive evaluation criteria.

[0016] According to another aspect of the present invention, there is also provided an artificial intelligence-based food mixing uniformity detection system, the artificial intelligence-based food mixing uniformity detection system comprising a surface image acquisition module, a surface uniformity evaluation module, a conductivity distribution acquisition module, an internal uniformity evaluation module and a comprehensive uniformity evaluation module; The surface image acquisition module, the surface uniformity evaluation module, the conductivity distribution acquisition module, the internal uniformity evaluation module and the comprehensive uniformity evaluation module are connected in sequence; A surface image acquisition module, used for acquiring an image of the surface of the food mixture and performing preprocessing; A surface uniformity assessment module, which is used to extract mixture features from the preprocessed images, analyze the mixture features and assess the uniformity of the food mixture surface; A conductivity distribution acquisition module, used for acquiring the conductivity distribution of the food mixture by using a food mixture inversion algorithm; An internal uniformity evaluation module, used for evaluating the internal uniformity of the food mixture based on the obtained conductivity distribution of the food mixture; The comprehensive uniformity evaluation module is used to combine the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture to comprehensively evaluate the uniformity of the food mixture.

[0017] The beneficial effects of the present invention are: 1. The present invention can comprehensively detect the uniformity of the food mixture by obtaining the surface and conductivity distribution of the food mixture and analyzing its uniformity, which not only helps to improve the quality control level in the food processing process, but also can achieve more accurate and faster quality assessment through intelligent detection means, thereby further improving production efficiency and product quality stability.

[0018] 2. The present invention utilizes an automated food mixture inversion algorithm to more quickly and accurately obtain the conductivity distribution of the food mixture, which not only greatly improves the speed of uniform detection, but also enhances the clarity and accuracy of the detection, so that operators on the production line can more intuitively understand the internal structure and uniformity of the food mixture, which helps to improve the operating efficiency of the production line.

[0019] 3. The present invention can reduce manual intervention in the production process through artificial intelligence detection processing and analysis technology, thereby reducing the possibility of human errors. It not only improves the stability and reliability of the production process, but also ensures consistent quality levels during long-term operation. At the same time, it optimizes production efficiency and saves valuable human and cost resources for the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 is a flow chart of a method for detecting food mixing uniformity based on artificial intelligence according to an embodiment of the present invention; Figure 2 The figure is a principle block diagram of a food mixing uniformity detection system based on artificial intelligence according to an embodiment of the present invention.

[0022] In the figure: 1. Surface image acquisition module; 2. Surface uniformity evaluation module; 3. Conductivity distribution acquisition module; 4. Internal uniformity evaluation module; 5. Comprehensive uniformity evaluation module. DETAILED DESCRIPTION

[0023] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0024] According to an embodiment of the present invention, a method and system for detecting food mixing uniformity based on artificial intelligence are provided.

[0025] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the food mixing uniformity detection method based on artificial intelligence according to an embodiment of the present invention, the method comprises the following steps: S1. Acquire an image of the surface of the food mixture and perform preprocessing.

[0026] It should be noted that, using the set image acquisition equipment, multiple images of different angles and areas are taken to obtain clear images; the images of the surface of the food mixture are processed by Gaussian filtering to eliminate or reduce the impact of noise on the image quality; the images after Gaussian filtering are contrast enhanced and the brightness distribution of the images is adjusted to make different parts of the image clearer; the contrast enhanced images are geometrically transformed to ensure that the images can be correctly aligned and compared in subsequent analysis.

[0027] S2. Extract mixture features from the preprocessed images, analyze the mixture features and evaluate the uniformity of the food mixture surface.

[0028] Specifically, extracting mixture features from the preprocessed image, analyzing the mixture features and evaluating the uniformity of the food mixture surface includes the following steps: S21. Perform frequency domain analysis on the preprocessed image using Fourier transform, and extract mixture features of the food mixture therefrom, wherein the mixture features include color features, texture features, and particle features.

[0029] It should be noted that the frequency domain analysis of the preprocessed image using Fourier transform and the extraction of the mixture characteristics of the food mixture therefrom include the following steps: The preprocessed image is subjected to Fourier transform to convert the image from the spatial domain to the frequency domain. In the frequency domain, the color features in the image are extracted by combining the frequency components and energy distribution of the color space. The texture features of the image are extracted in the frequency domain by analyzing the intensity and distribution of different frequency components in the image. The energy within a preset frequency range is used as a representation of the mixture characteristics to obtain the characteristics of the particulate matter in the food mixture.

[0030] In addition, the Fourier transform can use the short-range Fourier transform to convert the image from the time domain to the frequency domain; for images, the short-range Fourier transform can be used to calculate the frequency domain information of each pixel on the image. The short-range Fourier transform divides the signal into small time periods and applies the Fourier transform in each time period to obtain the frequency information.

[0031] For each image pixel, the short-range Fourier transform is used to obtain its frequency domain representation, where the phase information is a part of the short-range Fourier transform output. This phase information contains the phase change in the frequency domain and reflects the structural characteristics of the image in the frequency domain. Based on the frequency domain phase information, the local phase quantization algorithm is used to capture the color characteristics, texture characteristics and particle characteristics of the image.

[0032] S22. Analyze color features, texture features, and particle features, and use a food uniformity algorithm to obtain a uniformity classification of the surface of the food mixture.

[0033] Specifically, analyzing color features, texture features, and particle features, and using a food uniformity algorithm to obtain a uniformity classification of the surface of a food mixture includes the following steps: S221. According to a preset threshold, the color feature, the texture feature and the particle feature are quantified respectively to obtain a binary sequence.

[0034] It should be noted that, according to the preset thresholds, quantifying the color features, texture features and particle features respectively includes the following steps: Based on color features, according to the preset threshold and combined with the RGB space, the colors are divided into different intervals; for each pixel, the corresponding binary code is generated according to the interval in which its color value is located. If the color value of the pixel is within a certain interval, it is encoded as 1, otherwise it is encoded as 0.

[0035] Based on the texture features, one or more thresholds are set; for each pixel or pixel block, a corresponding binary code is generated according to the comparison result between its texture feature value and the preset threshold.

[0036] Based on the characteristics of the particles, one or more thresholds are set; for each detected particle, a corresponding binary code is generated according to the comparison result between its characteristic value and the threshold.

[0037] S222. According to a preset food uniformity model, a convolution operation is performed on the input binary sequence to extract color features, texture features, and particle features of the local mixture.

[0038] It should be noted that the food uniformity model is a principal component analysis network model. According to the preset food uniformity model, a convolution operation is performed on the input binary sequence to extract the color features, texture features and particle features of the local mixture. The color features, texture features and particle features of the local mixture are extracted, wherein: A convolution kernel is selected according to the task requirements and the color characteristics, texture characteristics and particle characteristics of the mixture; the convolution kernel is applied to the binary sequence, the convolution kernel is element-wise multiplied with each value or local area of ​​the binary sequence, and the results are added to obtain a new feature representation; the convolution kernel is slid on the binary sequence to extract the feature information of the local area. These local features can reflect the changes in the color characteristics, texture characteristics and particle characteristics of the mixture at different positions and scales in the binary sequence.

[0039] S223. Using a kernel principal component analysis algorithm, the local mixture color features, texture features, and particle features are mapped to a linear subspace in a high-dimensional space to obtain a color feature matrix, a texture feature matrix, and a particle feature matrix.

[0040] It should be noted that the kernel principal component analysis algorithm is used to map the local mixture color features, texture features and particle features to a linear subspace in a high-dimensional space, and obtaining a color feature matrix, a texture feature matrix and a particle feature matrix includes the following steps: A nonlinear kernel function is selected to calculate the kernel matrix of the local mixture color features, texture features and particle features in high-dimensional space; for each mixture color feature, texture feature and particle feature, the corresponding kernel matrix is ​​centered, that is, the mean of each row in the kernel matrix is ​​subtracted, and then the mean of each column is subtracted to eliminate the translation effect of the data; the eigenvalue decomposition of the centralized kernel matrix is ​​performed to obtain the eigenvalues ​​and the corresponding eigenvectors; according to the size of the eigenvalues, the eigenvectors corresponding to the largest eigenvalues ​​are selected for each feature type (color, texture, particle matter) as the principal components; the selected principal components (i.e., eigenvectors) are used to represent the color features, texture features and particle features, respectively, and the color feature matrix, texture feature matrix and particle feature matrix are constructed.

[0041] S224, fusing the color feature matrix, the texture feature matrix and the particle feature matrix, and combining them with a weighted algorithm to obtain an eigenvalue matrix.

[0042] It should be noted that each feature matrix is ​​standardized to eliminate the scale differences between different features; weights are assigned to each feature matrix according to the importance of color features, texture features, and particle features; the standardized color feature matrix, texture feature matrix, and particle feature matrix are weighted fused according to the assigned weights to obtain the eigenvalue matrix; the fused eigenvalue matrix can be calculated by the following formula: In the formula, M F Represented as an eigenvalue matrix; M C Represented as a color feature matrix; M t Represented as a texture feature matrix; M P It is expressed as a particle feature matrix; w c Represented as the weight of the color feature matrix; w t Represented as the weight of the texture feature matrix; w p Expressed as the weight of the particle feature matrix.

[0043] S225, dividing the eigenvalue matrix into a number of subsets, and performing histogram statistics on each subset; S226, combining the histograms of several subsets into a feature vector, and inputting the feature vector into a support vector machine classifier to obtain a uniformity classification of the surface of the food mixture.

[0044] S23. Based on the uniformity classification of the food mixture surface and in combination with a preset uniformity classification standard, the uniformity of the food mixture surface is evaluated.

[0045] It should be noted that the uniformity classification standard is usually formulated based on the surface color, texture, particle distribution and other characteristics of the food mixture, and is used to evaluate the uniformity of the mixture surface; the food mixture surface uniformity classification standard is: High uniformity includes consistent color without obvious color difference or color blocks; smooth texture without obvious particles or bumps; and even distribution of particles without obvious aggregation or vacancies.

[0046] Moderate uniformity includes basically consistent color, but there may be small color differences or color blocks; the texture is basically smooth, but there may be a small amount of particles or slight bumps; the particle distribution is relatively uniform, but there may be some small clusters or gaps.

[0047] Low uniformity includes uneven color, obvious color difference or color blocks; rough texture, obvious particles or bumps; uneven distribution of particles, obvious aggregation or large areas of vacancies.

[0048] S3. Using a food mixture inversion algorithm, obtain the conductivity distribution of the food mixture.

[0049] Specifically, using the food mixture inversion algorithm to obtain the conductivity distribution of the food mixture includes the following steps: S31. Voltage is applied sequentially using array electrodes pre-installed on the wall of the food mixing container to establish a uniform electric field.

[0050] It should be noted that a suitable voltage value is set according to the properties of the food mixture and the requirements of the electric field treatment; through the control system, each electrode in the array electrode is activated in turn so that voltage is applied to them respectively; the electric field distribution in the container is monitored by an electric field measuring device to establish a uniform electric field.

[0051] S32, obtaining voltage data of the uniform electric field, and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture.

[0052] Specifically, obtaining voltage data of a uniform electric field and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture includes the following steps: S321. Based on the uniform electric field, obtain voltage data of the array electrodes.

[0053] S322. According to the electrical properties of the food mixture, a mathematical model of voltage data and conductivity is established.

[0054] It should be noted that, based on the electrical properties of the food mixture, a linear model was selected as the mathematical model to describe the relationship between the voltage data and the conductivity; the slope and intercept were estimated, the voltage data was used to fit the linear model, and the goodness of fit of the linear model was calculated.

[0055] S323. Based on the mathematical model, the conductivity distribution inside the mixture is obtained by using the food mixing inversion algorithm.

[0056] Specifically, based on the mathematical model and using the food mixing inversion algorithm, obtaining the conductivity distribution inside the mixture includes the following steps: S3231. Use the food mixture as a sound source to obtain voltage data generated by the food mixture.

[0057] S3232. Based on the voltage data generated by the food mixture, combined with the mathematical model, the food mixture inversion algorithm is used to obtain the initial sound source distribution.

[0058] S3232. Divide the sound source distribution according to the initial sound source distribution obtained by the food mixture inversion algorithm and the prior information of the acoustic characteristics of the food mixture.

[0059] S3233. According to the acoustic parameters of each divided area, the sound source distribution is obtained by using the sound source distribution calculation formula.

[0060] Specifically, the sound source distribution calculation formula is: In the formula, ▽ represents the gradient operator; v Represented as voltage data of array electrodes; r It is expressed as the radius of the food mixing container; f(r,k) Sound waves expressed as food mixtures; θ(r) The sound source represented as a food mixture.

[0061] S3234. Based on the sound source distribution, the conductivity distribution is obtained using the eddy current vector source algorithm.

[0062] It should be noted that, based on the obtained sound source distribution image, the eddy current vector source algorithm is used to convert the sound source distribution image into the corresponding eddy current field distribution; based on the voltage theory, the physical relationship between eddy current and conductivity, a mathematical model between the eddy current field and the conductivity distribution is established; based on the calculation results of the eddy current vector source algorithm and the established mathematical model, the distribution of conductivity is calculated.

[0063] S4. Based on the obtained conductivity distribution of the food mixture, evaluate the homogeneity inside the food mixture.

[0064] Specifically, based on the obtained conductivity distribution of the food mixture, evaluating the homogeneity inside the food mixture includes the following steps: S41. Based on the acquired conductivity distribution, the conductivity distribution is converted into a conductivity distribution image using wavelet transform.

[0065] S42. Extract the mean value of conductivity from the conductivity image, and calculate the conductivity variation coefficient.

[0066] It should be noted that the calculation of the conductivity variation coefficient includes: calculating the square of the difference between each conductivity data point and the mean; adding up these square values ​​and dividing them by the number of data points minus 1 (i.e., sample variance); taking the square root of the obtained sample variance to obtain the standard deviation; dividing the standard deviation by the mean and multiplying by 100% to obtain the conductivity variation coefficient.

[0067] S43. Analyze the conductivity variation coefficient and identify the conductivity change area.

[0068] It should be noted that one or more thresholds are set according to the uniformity requirements of the food mixture. These thresholds can be based on experience, experimental data or industry standards. The conductivity coefficient of variation of each data point is compared with the set threshold. If the conductivity coefficient of variation of a data point exceeds the threshold range, the area where the point is located is considered to be a conductivity variation area.

[0069] S44. Evaluate the internal uniformity of the food mixture based on the conductivity variation coefficient and the conductivity variation area.

[0070] It should be noted that the uniformity within the food mixture can be evaluated based on the analysis results of the conductivity variation coefficient and the conductivity change area; if the coefficient of variation is small and the conductivity data distribution is relatively concentrated without obvious outliers, it means that the uniformity within the food mixture is good; otherwise, it means that the uniformity is poor.

[0071] S5. Combine the evaluation results of the surface uniformity of the food mixture with the internal uniformity of the food mixture to comprehensively evaluate the uniformity of the food mixture.

[0072] Specifically, combining the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture, the comprehensive evaluation of the uniformity of the food mixture includes the following steps: S51. Standardize the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture.

[0073] S52. Obtain comprehensive evaluation criteria and comprehensive weights based on the type, characteristics and evaluation requirements of the food mixture.

[0074] It should be noted that the comprehensive evaluation criteria include the numerical range of uniformity, its importance, and its impact on food quality, including: The numerical range of uniformity. Uniformity refers to the distribution consistency of food in terms of color, taste, nutritional content, etc. In order to quantitatively evaluate uniformity, a numerical range needs to be set. This range can be set according to the specific type of food, production process and consumer expectations. For example, when evaluating the uniformity of jam, the color, texture and particle size content of different parts can be measured, and a reasonable fluctuation range can be set. If the measured value is within this range, the jam is considered to have good uniformity.

[0075] The importance of uniformity. Uniformity has a crucial impact on food quality. First, uniform food has a more consistent taste, which can bring consumers a better eating experience. Second, uniform food has a more stable nutritional value, which can ensure that consumers take in a balanced amount of nutrients. In addition, uniform food is easier to control during production and processing, which can improve production efficiency and product quality.

[0076] The impact of uniformity on food quality, including the taste impact. The unevenness of food may lead to differences in taste, such as a certain part being too sweet, too sour, or containing too much pulp, which will affect the consumer's eating experience; the nutritional value impact. If the nutrients in the food are unevenly distributed, the nutritional value of some parts may be too high or too low, thus failing to meet the nutritional needs of consumers; the production control impact. Uneven food is difficult to control during the production process, which may lead to increased production costs and reduced product quality.

[0077] S52. Based on the obtained comprehensive weight, the evaluation results of the surface uniformity and internal uniformity of the food mixture are weighted averaged to obtain a numerical result of the comprehensive evaluation.

[0078] S53. Analyze and comprehensively evaluate the numerical results, identify the difference between the surface and internal uniformity of the food mixture, and visualize the difference between the surface and internal uniformity of the food mixture through graphs.

[0079] It should be noted that the absolute difference is used to calculate the difference between the surface and internal uniformity of the food mixture; the causes of the difference are analyzed in combination with the specific eigenvalue type and the characteristics of the food mixture; and the comparison results are presented in an intuitive form using visualization tools (such as scatter plots, box plots, heat maps, etc.); The calculation formula of absolute difference is: Z=|AB|; In the formula, Z Expressed as absolute error; A Expressed as measured values; B Represents the true value.

[0080] S54. Comprehensively evaluate the uniformity of food mixtures based on visual display results and comprehensive evaluation criteria.

[0081] like Figure 2 As shown, according to another embodiment of the present invention, there is also provided a food mixing uniformity detection system based on artificial intelligence, the food mixing uniformity detection system based on artificial intelligence comprising a surface image acquisition module 1, a surface uniformity evaluation module 2, a conductivity distribution acquisition module 3, an internal uniformity evaluation module 4 and a comprehensive uniformity evaluation module 5; The surface image acquisition module 1, the surface uniformity assessment module 2, the conductivity distribution acquisition module 3, the internal uniformity assessment module 4 and the comprehensive uniformity assessment module 5 are connected in sequence; A surface image acquisition module 1 is used to acquire an image of the surface of the food mixture and perform preprocessing; A surface uniformity assessment module 2, for extracting mixture features from the preprocessed image, analyzing the mixture features and assessing the uniformity of the food mixture surface; The conductivity distribution acquisition module 3 is used to obtain the conductivity distribution of the food mixture by using the food mixture inversion algorithm; An internal uniformity evaluation module 4, used for evaluating the internal uniformity of the food mixture based on the obtained conductivity distribution of the food mixture; The comprehensive uniformity evaluation module 5 is used to comprehensively evaluate the uniformity of the food mixture by combining the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture.

[0082] In summary, by means of the above technical scheme of the present invention, by obtaining the surface and conductivity distribution of the food mixture and analyzing its uniformity, the uniformity of the food mixture can be comprehensively detected, which not only helps to improve the quality control level in the food processing process, but also can achieve more accurate and faster quality assessment through intelligent detection means, thereby further improving production efficiency and product quality stability. By utilizing the automated food mixing inversion algorithm, the conductivity distribution of the food mixture can be obtained more quickly and accurately, which not only greatly improves the speed of uniform detection, but also enhances the clarity and accuracy of the detection, so that the operators on the production line can more intuitively understand the internal structure and uniformity of the food mixture, which helps to improve the operation efficiency of the production line. Through the artificial intelligence detection processing and analysis technology, the manual intervention in the production process can be reduced, thereby reducing the possibility of human error, not only improving the stability and reliability of the production process, but also ensuring that the quality level is maintained consistently during long-term operation, while optimizing production efficiency, saving valuable human and cost resources for the enterprise.

[0083] 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, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting food mixing uniformity based on artificial intelligence, characterized in that: The method comprises the following steps: S1, obtaining an image of the surface of the food mixture and performing preprocessing; S2, extracting mixture features from the preprocessed images, analyzing the mixture features and evaluating the uniformity of the food mixture surface; S3, using a food mixture inversion algorithm to obtain the conductivity distribution of the food mixture; S4. evaluating the homogeneity inside the food mixture based on the obtained conductivity distribution of the food mixture; S5. Combine the evaluation results of the surface uniformity of the food mixture with the internal uniformity of the food mixture to comprehensively evaluate the uniformity of the food mixture.

2. The method for detecting food mixing uniformity based on artificial intelligence according to claim 1, characterized in that: The method of extracting mixture features from the preprocessed image, analyzing the mixture features and evaluating the uniformity of the surface of the food mixture comprises the following steps: S21, performing frequency domain analysis on the preprocessed image using Fourier transform, and extracting mixture features of the food mixture therefrom, wherein the mixture features include color features, texture features, and particle features; S22, analyzing color features, texture features, and particle features, and using a food uniformity algorithm to obtain a uniformity classification of the surface of the food mixture; S23. Based on the uniformity classification of the food mixture surface and in combination with a preset uniformity classification standard, the uniformity of the food mixture surface is evaluated.

3. The method for detecting food mixing uniformity based on artificial intelligence according to claim 2, characterized in that: The method of analyzing the color features, texture features and particle features and using the food uniformity algorithm to obtain the uniformity classification of the surface of the food mixture comprises the following steps: S221, quantifying the color features, texture features and particle features respectively according to a preset threshold value to obtain a binary sequence; S222, performing a convolution operation on the input binary sequence according to a preset food uniformity model to extract color features, texture features, and particle features of the local mixture; S223, using a kernel principal component analysis algorithm, mapping the local mixture color features, texture features, and particle features to a linear subspace in a high-dimensional space to obtain a color feature matrix, a texture feature matrix, and a particle feature matrix; S224, fusing the color feature matrix, the texture feature matrix, and the particle feature matrix, and combining them with a weighted algorithm to obtain an eigenvalue matrix; S225, dividing the eigenvalue matrix into a number of subsets, and performing histogram statistics on each subset; S226, combining the histograms of several subsets into a feature vector, and inputting the feature vector into a support vector machine classifier to obtain a uniformity classification of the surface of the food mixture.

4. The method for detecting food mixing uniformity based on artificial intelligence according to claim 1, characterized in that: The method of obtaining the conductivity distribution of the food mixture by using the food mixture inversion algorithm comprises the following steps: S31, applying voltage in sequence using array electrodes pre-installed on the wall of the food mixing container to establish a uniform electric field; S32, obtaining voltage data of the uniform electric field, and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture.

5. The method for detecting food mixing uniformity based on artificial intelligence according to claim 4, characterized in that: The step of obtaining voltage data of a uniform electric field and processing the voltage data using a food mixture inversion algorithm to obtain the conductivity distribution inside the food mixture comprises the following steps: S321, acquiring voltage data of the array electrodes based on the uniform electric field; S322. Establishing a mathematical model of voltage data and electrical conductivity according to the electrical properties of the food mixture; S323. Based on the mathematical model, the conductivity distribution inside the mixture is obtained by using the food mixing inversion algorithm.

6. The method for detecting food mixing uniformity based on artificial intelligence according to claim 5, characterized in that: The method of obtaining the conductivity distribution inside the mixture based on the mathematical model and using the food mixing inversion algorithm comprises the following steps: S3231, using the food mixture as a sound source to obtain voltage data generated by the food mixture; S3232, based on the voltage data generated by the food mixture obtained, combined with the mathematical model, using the food mixture inversion algorithm to obtain the initial sound source distribution; S3232, dividing the sound source distribution according to the initial sound source distribution obtained by the food mixture inversion algorithm and combining the acoustic characteristic prior information of the food mixture; S3233, according to the acoustic parameters of each divided area, using the sound source distribution calculation formula, to obtain the sound source distribution; S3234. Based on the sound source distribution, the conductivity distribution is obtained using the eddy current vector source algorithm.

7. The method for detecting food mixing uniformity based on artificial intelligence according to claim 6, characterized in that: The sound source distribution calculation formula is: In the formula, ▽ represents the gradient operator; v Represented as voltage data of array electrodes; r It is expressed as the radius of the food mixing container; f(r,k) Sound waves expressed as food mixtures; θ(r) The sound source represented as a food mixture.

8. The method for detecting food mixing uniformity based on artificial intelligence according to claim 1, characterized in that: The method of evaluating the homogeneity of the food mixture based on the obtained conductivity distribution of the food mixture comprises the following steps: S41, based on the obtained conductivity distribution, using wavelet transform to convert the conductivity distribution into a conductivity distribution image; S42, extracting the mean value of conductivity from the conductivity image, and calculating the conductivity variation coefficient; S43, analyzing the conductivity variation coefficient and identifying the conductivity change area; S44. Evaluate the internal uniformity of the food mixture based on the conductivity variation coefficient and the conductivity variation area.

9. The method for detecting food mixing uniformity based on artificial intelligence according to claim 1, characterized in that: The method of comprehensively evaluating the uniformity of the food mixture by combining the evaluation results of the uniformity of the surface of the food mixture and the uniformity of the interior of the food mixture comprises the following steps: S51. Standardizing the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture; S52. Obtain comprehensive evaluation criteria and comprehensive weights based on the type, characteristics and evaluation requirements of the food mixture; S52. Based on the obtained comprehensive weight, the evaluation results of the surface uniformity and the internal uniformity of the food mixture are weighted averaged to obtain a numerical result of the comprehensive evaluation; S53. Analyze and comprehensively evaluate the numerical results, identify the difference between the surface and internal uniformity of the food mixture, and visualize the difference between the surface and internal uniformity of the food mixture through graphs; S54. Comprehensively evaluate the uniformity of food mixtures based on visual display results and comprehensive evaluation criteria.

10. A food mixing uniformity detection system based on artificial intelligence, used to implement the food mixing uniformity detection method based on artificial intelligence as claimed in any one of claims 1 to 9, characterized in that: The food mixing uniformity detection system based on artificial intelligence includes a surface image acquisition module, a surface uniformity evaluation module, a conductivity distribution acquisition module, an internal uniformity evaluation module and a comprehensive uniformity evaluation module; The surface image acquisition module, the surface uniformity assessment module, the conductivity distribution acquisition module, the internal uniformity assessment module and the comprehensive uniformity assessment module are connected in sequence; The surface image acquisition module is used to acquire an image of the surface of the food mixture and perform preprocessing; The surface uniformity assessment module is used to extract mixture features from the preprocessed image, analyze the mixture features and assess the uniformity of the food mixture surface; The conductivity distribution acquisition module is used to acquire the conductivity distribution of the food mixture by using a food mixture inversion algorithm; The internal uniformity evaluation module is used to evaluate the internal uniformity of the food mixture based on the obtained conductivity distribution of the food mixture; The comprehensive uniformity evaluation module is used to comprehensively evaluate the uniformity of the food mixture by combining the evaluation results of the surface uniformity of the food mixture and the internal uniformity of the food mixture.

Images