A digital evaluation method for tea leaf grade
The tea strip index is calculated through python programming and the edge detection algorithm of the opencv library, and combined with the normal distribution principle and data transformation, an objective evaluation standard is established, which solves the subjectivity and lack of quantification of the existing tea strip evaluation method, and realizes the Digital evaluation and objective scoring of tea strips.
Patent Information
- Application Number
- CN202211200229.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-29
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing tea strip evaluation method relies on professionals, is subjective and lacks objectivity, cannot give quantitative values, and has not achieved digital evaluation of quality.
Using python programming and edge detection and skeleton detection algorithms in the opencv library, tea images are obtained through shadowless lamp photography, calculating the wire index, and using normal distribution principles and data transformation methods to establish objective evaluation standards to realize the quantification and division of wire index index files.
The quantitative evaluation and objective scoring of tea strips are realized, and the strips of black tea and green tea can be quickly and accurately evaluated, improving the digital evaluation ability of tea quality.
Smart Images

Figure CN115493519B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of tea quality control, and in particular to a digital evaluation method for tea strip grades. Background Art
[0002] The strip of tea refers to the overall tightness of dry tea, which means that the tea is in the shape of strips, as tight as a twisted rope, such as Xinyang Maojian, Lishui fragrant tea, Gongfu black tea, etc. Strip is one of the important evaluation indicators of tea appearance quality.
[0003] The existing evaluation methods have technical deficiencies including:
[0004] 1. The evaluation of the strips depends on professionals, the evaluation method is subjective and the comments are relatively general;
[0005] 2. There is no objective evaluation method, and the quantitative values of the strip indicators cannot be given. The digital evaluation of the strip quality has not been achieved, and there is no objective evaluation standard.
[0006] Therefore, establishing a quantitative method for tea (dry tea) strips and an objective evaluation method for their grade is of great significance for the objective evaluation and digital improvement of tea quality. Summary of the invention
[0007] The purpose of the present invention is to provide a digital evaluation method for tea leaves, which can quickly complete the digital evaluation of black tea and green tea strips (dry tea) to solve the problems raised in the above background technology.
[0008] To achieve the above object, the present invention provides the following technical solutions:
[0009] A digital evaluation method for tea strip grade comprises the following steps:
[0010] S1. Collect no less than 50 dried tea leaves at a time and spread them evenly on a white wooden tray so that the leaves do not overlap each other;
[0011] S2. Obtain sample images: Use a camera to shoot vertically from the top of the shadowless light studio through a shadowless light to collect clear images of tea leaves without shadows;
[0012] S3. Using Python programming, the edge detection and skeleton detection algorithms in the OpenCV library are used to automatically identify and extract the image of the image evaluation part, and the stripe index of this batch of dry tea is measured using the stripe index measurement calculation method. The stripe index measurement calculation formula is:
[0013] Strip index ;
[0014] Further, TI: stripe index; LA: tea leaf area; SL: tea leaf skeleton length; C: constant, take 10, quantify the stripe index to a suitable dimension, take the single digit;
[0015] S4. Establish an objective evaluation standard for the tea strips by using the standard samples of the tea to be tested collected in the early stage, that is, adopt the normal distribution principle for grading;
[0016] S5. Using the objective evaluation standard of the tea leaves established in the early stage, each tea leaf is graded as "tight, medium, loose", and the proportion of "tight, medium, loose" and the average value of the tea leaves index are obtained;
[0017] Furthermore, the tea skeleton refers to a support for the basic outline of the tea leaves, and the skeleton extraction is obtained through image refinement, wherein the image refinement is performed by peeling off layer by layer, removing some points from the original image while maintaining the original shape until the skeleton of the image is obtained.
[0018] Furthermore, the objective evaluation method of the strip grade is as follows: based on a certain type of black tea or green tea strip, 4-6 levels of standard samples are taken, and the strip index of the samples is measured. If the sample population conforms to the normal distribution, the normal distribution principle can be used for grading; if the sample population does not conform to the normal distribution, the original sample is converted into a normal population through inverse transformation, logarithmic transformation or Box-Cox transformation data conversion method, and then the normal distribution principle is used for grading.
[0019] Furthermore, the normal distribution classification includes:
[0020] If the array x approximately follows the normal distribution x~N(μ,σ 2 ), then this set of values is usually considered to be normally distributed;
[0021] Those exceeding μ + σ are rated as A;
[0022] The distribution between μ and μ+σ is rated as B;
[0023] Those distributed between μ- σ and μ are rated C;
[0024] μ-σ to μ-2σ are rated as D;
[0025] Those below μ-2σ are rated as E.
[0026] Furthermore, 5 grades of standard samples of the tea "Guizhou Babu Zijuan Black Tea" were taken, with about 250-300 samples in each grade, and a total of 1343 samples were drawn. The stripe index was calculated according to the method of the present invention. After testing, the skewness = 0.72>0, p<0.05, and the stripe index of the "Guizhou Babu Zijuan Black Tea" was not normally distributed as a whole.
[0027] Furthermore, 5 grades of standard samples of "Lishui Fragrant Tea" were taken, with about 250-300 samples in each grade, and a total of 1275 samples were drawn. The stripe index was calculated according to the method of the present invention. After inspection, the skewness = 0.746>0, p<0.05, and the stripe index of "Lishui Fragrant Tea" was not normally distributed overall.
[0028] Furthermore, according to the principles of probability theory and mathematical statistics, in order to make the original data generally normally distributed, the stripe index of the Guizhou Babu Zijuan black tea standard sample was Box-Cox transformed [2]. After testing, the skewness was ≈0, p>0.05, and the tea was generally normally distributed after the transformation.
[0029] Similarly, the stripe index of the Lishui fragrant tea standard sample was transformed by Box-Cox transformation[2]. After testing, the skewness was ≈0, p>0.05. After the transformation, the tea was generally normally distributed.
[0030] Furthermore, according to the principles of probability theory and mathematical statistics, a set of values is divided into 5 levels. After appropriate adjustments, it is adjusted to 3 levels, which are [0, μ-σ], [μ-σ, μ+σ], and [μ+σ, +∞], accounting for 15.87%, 68.27%, and 15.87% respectively. The physical standard samples of the two kinds of tea are divided into 3 levels respectively. Among them, the grading threshold is obtained by the normal parameter through the Box-Cox inverse transformation:
[0031] Guizhou Babu Zijuan black tea classification threshold interval: [0, 1.89], [1.89, 2.66], [2.66, +∞];
[0032] Lishui fragrant tea classification threshold interval: [0,1.66], [1.66,2.89], [2.89,+∞];
[0033] Technical effects and advantages of the present invention:
[0034] 1. The present invention proposes a calculation formula for the stripe index, uses Python programming, and adopts the edge detection and skeleton detection algorithms in the opencv library to obtain the stripe index, thereby realizing the quantification of the stripes and achieving quantitative evaluation;
[0035] 2. The present invention collects standard samples of a certain type of tea, adopts statistical principles, establishes the threshold range of the strip index, establishes an objective evaluation method, and realizes objective evaluation. It can be used for the strip evaluation of black tea and green tea strips (dry tea). BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 A flow chart of a digital evaluation method for tea strip grade of the present invention;
[0037] Figure 2 It is a schematic diagram of the process of extracting tea skeleton and calculating the tea leaf index of the present invention;
[0038] Figure 3 This is an example diagram of the skeleton morphology of 5 samples of Guizhou Babu Zijuan black tea leaves of the present invention;
[0039] Figure 4 This is an example diagram of the skeleton morphology of 5 samples of Lishui fragrant tea leaves of the present invention;
[0040] Figure 5 It is the standard spline index diagram of Guizhou Babu Zijuan black tea of the present invention;
[0041] Figure 6 It is the standard spline index diagram of Lishui fragrant tea of the present invention;
[0042] Figure 7 The frequency histogram and normal fitting result after the black tea data conversion of the present invention;
[0043] Figure 8 The frequency histogram and normal fitting result of the scented tea data after conversion of the present invention;
[0044] Fig. 9 This is the graded diagram of Guizhou Babu Zijuan black tea strips of the present invention;
[0045] Fig.10 This is a graded diagram of the Lishui fragrant tea strips of the present invention;
[0046] Fig.11 Collecting images of Guizhou Babu Zijuan black tea (dry tea) of the present invention;
[0047] Fig.12 This is a comparison chart of the proportions of three rope levels and the mean values of the rope index in 58 tea leaves in Example 1 of the present invention;
[0048] Fig.13 Collecting images of Lishui fragrant tea samples of the present invention;
[0049] Fig.14 This is a comparison chart of the proportions of three rope grades and the mean values of the rope index in 56 tea leaves in Example 2 of the present invention. DETAILED DESCRIPTION
[0050] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0051] The tea skeleton refers to a support for the basic outline of the tea leaves. The skeleton extraction is obtained through image refinement. The image refinement is to remove some points from the original image through layer-by-layer peeling while maintaining the original shape until the skeleton of the image is obtained.
[0052] The present invention provides a digital evaluation method for tea leaf grades: Example 1
[0053] A digital evaluation method for Guizhou Babu Zijuan black tea (dry tea) strips specifically comprises the following steps:
[0054] S1. Randomly take 58 pieces of Guizhou Babu Zijuan black tea dry tea samples and evenly spread them on a white background cloth under a shadowless lamp so that the tea leaves do not overlap each other;
[0055] S2. Use photographic equipment to shoot vertically from the top of the shadowless light studio to capture a clear image of the tea leaves without shadows, so that all the tea leaves are in the photo without too much blank space, such as Fig.11 As shown;
[0056] S3, import the image into Python, use the target recognition algorithm to automatically identify and extract the image of the tea evaluation part, and use the stripe index measurement and calculation method to measure the stripe index of 58 pieces of dry tea;
[0057] S4. Using the standard samples of Guizhou Babu Zijuan black tea collected in the early stage, an objective evaluation standard for the tested tea samples was constructed, that is, the normal distribution principle was used for classification;
[0058] S5. Based on the objective evaluation criteria of the stripe index of Guizhou Babu Zijuan black tea, the proportion of the three stripe levels in 58 tea leaves and the mean of the stripe index were obtained, such as Fig.12 shown.
[0059] Specifically, a total of 5 grades of standard samples of "Guizhou Babu Zijuan Black Tea" were taken, with about 250-300 samples of each grade, and a total of 1343 samples were drawn. The stripe index was calculated according to the method of the present invention, such as Figure 5 As shown, after testing, the skewness = 0.72>0, p<0.05, the stripe index of "Guizhou Babu Zijuan Black Tea" was not normally distributed overall.
[0060] Specifically, according to the principles of probability theory and mathematical statistics, in order to make the original data generally normally distributed, the stripe index of the Guizhou Babu Zijuan black tea standard sample was Box-Cox transformed, as follows: Figure 7 As shown, after testing, the skewness is ≈ 0, p> 0.05, and the tea leaves are generally normally distributed after transformation.
[0061] Specifically, according to the principles of probability theory and mathematical statistics, a set of values is divided into 5 levels. After appropriate adjustments, it is adjusted to 3 levels, which are [0, μ-σ], [μ-σ, μ+σ], and [μ+σ, +∞], accounting for 15.87%, 68.27%, and 15.87% respectively. The physical standard samples of the two kinds of tea are divided into 3 levels respectively. Among them, the grading threshold is obtained by the normal parameter through the Box-Cox inverse transformation:
[0062] The threshold interval for Guizhou Babu Zijuan black tea classification is: [0, 1.89], [1.89, 2.66], [2.66, +∞], such as Fig. 9 shown. Example 2
[0063] A digital evaluation method for Lishui fragrant tea (dry tea) strips specifically comprises the following steps:
[0064] S1. Randomly take 56 pieces of Lishui fragrant tea dry tea samples and evenly spread them on a white background cloth under a shadowless lamp so that the tea leaves do not overlap each other;
[0065] S2. Use photographic equipment to shoot vertically from the top of the shadowless light studio to capture clear images of tea leaves without shadows, so that all tea leaves are in the photo without excessive blank space;
[0066] S3, import the image into Python, use the target recognition algorithm to automatically identify and extract the image of the tea evaluation part, and use the stripe index measurement and calculation method to measure the stripe index of 56 pieces of dry tea;
[0067] S4. Using the standard samples of Lishui fragrant tea collected in the early stage, an objective evaluation standard for the tested tea samples was constructed, that is, the normal distribution principle was used for classification;
[0068] S5. Based on the objective evaluation criteria of the stripe index of Lishui fragrant tea, the proportion of three stripe grades in 56 tea leaves and the average of the stripe index were obtained.
[0069] Specifically, the objective evaluation method for the strip grade is as follows: based on a certain type of black tea or green tea strip, 4-6 levels of standard samples are taken, and the strip index of the samples is measured. If the sample population conforms to the normal distribution, the normal distribution principle can be used for grading; if the sample population does not conform to the normal distribution, the original sample is converted into a normal population through inverse transformation, logarithmic transformation or Box-Cox transformation data conversion method, and then the normal distribution principle is used for grading.
[0070] Specifically, the normal distribution classification includes:
[0071] If the array x approximately follows the normal distribution x~N(μ,σ 2 ), then this set of values is usually considered to be normally distributed;
[0072] Those exceeding μ + σ are rated as A;
[0073] The distribution between μ and μ+σ is rated as B;
[0074] Those distributed between μ- σ and μ are rated C;
[0075] μ-σ to μ-2σ are rated as D;
[0076] Those below μ-2σ are rated as E.
[0077] Specifically, 5 grades of standard samples of "Lishui Fragrant Tea" were taken, with about 250-300 samples for each grade, and a total of 1275 samples were drawn. The stripe index was calculated according to the method of the present invention, such as Figure 6 As shown, after testing, the skewness = 0.746>0, p<0.05, the stripe index of "Lishui Fragrant Tea" was not normally distributed overall.
[0078] Similarly, the stripe index of the Lishui fragrant tea standard sample is transformed by Box-Cox transformation [2], as follows: Figure 8 As shown, after testing, the skewness is ≈ 0, p> 0.05, and the tea leaves are generally normally distributed after transformation.
[0079] Lishui fragrant tea classification threshold interval: [0,1.66], [1.66,2.89], [2.89, +∞], such as Fig.10 As shown;
[0080] Using the large database established in the early stage and the obtained classification threshold, an objective evaluation standard for the strips of Guizhou Babu Zijuan black tea and Lishui fragrant tea was established. The objective evaluation standards for the two tea strips defined in the present invention are as follows:
[0081]
[0082] Table 1: Evaluation criteria for Guizhou Babu Zijuan black tea strips.
[0083]
[0084] Table 2: Evaluation criteria for Lishui fragrant tea strips.
[0085] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
[0086] It should be noted that, in this article, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or apparatus.
Claims
1. A digital evaluation method for tea leaf grades, characterized by: The steps include: S1. Collect no less than 50 dried tea leaves at a time and spread them evenly on a white wooden tray so that the leaves do not overlap each other; S2. Obtain sample images: Use a camera to shoot vertically from the top of the shadowless light studio through a shadowless light to collect clear images of tea leaves without shadows; S3. Using Python programming, the edge detection and skeleton detection algorithms in the OpenCV library are used to automatically identify and extract the image of the image evaluation part, and the stripe index of this batch of dry tea is measured using the stripe index measurement calculation method. The stripe index measurement calculation formula is: Strip index ; Wherein, TI: stripe index; LA: tea leaf area; SL: tea leaf skeleton length; C: any constant; in order to better quantify the numerical range, C=10 is taken, and the stripe index is quantified to a suitable dimension, taking the single digit; S4. Establish an objective evaluation standard for the tea strips by using the standard samples of the tea to be tested collected in the early stage, that is, adopt the normal distribution principle for grading; S5. Using the objective evaluation standard of the tea leaves established in the early stage, each tea leaf is graded as "tight, medium, loose", and the proportion of "tight, medium, loose" and the average value of the tea leaves index are obtained; The tea skeleton refers to a support for the basic outline of the tea leaves. The skeleton is obtained by image thinning. The image thinning is to remove some points from the original image layer by layer while maintaining the original shape until the skeleton of the image is obtained. The width is one pixel, and the area of the pixel is the length SL of the tea skeleton. In the stripe index TI formula, after the tea image is binarized, all pixels with grayscale greater than or equal to a self-defined threshold are determined to belong to a specific object, and the pixel size is the tea leaf area LA.
2. A digital evaluation method for tea strand grade according to claim 1, characterized in that: The objective evaluation method of the strip grade is as follows: based on a certain type of black tea or green tea strip, 4-6 levels of standard samples are taken, and the strip index of the samples is measured. If the sample population conforms to the normal distribution, the normal distribution principle can be used for grading; if the sample population does not conform to the normal distribution, the original sample is converted into a normal population through inverse transformation, logarithmic transformation or Box-Cox transformation data conversion method, and then the normal distribution principle is used for grading.
3. A digital evaluation method for tea strand grade according to claim 2, characterized in that: The normal distribution classification includes: If the array x approximately follows the normal distribution x~N(μ,σ 2 ), then this set of values is usually considered to be normally distributed; Those exceeding μ + σ are rated as A; The distribution between μ and μ+σ is rated as B; Those distributed between μ- σ and μ are rated C; μ-σ to μ-2σ are rated as D; Those below μ-2σ are rated as E.
4. A digital evaluation method for tea strand grade according to claim 3, characterized in that: According to the principles of probability theory and mathematical statistics, a group of values are divided into 5 levels. After appropriate adjustments, they are adjusted to 3 levels, which are [0, μ-σ], [μ-σ, μ+σ], and [μ+σ, +∞], accounting for 15.87%, 68.27%, and 15.87% respectively. The physical standard samples of the tested tea are divided into 3 levels respectively. Among them, the grading threshold is obtained by the normal parameter through the Box-Cox inverse transformation.
