Method for identifying taste quality difference of Hainan mulberry leaf green tea with different tenderness

Abstract

The invention relates to the technical field of agricultural product detection, in particular to a method for identifying the taste quality difference of Hainan mulberry leaf green tea with different tenderness, which comprises the following steps: obtaining Hainan mulberry leaf green tea samples, measuring the near infrared spectrum absorption value of each sample at 800-2500nm, recording the absorption values in a data array, and generating a spectrum absorption data array. According to the method disclosed by the invention, the chemical components and the taste quality of the Hainan mulberry leaf tea can be evaluated through differential processing of the spectral absorption data and color space conversion from RGB (Red Green Blue) to HSV (Hue Saturation Value). The quality of the tea leaves is comprehensively evaluated by statistically analyzing spectrum change indexes and color characteristic indexes, and quantitative data support is provided. And particularly, correlation coefficients are used for calculation, so that the correlation degree among different quality indexes is defined, and the quality evaluation is more scientific and systematic. According to the method, the quality control is more targeted in the production process, and producers are helped to adjust the process and optimize the product quality.

Description

Technical Field

[0001] The invention relates to the technical field of agricultural product detection, and in particular to a method for identifying taste quality differences of Hainan mulberry leaf green teas with different tenderness. Background Art

[0002] Mulberry leaf is the leaf of mulberry tree, which has been widely used since ancient times, especially in China. It is not only the main feed for silkworm breeding, but also has important medicinal value and therapeutic effect. Studies have shown that mulberry leaf has medicinal value such as lowering blood lipids, lowering blood sugar, anti-diabetes, anti-hypertension, anti-obesity and anti-atherosclerosis. It can be used for dietary therapy to achieve the effects of clearing lungs and moisturizing dryness, clearing liver and improving eyesight, and evacuating wind and heat. However, few studies have reported on the metabolism and taste quality of Hainan mulberry leaf tea, and there is a lack of systematic research on the chemical composition and taste quality of mulberry leaf tea. For food and tea products, consumer acceptance depends largely on sensory quality. The lack of systematic quality evaluation will lead to difficulties in product development and market promotion. Summary of the invention

[0003] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness.

[0004] In order to achieve the above object, the present invention adopts the following technical scheme, a method for identifying the taste quality difference of Hainan mulberry leaf green tea with different tenderness, comprising the following steps:

[0005] Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample at 800nm ​​to 2500nm, record the absorption value into a data array, generate a spectrum absorption data array, take an image of each sample, extract RGB color values ​​from the image, and generate an image color data array;

[0006] Based on the spectral absorption data array, the absorption values ​​of each sample between adjacent wavelengths are subtracted, the obtained difference is stored in a new data array, and a spectral difference data array is generated; the numerical average value in the spectral difference data array is calculated for each sample to obtain a numerical value representing the spectral change, and a spectral change index is generated; the spectral change index of each sample is compared, the spectral difference pattern is identified, and a spectral tenderness score is generated;

[0007] Based on the image color data array, converting the RGB color value of each sample into the HSV color space, calculating the hue, saturation and lightness components, generating the HSV color data array, calculating the average values ​​of the hue, saturation and lightness in the HSV color data array for each sample, obtaining a numerical value representing the color feature, generating a color feature index, analyzing the distribution of the color feature index of each sample, identifying the change of color attributes, and generating a color tenderness score;

[0008] The spectral tenderness score and color tenderness score of each sample were combined, the correlation coefficient between the scores was calculated, and the samples were sorted according to the correlation coefficient to generate a tenderness quality ranking.

[0009] Preferably, the steps of acquiring the spectral absorption data array are:

[0010] Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample in the range of 800nm ​​to 2500nm, and obtain spectrum absorption data of each sample;

[0011] Arrange the spectral absorption data of each sample into a data array and store them in order of wavelength to obtain an arranged spectral absorption data array;

[0012] The integrity and accuracy of the sorted spectral absorption data array are verified to obtain the spectral absorption data array.

[0013] Preferably, the steps of acquiring the image color data array are:

[0014] Collect images of each Hainan mulberry leaf green tea sample to obtain an image dataset of each sample;

[0015] Based on the image data set of each sample, each image is analyzed, the red, green and blue channel values ​​in the image are extracted, the numerical distribution of each color is recorded, and the RGB color data of each sample is obtained;

[0016] According to the RGB color data of each sample, data arrangement and standardization are performed, and the RGB values ​​are arranged into the same data structure according to the sample to form an image color data array.

[0017] Preferably, the steps of obtaining the spectral change index are:

[0018] Based on the spectral absorption data array, the absorption values ​​of each sample at adjacent wavelengths are subtracted, and the difference in absorption values ​​between two consecutive wavelength points is calculated each time to obtain absorption difference data;

[0019] Based on the absorption difference data, the numerical average of all differences in the spectral difference data array of each sample is calculated to obtain a spectral change index. The calculation formula of the numerical average is:

[0020] ;

[0021] in, is the numerical average of the spectral variation, is the total number of wavelength points, and are the absorbance values ​​at consecutive wavelength points.

[0022] Preferably, the steps for obtaining the spectral tenderness score are:

[0023] Collect the spectral change index of each sample to obtain preliminary difference pattern data;

[0024] Based on the preliminary difference pattern data, the spectral tenderness score of each sample was calculated using the formula:

[0025] ;

[0026] in, Representative samples Spectral tenderness score, and The samples and samples The spectral variation index, is the total number of samples, and The samples and samples The standard deviation of the spectral variation index;

[0027] Based on the calculated spectral tenderness score of each sample, the scores were collated and sorted to obtain a spectral tenderness score.

[0028] Preferably, the steps of obtaining the color characteristic index are:

[0029] Based on the image color data array, convert the RGB color value of each sample into the HSV color space to obtain the HSV color data array of each sample;

[0030] Calculate the hue, saturation and lightness values ​​of each sample from the HSV color data array, and calculate the average value of each component to form the hue average value, saturation average value and lightness average value;

[0031] The average hue, saturation and lightness of each sample are integrated to form a color characteristic index.

[0032] Preferably, the steps for obtaining the color tenderness score are:

[0033] Analyzing the hue, saturation and brightness distribution of each sample from the color characteristic index, identifying the color attribute changes between samples, and obtaining color attribute change data;

[0034] Based on the color attribute change data, the color tenderness score of each sample is calculated using the following formula:

[0035] ;

[0036] in, For samples Color tenderness score, The samples The average value of hue, saturation and lightness, is the adjustment factor, For samples The weight coefficient of is the total number of samples, For comparison sample index.

[0037] Preferably, the steps for obtaining the tenderness quality ranking are:

[0038] The spectral tenderness score and color tenderness score of each sample were summarized, and the correlation coefficient between the spectral tenderness score and the color tenderness score was calculated using the following formula:

[0039] ;

[0040] in, represents the correlation coefficient between the spectral tenderness score and the color tenderness score, and Samples Spectral tenderness score and color tenderness score, and are the averages of the spectral tenderness score and the color tenderness score, is the total number of samples;

[0041] Based on the correlation coefficients, the samples were ranked to generate a tenderness quality ranking.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are:

[0043] The present invention can evaluate the chemical composition and taste quality of Hainan mulberry leaf tea by differential processing of spectral absorption data and color space conversion from RGB to HSV. By statistically analyzing spectral change indicators and color characteristic indicators, the quality of tea is comprehensively evaluated, providing quantitative data support. In particular, by using correlation coefficient calculation, the correlation between different quality indicators is clarified, making the quality evaluation more scientific and systematic. This method makes quality control more targeted in the production process, which helps producers adjust the process and optimize product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0046] See also Figure 1 The present invention provides a technical solution, a method for identifying the taste quality difference of Hainan mulberry leaf green tea with different tenderness, comprising the following steps:

[0047] Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample at 800nm ​​to 2500nm, record the absorption value into a data array, generate a spectrum absorption data array, take an image of each sample, extract RGB color values ​​from the image, and generate an image color data array;

[0048] Based on the spectral absorption data array, the absorption values ​​of each sample at adjacent wavelengths are subtracted, the obtained difference is stored in a new data array, a spectral difference data array is generated, the numerical average value in the spectral difference data array is calculated for each sample, a numerical value representing the spectral change is obtained, a spectral change index is generated, the spectral change index of each sample is compared, the spectral difference pattern is identified, and a spectral tenderness score is generated;

[0049] Based on the image color data array, the RGB color value of each sample is converted into the HSV color space, the hue, saturation and lightness components are calculated, and the HSV color data array is generated. For each sample, the average value of the hue, saturation and lightness in the HSV color data array is calculated to obtain the value representing the color characteristics, generate the color characteristic index, analyze the distribution of the color characteristic index of each sample, identify the change of color attributes, and generate the color tenderness score;

[0050] The spectral tenderness score and color tenderness score of each sample were combined, the correlation coefficient between the scores was calculated, and the samples were sorted according to the correlation coefficient to generate a tenderness quality ranking.

[0051] The steps to obtain the spectral absorption data array are:

[0052] Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample in the range of 800nm ​​to 2500nm, and obtain spectrum absorption data of each sample;

[0053] Arrange the spectral absorption data of each sample into a data array and store them in order of wavelength to obtain an arranged spectral absorption data array;

[0054] The integrity and accuracy of the sorted spectral absorption data array are verified to obtain the spectral absorption data array.

[0055] Specifically, after obtaining the Hainan mulberry leaf green tea samples, each sample is numbered, and then the wavelength range is set to 800nm ​​to 2500nm using a spectrometer to ensure that the measurement covers the required spectral region, and the spectral absorption value of each sample is measured one by one. During the measurement process, the operator will repeat the measurement of each wavelength point at least three times to ensure the repeatability and accuracy of the data. After each measurement, the data for each wavelength point is averaged to calculate a stable absorption value, and these averages are recorded in a data table. In addition, for each measured value, the error range is checked to ensure that all data points are within the error threshold of the spectrometer.

[0056] At the beginning of the verification process, technicians check the sorted spectral absorption data array and compare the deviation of each data point with the standard data to ensure that the data is within an acceptable error range. For abnormal data points detected, they are reviewed to confirm whether they are instrument errors or sample abnormalities. After confirming the consistency and reliability of the data, the confirmed data is stored as a verified spectral absorption data array.

[0057] The steps to obtain the image color data array are:

[0058] Collect images of each Hainan mulberry leaf green tea sample to obtain an image dataset of each sample;

[0059] Based on the image data set of each sample, each image is analyzed, the red, green and blue channel values ​​in the image are extracted, the numerical distribution of each color is recorded, and the RGB color data of each sample is obtained;

[0060] According to the RGB color data of each sample, data arrangement and standardization are performed, and the RGB values ​​are arranged into the same data structure according to the sample to form an image color data array.

[0061] Specifically, images of each Hainan mulberry leaf green tea sample are collected and photographed under a uniform light source to ensure that each image evenly displays the natural color and details of the tea leaves for color analysis. These images are collected to form the image dataset for each sample.

[0062] Based on the image dataset of each sample, each image was analyzed, the red, green, and blue channels were separated, and the color value distribution of each channel was calculated separately.

[0063] According to the RGB color data of each sample, the RGB data is normalized and the erroneous data is corrected to ensure the comparability of the data of all samples. During the compilation process, each color value is organized into the same data structure according to the sample number.

[0064] The steps for obtaining the spectral change index are:

[0065] Based on the spectral absorption data array, the absorption values ​​of each sample at adjacent wavelengths are subtracted, and the difference in absorption values ​​between two consecutive wavelength points is calculated each time to obtain the absorption difference data;

[0066] Based on the absorption difference data, the numerical average of all differences in the spectral difference data array of each sample is calculated to obtain the spectral change index. The calculation formula of the numerical average is:

[0067] ;

[0068] in, is the numerical average of the spectral variation, is the total number of wavelength points, and are the absorbance values ​​at consecutive wavelength points.

[0069] Specifically, based on the spectral absorption data array, the differential calculation of the absorption value of each sample between adjacent wavelengths is performed. First, the absorption value of each sample at each wavelength point is extracted one by one from the data array, and then the absorption values ​​at two consecutive wavelength points are subtracted. Two consecutive data points are extracted each time, and their difference is calculated. The difference obtained by subtraction reflects the influence of wavelength changes on the absorption value. The absolute value of the difference is to eliminate the influence of directionality, to ensure that the result only reflects the amplitude of the change in the absorption value, and is not affected by the direction of increase or decrease, so as to obtain the absorption difference data.

[0070] The benefit of the formula is that it provides a method to quantify the spectral changes of a sample, and by calculating the average change in the absorption values ​​between adjacent wavelengths, the spectral characteristics of the sample can be revealed; The parameter acquisition step is to count the total number of wavelength points of each sample in the spectral absorption data array, which is usually determined based on the range and resolution set by the spectral measurement equipment. and The step of obtaining the parameters is to directly extract the absorption values ​​of adjacent wavelength points from the spectrum absorption data array;

[0071] Calculation process: The sample has absorption values ​​at 5 wavelengths: [100, 105, 103, 108, 107], The formula is 5. The calculation process is as follows: first calculate the absolute value of the difference between the absorption values ​​of adjacent wavelengths to obtain [5, 2, 5, 1], and then calculate the average value of these differences:

[0072] ;

[0073] The results show that the sample's absorption value changes by 3.25 units for every one wavelength unit change within the measured wavelength range, reflecting the sample's spectral variation properties and the sensitivity of the spectral response.

[0074] The steps to obtain the spectral tenderness score are as follows:

[0075] Collect the spectral change index of each sample to obtain preliminary difference pattern data;

[0076] Based on the preliminary difference pattern data, the spectral tenderness score of each sample was calculated using the formula:

[0077] ;

[0078] in, Representative samples Spectral tenderness score, and The samples and samples The spectral variation index, is the total number of samples, and The samples and samples The standard deviation of the spectral variation index;

[0079] Based on the calculated spectral tenderness score of each sample, the scores were collated and sorted to obtain a spectral tenderness score.

[0080] Specifically, the spectral change index of each sample is collected, and the near-infrared spectrum of the sample is scanned using a spectrometer. The absorption value of each sample within a specific wavelength range is continuously recorded, and then the collected raw spectral data is preprocessed, including denoising, normalization, and baseline correction to ensure the accuracy and comparability of the data. By comparing the absorption differences of different samples at the same wavelength point, the spectral change index of each sample is calculated. These spectral change indicators reflect the slight differences in the chemical composition and structure of the samples. Through the precise calculation of these spectral change indicators, preliminary difference pattern data are finally obtained, which provides a basis for subsequent data analysis and scoring;

[0081] The benefit of the formula is that by adding cubic terms and the square root of the standard deviation, the influence of spectral differences between samples is emphasized, making the score more sensitive to samples with significant spectral differences. Representative samples The spectral tenderness score represents the comprehensive difference of the sample in spectral data relative to other samples; and The samples and samples The spectral change index represents the absorption intensity of each sample at a specific wavelength and is obtained through actual spectral measurement; is the total number of samples, indicating the number of samples involved in the comparison, which is directly obtained from the experimental design; and The samples and samples The standard deviation of the spectral variation index is obtained through statistical analysis, which reflects the variability of the spectral data of each sample. The detailed explanation of the formula and the derivation process of the formula calculation are as follows: Assume that the total number of samples M=10, the spectral variation indexes of sample i and sample j are 0.5 and 0.6 respectively, and their standard deviations are 0.05 and 0.04 respectively. Substitute them into the formula for calculation:

[0082] ;

[0083] This result indicates the position of sample i in the spectral tenderness score, with higher scores indicating greater differences from other samples, reflecting its unique chemical or structural characteristics;

[0084] Based on the calculated spectral tenderness score for each sample, this process involves sorting and analyzing the calculated score data. First, the spectral tenderness scores of all samples are summarized into a list, and then a sorting algorithm is used to sort these data from high to low. During the sorting process, the samples are ranked according to the score. In this way, you can intuitively see which samples have more unique spectral features and which are more common. Finally, through this series of data processing and sorting operations, the spectral tenderness score is obtained. This score result not only reflects the relative differences between samples, but also can be used as a basis for subsequent research and quality control.

[0085] The steps to obtain color feature indicators are:

[0086] Based on the image color data array, perform the conversion of the RGB color value of each sample to the HSV color space to obtain the HSV color data array of each sample;

[0087] From the HSV color data array, the hue, saturation and lightness values ​​of each sample are calculated, and the average value of each component is calculated to form the hue average value, saturation average value and lightness average value;

[0088] The average hue, saturation and lightness of each sample are integrated to form a color characteristic index.

[0089] Specifically, converting RGB color values ​​to HSV color space includes the steps of calculating hue (H), saturation (S), and value (V). The conversion is obtained by calculating the red, green, and blue component values ​​of RGB respectively. The hue (H) calculation takes into account the wavelength distribution of the color, the saturation (S) measures the purity of the color, and the value (V) represents the brightness of the color. After the conversion is completed, the corresponding HSV color data is generated for each sample.

[0090] The average values ​​of hue, saturation, and brightness are calculated from the HSV color data. The average value of each color component is obtained by adding the corresponding color component values ​​of all pixels in the sample and dividing by the total number of pixels. Such calculation not only reflects the visual average color performance of the sample, but also shows the color changes of the sample under different lighting and viewing angles.

[0091] The calculated average values ​​of hue, saturation and brightness are integrated to form the color characteristic value of each sample. This value is obtained by comprehensively considering the statistical data of each color component and reflects the color characteristics of the sample during the entire observation period.

[0092] The steps to obtain the color tenderness score are:

[0093] From the color characteristic index, analyze the hue, saturation and brightness distribution of each sample, identify the color attribute changes between samples, and obtain color attribute change data;

[0094] Based on the color attribute change data, the color tenderness score of each sample is calculated using the following formula:

[0095] ;

[0096] in, For samples Color tenderness score, Samples The average value of hue, saturation and lightness, is the adjustment factor, For samples The weight coefficient of is the total number of samples, For comparison sample index.

[0097] Specifically, when analyzing the obtained color characteristic indicators, we first perform quantitative measurements of hue, saturation and brightness for each sample, and calculate the statistical data of each color component. These data mainly include the average value, maximum value, minimum value and standard deviation of each component; then, by comparing the differences in the color data of each sample, we analyze the similarities and differences in their color distribution, so as to identify the changing trends of color attributes.

[0098] The formula is beneficial in that it makes the score more sensitive to small changes in color by accounting for the differences in hue, saturation, and lightness between each sample and the others, and by amplifying the effects of these differences through an exponential function.

[0099] Calculation process: Set the parameters as: , , , , , the other two samples They are and .

[0100] The calculation steps are as follows: For the first sample (j=1), the weight , difference calculation:

[0101] ;

[0102] Index part calculation:

[0103] ;

[0104] Weights applied:

[0105] ;

[0106] For the second sample (j=2), the weight , difference calculation:

[0107] ;

[0108] Index part calculation:

[0109]

[0110] Weights applied:

[0111]

[0112] Final color tenderness score Calculation:

[0113]

[0114] The results show that the sample The color tenderness score of the sample was 0.68, which reflects the The degree of difference in hue, saturation, and value from other samples, with a higher score indicating a greater difference, thus indicating that the sample The color attributes of the product are more unique or distinct.

[0115] The steps to obtain the tenderness quality ranking are:

[0116] The spectral tenderness score and color tenderness score of each sample were summarized, and the correlation coefficient between the spectral tenderness score and the color tenderness score was calculated using the following formula:

[0117]

[0118] in, represents the correlation coefficient between the spectral tenderness score and the color tenderness score, and Samples Spectral tenderness score and color tenderness score, and are the averages of the spectral tenderness score and the color tenderness score, is the total number of samples;

[0119] Based on the correlation coefficient, the samples were sorted to generate a tenderness quality ranking.

[0120] Specifically, the collection and summary of spectral tenderness scores and color tenderness scores need to be carried out immediately after each test phase. First, the score data of each sample is entered into the database completely and accurately, and data verification is performed to check the completeness and accuracy of the data and eliminate any obvious input errors or outliers. The data verification criteria are based on the data range of previous experiments to ensure that all data are within safe and reasonable thresholds.

[0121] Formula calculation process: Set the number of samples n=10, and get the spectral score of each sample from the experimental data and color rating , and the average and For example, the spectral score and color score of a sample are 75 and 80 respectively, and the average spectral score is =70, average color rating = 75. The difference between each score and its mean is calculated first, and then the product and sum of squares of the differences are calculated. These values ​​are then used to calculate the correlation coefficient using the formula , confirm the size of the correlation coefficient, which will be used to guide the final quality ranking; if is close to 1, indicating that the two scores have a high positive correlation, meaning that increasing spectral tenderness will generally also increase color tenderness, and vice versa. If it is close to -1, it is a negative correlation, and the two scores show opposite performance. If it is close to 0, it means there is no obvious linear relationship between the two.

[0122] The calculated correlation coefficient is used to rank the samples according to their comprehensive tenderness quality using quick sort or merge sort. During the sorting process, the score of each sample will be converted into a ranking index. This conversion is performed based on the size of the correlation coefficient to ensure that samples with high scores are ranked higher.

[0123] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness, characterized in that: The following steps are involved: Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample at 800nm ​​to 2500nm, record the absorption value into a data array, generate a spectrum absorption data array, take an image of each sample, extract RGB color values ​​from the image, and generate an image color data array; Based on the spectral absorption data array, the absorption values ​​of each sample between adjacent wavelengths are subtracted, the obtained difference is stored in a new data array, and a spectral difference data array is generated; the numerical average value in the spectral difference data array is calculated for each sample to obtain a numerical value representing the spectral change, and a spectral change index is generated; the spectral change index of each sample is compared, the spectral difference pattern is identified, and a spectral tenderness score is generated; Based on the image color data array, converting the RGB color value of each sample into the HSV color space, calculating the hue, saturation and lightness components, generating the HSV color data array, calculating the average values ​​of the hue, saturation and lightness in the HSV color data array for each sample, obtaining a numerical value representing the color feature, generating a color feature index, analyzing the distribution of the color feature index of each sample, identifying the change of color attributes, and generating a color tenderness score; The spectral tenderness score and color tenderness score of each sample were combined, the correlation coefficient between the scores was calculated, and the samples were sorted according to the correlation coefficient to generate a tenderness quality ranking.

2. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps of acquiring the spectral absorption data array are as follows: Acquire Hainan mulberry leaf green tea samples, measure the near infrared spectrum absorption value of each sample in the range of 800nm ​​to 2500nm, and obtain spectrum absorption data of each sample; Arrange the spectral absorption data of each sample into a data array and store them in order of wavelength to obtain an arranged spectral absorption data array; The integrity and accuracy of the sorted spectral absorption data array are verified to obtain the spectral absorption data array.

3. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps for obtaining the image color data array are: Collect images of each Hainan mulberry leaf green tea sample to obtain an image dataset of each sample; Based on the image data set of each sample, each image is analyzed, the red, green and blue channel values ​​in the image are extracted, the numerical distribution of each color is recorded, and the RGB color data of each sample is obtained; According to the RGB color data of each sample, data arrangement and standardization are performed, and the RGB values ​​are arranged into the same data structure according to the sample to form an image color data array.

4. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps for obtaining the spectrum change index are: Based on the spectral absorption data array, the absorption values ​​of each sample at adjacent wavelengths are subtracted, and the difference in absorption values ​​between two consecutive wavelength points is calculated each time to obtain absorption difference data; Based on the absorption difference data, the numerical average of all differences in the spectral difference data array of each sample is calculated to obtain a spectral change index. The calculation formula of the numerical average is: ； in, is the numerical average of the spectral variation, is the total number of wavelength points, and are the absorbance values ​​at consecutive wavelength points.

5. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps for obtaining the spectral tenderness score are as follows: Collect the spectral change index of each sample to obtain preliminary difference pattern data; Based on the preliminary difference pattern data, the spectral tenderness score of each sample was calculated using the formula: ； in, Representative samples Spectral tenderness score, and Samples and samples The spectral variation index, is the total number of samples, and The samples and samples The standard deviation of the spectral variation index; Based on the calculated spectral tenderness score of each sample, the scores were collated and sorted to obtain a spectral tenderness score.

6. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps of obtaining the color characteristic index are as follows: Based on the image color data array, convert the RGB color value of each sample into the HSV color space to obtain the HSV color data array of each sample; Calculate the hue, saturation and lightness values ​​of each sample from the HSV color data array, and calculate the average value of each component to form the hue average value, saturation average value and lightness average value; The average hue, saturation and lightness of each sample are integrated to form a color characteristic index.

7. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps for obtaining the color tenderness score are: Analyzing the hue, saturation and brightness distribution of each sample from the color characteristic index, identifying the color attribute changes between samples, and obtaining color attribute change data; Based on the color attribute change data, the color tenderness score of each sample is calculated using the following formula: ； in, For samples Color tenderness score, The samples The average value of hue, saturation and lightness, is the adjustment factor, For samples The weight coefficient of is the total number of samples, For comparison sample index.

8. The method for identifying the difference in taste quality of Hainan mulberry leaf green tea with different tenderness according to claim 1, characterized in that: The steps for obtaining the tenderness quality ranking are: The spectral tenderness score and color tenderness score of each sample were summarized, and the correlation coefficient between the spectral tenderness score and the color tenderness score was calculated using the following formula: ； in, represents the correlation coefficient between the spectral tenderness score and the color tenderness score, and The samples Spectral tenderness score and color tenderness score, and are the averages of the spectral tenderness score and the color tenderness score, is the total number of samples; Based on the correlation coefficients, the samples were ranked to generate a tenderness quality ranking.

Images