A citrus fruit dry sorting method for internal quality detection

Abstract

The application discloses a citrus fruit drought detection method, and relates to the technical field of fruit and vegetable sorting and detection. The citrus fruit is initially screened by judging whether the citrus fruit is in a normal condition about density within a membership range. The citrus fruit is secondarily screened by jointly judging the citrus fruit about titratable acid and the citrus fruit about sugar-acid ratio within a membership function. The citrus fruit is classified into a normal condition, an obvious drought condition and a slight drought condition to divide levels. In the sorting process, a hyperspectral imager is selectively used to obtain spectral data of part of the citrus fruit to obtain predicted values of the titratable acid and the sugar-acid ratio. The accuracy and operability of the citrus fruit drought sorting are effectively improved, the processing and calculation amount of the spectral data is reduced, and the citrus fruit slight drought condition has a better application effect.

Description

Technical Field

[0003] This invention relates to the field of fruit and vegetable sorting and testing technology, and in particular to a method for sorting citrus fruits by removing moisture for internal quality testing. Background Technology

[0005] Citrus fruit dehydration is a common physiological disorder affecting citrus fruits during ripening and post-harvest storage, severely impacting fruit quality and storage performance. Dehydration is characterized by significant quality loss in citrus fruits, separation of peel and pulp, thickening and hardening of segments, granulation of juice vesicles, substantial reduction in nutrients and juice content, and loss of sweetness and aroma. Citrus fruit dehydration is a complex physiological process closely related to cell development, nutrient metabolism, mineral element content, and endogenous hormone levels. It is also influenced by factors such as species, variety, rootstock, pre-harvest environment, pre-harvest management, harvest maturity, and post-harvest storage and treatment.

[0006] Dehydration of citrus pulp negatively impacts the commercial quality of navel oranges, leading to a downgrade in sorting grade. Severe dehydration can almost completely destroy their edible value, affecting their market worth. During the commercial processing and sorting of citrus, existing mass and image sensors can typically be used to calculate fruit density and identify dehydration, enabling rapid, non-destructive testing. However, in practice, the irregular shape of citrus fruits leads to significant errors in volume estimation, making it difficult to accurately identify even minor dehydration and thus affecting the overall performance of internal quality testing. Therefore, addressing these issues is a crucial aspect of current research in citrus internal quality testing technology. Summary of the Invention

[0008] This invention proposes a method for sorting citrus fruits with slight water loss for internal quality detection. Considering the significant estimation deviation of volume during the commercial processing and sorting of citrus fruits, which affects the accuracy of fruit density, the method first performs initial screening using the membership function of citrus fruits with normal density. Then, it performs secondary screening using the membership functions of citrus fruits with normal titratable acidity and normal sugar-acid ratio. During the sorting process, a hyperspectral imager is selectively used to acquire spectral data from a portion of the citrus fruits to obtain predicted values ​​for titratable acidity and sugar-acid ratio. This effectively improves the accuracy and operability of sorting citrus fruits with slight water loss while reducing the computational burden of spectral data processing. It also shows good application results for detecting slight water loss in citrus fruits.

[0009] The inventive concept of this invention is as follows: a method for sorting citrus fruits by moisture content based on internal quality detection. This method involves setting membership functions for citrus fruits regarding normal density, normal titratable acidity, and normal sugar-acid ratio. Initial screening is achieved by determining whether the fruit falls within the membership range of normal density or the membership range of moisture content based on density. For other cases, secondary screening is achieved through a joint judgment using the membership functions of normal titratable acidity and normal sugar-acid ratio. This allows citrus fruits to be categorized into normal, significantly moisture-depleted, and slightly moisture-depleted conditions, effectively classifying citrus fruits based on the presence or absence of moisture-depleted fruit in the internal quality detection process.

[0010] This invention is achieved through the following measures: a method for sorting citrus fruits by moisture content for internal quality testing, comprising the following steps:

[0011] Step 1: Set the membership function of citrus fruit density under normal conditions for the initial screening of citrus fruit moisture selection;

[0012] Step 2: Set the membership functions of citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio for the secondary screening of citrus fruit with dry water content.

[0013] Step 3: For citrus fruits that need to be sorted by dryness, the initial screening can be made by judging the membership function of the citrus fruit density relative to the normal condition. If the citrus fruit density is within the membership range of the normal condition, it is judged to be in normal condition and can be classified as first grade or above. If the citrus fruit density is within the membership range of the dry condition, it is judged to be in obvious dry condition and can be classified as not eligible for grade. Other cases will proceed to the second screening in Step 4.

[0014] Step 4: After the citrus fruits that need to be sorted by moisture are put into the second screening, a joint judgment can be made by the membership function of the citrus fruit with respect to the normal condition of titratable acid and the membership function of the citrus fruit with respect to the normal condition of sugar-acid ratio. If the citrus fruit is within the membership range of both the normal condition of titratable acid and the normal condition of sugar-acid ratio, it is judged to be in normal condition and can be classified as first grade or above. If the citrus fruit is in the membership range of either the condition of moisture loss or the condition of moisture loss of sugar-acid ratio, it is judged to be in obvious moisture loss and can be classified as not eligible for grade. In other cases, it is judged to be in slight moisture loss and can be classified as second grade.

[0015] As a further optimization of the citrus dryness separation method for intrinsic quality detection provided by the present invention, the specific content of step 1 is as follows:

[0016] For a certain type of citrus that requires dry sorting, the selection... A citrus fruit sample set will be constructed by identifying citrus fruits that are in good condition and ready for market but may be dehydrated, and by assigning numbers to these fruits. ,in, The number of samples of this type of citrus used in the drought test, and the citrus fruit number. satisfy and ; by obtaining the Construct a citrus fruit density dataset based on the density of different citrus fruits. Among them, for citrus fruits density It is possible The result is given after measurement and calculation. For the citrus fruit The mass measured by an electronic scale For the citrus fruit The volume measured by the displacement method; by obtaining the... Constructing a spectral dataset of citrus fruits using spectral data from different citrus fruits. Among them, for citrus fruits spectral data Citrus fruits can be observed using a hyperspectral imager The results are obtained after detection and spectral preprocessing; by the... After cutting citrus fruits along the equatorial plane, it is manually determined whether they are dehydrated. Citrus fruits that are determined by the naked eye not to be dehydrated are classified into the normal citrus fruit subset. Citrus fruits that are visually identified as being in a state of water scarcity are categorized into a subset of citrus fruits in a state of water scarcity. and according to the and Construct datasets of citrus fruit density under normal conditions. Data set on citrus fruit density under drought conditions Considering that the fruit density of citrus fruits under normal conditions is greater than that under drought conditions, we define the domain of discourse for citrus fruit density as being expressible through fuzzy mathematics. U = [min( C 2 , 2 ),max( C 2 , 1 )] ,in, To find the minimum value of a set, To find the function that maximizes a set, we use fuzzy subsets. To indicate "citrus fruit density is at a normal level" to distinguish it from "citrus fruit density is in a low-water condition," the following settings are configured. membership function for:

[0017] m A ˜ ( x ) = { 1 ， x ≥ E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) [ x − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) ] k A ˜ , E ( C 2 , 2 ) + 2 D ( C 2 , 2 ) < x < E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) 0 , x ≤ E ( C 2 , 2 ) + 2 D ( C 2 , 2 )

[0018] in, As the independent variable, To find the function of the average of a set, To find a function of the population standard deviation of a set, For the The power exponent can be used to set the boundary membership value of citrus fruits under normal conditions. Then, the aforementioned Determined as log [ min( C 2 , 1 ) − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) ] x ;

[0019] As a further optimization of the citrus dryness separation method for intrinsic quality detection provided by the present invention, the specific content of step 2 is as follows:

[0020] The and After juicing individual citrus fruits cut in half, titratable acid and sugar-acid ratio were measured to construct a dataset of titratable acid from citrus fruits under normal conditions. Data set of titratable acidity data for citrus fruits under drought conditions Data set of sugar-acid ratio of citrus fruits under normal conditions And data set on the sugar-acid ratio of citrus fruits under drought conditions. ,in, For the citrus fruit The titratable acid value was obtained through quality analysis. This is to determine the sugar-acid ratio of citrus fruits through quality analysis; a setting can be used to... The citrus fruit in the text corresponds to the After filtering for specific long wavelengths from the spectral data, principal component analysis was used to process them separately. and Therefore, predictive models for titratable acidity and sugar-acid ratio of citrus fruits were established, respectively. Using the union notation, the predictive model for titratable acidity in citrus fruits can be represented as follows: The predictive model for the sugar-acid ratio of citrus fruits can be expressed as follows: , Citrus fruit Titratable acid values ​​were obtained using a predictive model of titratable acidity in citrus fruits. Citrus fruit Titratable acid values ​​obtained from a predictive model of the sugar-acid ratio in citrus fruits. and These are the constant terms in the prediction models for titratable acid and sugar-acid ratio of citrus fruit, respectively. This is the set of wavelengths used in the predictive model of titratable acidity for citrus fruits. This is the set of wavelengths used in the predictive model of sugar-acid ratio for citrus fruits. and The respective and Wavelength number in wavelength Weighting coefficients in the predictive model of titratable acidity in citrus fruit. wavelength Weighting coefficients in the prediction model of sugar-acid ratio for citrus fruits. Citrus fruit Regarding wavelength In the The values ​​corresponding to the mid-spectral data, Citrus fruit Regarding wavelength In the The values ​​corresponding to the spectral data;

[0021] Considering that the titratable acid values ​​of citrus fruits under normal conditions are greater than those under dry conditions, we define the domain of discourse regarding the titratable acid values ​​of citrus fruits as being expressible through fuzzy mathematics. V = [min( C 3 , 2 ),max( C 3 , 1 )] Through fuzzy subsets To indicate that "citrus fruit is in a normal condition regarding titratable acidity" to distinguish it from "citrus fruit is in a low-water condition regarding titratable acidity," the following settings are configured. membership function for:

[0022] m B ˜ ( y ) = { 1 ， y ≥ E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) [ y − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) ] k B ˜ , E ( C 3 , 2 ) + 2 D ( C 3 , 2 ) < y < E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) 0 , y ≤ E ( C 3 , 2 ) + 2 D ( C 3 , 2 )

[0023] in, As the independent variable, For the The power exponent can be used to describe Determined as log [ min( C 3 , 1 ) − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) ] x ;

[0024] Considering that the sugar-acid ratio of citrus fruits under normal conditions is lower than that of citrus fruits under dry conditions, we define the domain of discourse regarding the titratable acid of citrus fruits as being expressible through fuzzy mathematics. W = [min( C 4 , 1 ),max( C 4 , 2 )] Through fuzzy subsets To indicate that "the sugar-acid ratio of citrus fruit is in a normal state" to distinguish it from "the sugar-acid ratio of citrus fruit is in a low-water state," the following settings are configured. membership function for:

[0025] m C ˜ ( z ) = { 0 ， z ≥ E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) [ E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − z E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − E ( C 4 , 1 ) − 2 D ( C 4 , 1 ) ] k C ˜ , E ( C 4 , 1 ) + 2 D ( C 4 , 1 ) < y < E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) 1 , z ≤ E ( C 4 , 1 ) + 2 D ( C 4 , 1 )

[0026] in, As the independent variable, For the The power exponent can be used to describe Determined as log [ E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − max( C 4 , 1 ) E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − E ( C 4 , 1 ) − 2 D ( C 4 , 1 ) ] x ;

[0027] As a further optimization of the citrus dryness separation method for intrinsic quality detection provided by the present invention, the specific content of step 3 is as follows:

[0028] During the citrus sorting process, in the dry season detection and sorting stage...

[0029] For citrus fruits requiring dry-water sorting, the initial screening can be conducted using a membership function of the citrus fruit's density relative to a normal state. The membership ranges for citrus fruits with normal density and those with dry-water conditions can be set as follows: [ t 1 , 1 , 1 ] and [ 0 , t 1 , 2 ] ,in, and These are the lower limit of membership for fruits with normal density and the upper limit of membership for fruits with low water content; for citrus fruits that need to be sorted by water content. ,in This refers to the numbering of citrus fruits during the dry season sorting process. (The last sentence appears to be incomplete and possibly refers to a separate context.) density for Afterwards, among them, The citrus sorting equipment uses an online weighing method to measure the weight of citrus fruits. The weight, Three-dimensional laser scanning is used to measure citrus fruits in citrus sorting equipment. The volume of the citrus fruit can determine its size. Regarding the membership function value when the density is under normal conditions for:

[0030] m A ˜ ( r h ) = { 1 ， r h ≥ E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) [ r h − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) − E ( C 2 , 2 ) − 2 D ( C 2 , 2 ) ] k A ˜ , E ( C 2 , 2 ) + 2 D ( C 2 , 2 ) < r h < E ( C 2 , 1 ) − 2 D ( C 2 , 1 ) 0 , r h ≤ E ( C 2 , 2 ) + 2 D ( C 2 , 2 )

[0031] Considering that in the citrus grading system, premium and first-grade citrus fruits are required to be free of withered fruit, while second-grade fruits are allowed to have slight water loss, the following setting is made: m A ˜ ( r h ) ∈ [ t 1 , 1 , 1 ] Then determine the citrus fruit This is the normal condition and can be classified as first-class or above. m A ˜ ( r h ) ∈ [ 0 , t 1 , 2 ] Then determine the citrus fruit If the water level is clearly low and the level can be classified as not eligible for further classification, then... Then proceed to the secondary screening in step 4;

[0032] As a further optimization of the citrus dryness separation method for intrinsic quality detection provided by the present invention, the specific content of step 4 is as follows:

[0033] During the secondary screening stage of citrus sorting, specifically the water shortage detection stage, a joint judgment can be made using the membership functions of citrus fruit regarding whether its titratable acidity is within a normal range and the membership function of citrus fruit regarding whether its sugar-acid ratio is within a normal range. The membership ranges for citrus fruit regarding whether its titratable acidity is within a normal range and the membership ranges for citrus fruit regarding whether its titratable acidity is within a water shortage range can be set as follows: [ t 2 , 1 , 1 ] and [ 0 , t 2 , 2 ] It is possible to set the membership range for citrus fruits with a normal sugar-acid ratio and the membership range for citrus fruits with a low sugar-acid ratio. [ t 3 , 1 , 1 ] and [ 0 , t 3 , 2 ] ,in, and These are the lower limit of membership for titratable acid under normal conditions and the upper limit of membership for titratable acid under low water conditions, respectively. and These represent the lower limit of membership for a sugar-acid ratio under normal conditions and the upper limit of membership for a sugar-acid ratio under low water conditions, respectively.

[0034] For citrus fruits that need to be sorted by dryness During the secondary screening, spectral data after spectral preprocessing can be obtained using a hyperspectral imager on the citrus sorting equipment. This can identify citrus fruits. Membership function values ​​of titratable acids under normal conditions And the membership function value with respect to the normal sugar-acid ratio. They are respectively:

[0035] m B ˜ ( f ^ h ) = { 1 ， f ^ h ≥ E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) [ f ^ h − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) − E ( C 3 , 2 ) − 2 D ( C 3 , 2 ) ] k B ˜ , E ( C 3 , 2 ) + 2 D ( C 3 , 2 ) < f ^ h < E ( C 3 , 1 ) − 2 D ( C 3 , 1 ) 0 , f ^ h ≤ E ( C 3 , 2 ) + 2 D ( C 3 , 2 )

[0036] m C ˜ ( ϕ ^ h ) = { 0 ， ϕ ^ h ≥ E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) [ E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − ϕ ^ h E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) − E ( C 4 , 1 ) − 2 D ( C 4 , 1 ) ] k C ˜ , E ( C 4 , 1 ) + 2 D ( C 4 , 1 ) < ϕ ^ h < E ( C 4 , 2 ) − 2 D ( C 4 , 2 ) 1 , ϕ ^ h ≤ E ( C 4 , 1 ) + 2 D ( C 4 , 1 )

[0037] in, Citrus fruit Titratable acid values ​​obtained from a predictive model of titratable acidity in citrus fruits can be obtained through [further details needed]. Give, Citrus fruit Regarding wavelength In the The corresponding value in Citrus fruit The sugar-acid ratio values ​​obtained from the prediction model of sugar-acid ratio in citrus fruits can be specifically obtained through... Give, Citrus fruit Regarding wavelength In the The corresponding value in;

[0038] If the setting is m B ˜ ( f ^ h ) ∈ [ t 2 , 1 , 1 ] and m C ˜ ( ϕ ^ h ) ∈ [ t 3 , 1 , 1 ] Then determine the citrus fruit This is the normal condition and can be classified as first-class or above. m B ˜ ( f ^ h ) ∈ [ 0 , t 2 , 2 ] or m C ˜ ( ϕ ^ h ) ∈ [ 0 , t 3 , 2 ] Then determine the citrus fruit If the water level is clearly low, the grade can be classified as "not eligible for grading"; otherwise, the citrus fruit will be judged as "unsuitable for grading". It is in a state of slight drought and can be classified as a second-class product. Attached Figure Description

[0040] Figure 1 The diagram shows a flowchart of a method for separating citrus fruits by moisture content for internal quality testing, as proposed in this invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0043] The present invention proposes a method for sorting citrus fruits by moisture content for internal quality detection, such as... Figure 1 As shown, the specific steps include the following:

[0044] Step 1: Set the membership function for the initial screening of citrus fruit density under normal conditions in the water-saturated citrus selection. The specific content is as follows:

[0045] The setup requires water-based sorting of Newhall navel oranges. A citrus fruit sample set will be constructed by identifying citrus fruits that are in good condition and ready for market but may be dehydrated, and by assigning numbers to these fruits. The number of samples of this type of citrus used in the drought test. Set to 300, the number of citrus fruits. satisfy and ; by obtaining the Construct a citrus fruit density dataset based on the density of different citrus fruits. Among them, for citrus fruits density It is possible The result is calculated after measurement, with the unit set to grams per cubic centimeter. For the citrus fruit The mass measured by an electronic scale For the citrus fruit The volume measured by the displacement method; by obtaining the... Constructing a spectral dataset of citrus fruits using spectral data from different citrus fruits. Among them, for citrus fruits spectral data Citrus fruits can be observed using a hyperspectral imager The results are obtained after detection and spectral preprocessing; by the... After cutting citrus fruits along the equatorial plane, it is manually determined whether they are dehydrated. Citrus fruits that are determined by the naked eye not to be dehydrated are classified into the normal citrus fruit subset. Citrus fruits that are visually identified as being in a state of water scarcity are categorized into a subset of citrus fruits in a state of water scarcity. and according to the and Construct datasets of citrus fruit density under normal conditions. Data set on citrus fruit density under drought conditions Considering that the fruit density of citrus fruits under normal conditions is greater than that under drought conditions, we define the domain of discourse for citrus fruit density as being expressible through fuzzy mathematics. U = [min( C 2 , 2 ),max( C 2 , 1 )] Through fuzzy subsets This is used to indicate "citrus fruit density is in a normal state" to distinguish it from "citrus fruit density is in a low-water state"; Table 1 shows one setting of some parameters in the membership function for citrus fruit density being in a normal state in this embodiment;

[0046] Table 1. Parameter settings in the membership function of citrus fruits with respect to normal density in this embodiment (unit: g / cm³).

[0047] parameters <![CDATA[min(Γ 2,1 )]]> <![CDATA[E(Γ 2,1 )]]> <![CDATA[D(Γ 2,1 )]]> <![CDATA[E(Γ 2,2 )]]> <![CDATA[D(Γ 2,2 )]]> numerical value 0.795 0.853 0.019 0.703 0.027

[0048] In the membership function The parameters in Table 1 are selected, and the boundary membership values ​​for citrus fruits under normal conditions are set. After setting it to 0.9, it can be seen that the aforementioned membership function power index If the value is 0.261, then the aforementioned value can be set. membership function for:

[0049]

[0050] in, As the independent variable;

[0051] Step 2: Set the membership functions of citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio for the secondary screening of citrus fruit with dryness. The specific content is as follows:

[0052] The and After juicing individual citrus fruits cut in half, titratable acid and sugar-acid ratio were measured to construct a dataset of titratable acid from citrus fruits under normal conditions. Data set of titratable acidity data for citrus fruits under drought conditions Data set of sugar-acid ratio of citrus fruits under normal conditions And data set on the sugar-acid ratio of citrus fruits under drought conditions. ,in, For the citrus fruit The titratable acid value was obtained through quality analysis. This is to determine the sugar-acid ratio of citrus fruits through quality analysis; a setting can be used to... The citrus fruit in the text corresponds to the After filtering for specific long wavelengths from the spectral data, principal component analysis was used to process them separately. and Therefore, predictive models for titratable acidity and sugar-acid ratio of citrus fruits were established, respectively. Using the union notation, the predictive model for titratable acidity in citrus fruits can be represented as follows: The predictive model for the sugar-acid ratio of citrus fruits can be expressed as follows: , Citrus fruit Titratable acid values ​​were obtained using a predictive model of titratable acidity in citrus fruits. Citrus fruit Titratable acid values ​​obtained from a predictive model of the sugar-acid ratio in citrus fruits. and These are the constant terms in the prediction models for titratable acid and sugar-acid ratio of citrus fruit, respectively. This is the set of wavelengths used in the predictive model of titratable acidity for citrus fruits. This is the set of wavelengths used in the predictive model of sugar-acid ratio for citrus fruits. and The respective and Wavelength number in wavelength Weighting coefficients in the predictive model of titratable acidity in citrus fruit. wavelength Weighting coefficients in the prediction model of sugar-acid ratio for citrus fruits. Citrus fruit Regarding wavelength In the The values ​​corresponding to the mid-spectral data, Citrus fruit Regarding wavelength In the The values ​​corresponding to the spectral data;

[0053] Considering that the titratable acid values ​​of citrus fruits under normal conditions are greater than those under dry conditions, we define the domain of discourse regarding the titratable acid values ​​of citrus fruits as being expressible through fuzzy mathematics. V = [min( C 3 , 2 ),max( C 3 , 1 )] Through fuzzy subsets This is used to indicate that "citrus fruit is in a normal state with respect to titratable acid" to distinguish it from "citrus fruit is in a state of dehydration with respect to titratable acid". In this embodiment, some parameters in the membership function of citrus fruit being in a normal state with respect to titratable acid can be set as shown in Table 2.

[0054] Table 2. Parameter settings of the membership function of citrus fruit with respect to titratable acidity under normal conditions in this embodiment (unit: %)

[0055] parameters <![CDATA[min(Γ 3,1 )]]> <![CDATA[E(Γ 3,1 )]]> <![CDATA[D(Γ 3,1 )]]> <![CDATA[E(Γ 3,2 )]]> <![CDATA[D(Γ 3,2 )]]> numerical value 0.42 0.50 0.026 0.29 0.030

[0056] In the membership function By selecting the parameter settings in Table 2, it can be seen that... membership function power index If it is 0.313, then the above can be set. membership function for:

[0057]

[0058] in, As the independent variable;

[0059] Considering that the sugar-acid ratio of citrus fruits under normal conditions is lower than that of citrus fruits under dry conditions, we define the domain of discourse regarding the titratable acid of citrus fruits as being expressible through fuzzy mathematics. W = [min( C 4 , 1 ),max( C 4 , 2 )] Through fuzzy subsets This is used to indicate that "the sugar-acid ratio of citrus fruit is in a normal state" to distinguish it from "the sugar-acid ratio of citrus fruit is in a state of dehydration"; Table 3 shows one setting of some parameters in the membership function of the citrus fruit with a sugar-acid ratio in a normal state in this embodiment.

[0060] Table 3. Parameter settings of the membership function of citrus fruit with respect to a normal sugar-acid ratio in this embodiment.

[0061] parameters <![CDATA[max(Γ 4,1 )]]> <![CDATA[E(Γ 4,1 )]]> <![CDATA[D(Γ 4,1 )]]> <![CDATA[E(Γ 4,2 )]]> <![CDATA[D(Γ 4,2 )]]> numerical value 24.32 21.2 1.04 35.7 2.47

[0062] In the membership function Using the parameter settings in Table 3, it can be seen that... membership function power index If it is 0.7038, then set the above... membership function for:

[0063]

[0064] in, As the independent variable;

[0065] Step 3: For citrus fruits requiring moisture-controlled sorting, the initial screening can be conducted using the membership function of the citrus fruit's density relative to normal conditions. If the density falls within the normal range, the fruit is considered normal and can be graded as Grade 1 or above. If the density falls within the dry-water range, the fruit is considered significantly dry and cannot be graded. All other cases proceed to the second screening in Step 4, which details:

[0066] During the citrus sorting process, in the dry season detection and sorting stage, the initial screening can be conducted using a membership function of the citrus fruit's density relative to a normal state. The membership ranges for citrus fruit with respect to normal density and citrus fruit with respect to dry season density can be set as follows: [ t 1 , 1 , 1 ] and [ 0 , t 1 , 2 ] It allows setting a lower limit for membership when the density is under normal conditions. The upper limit of the membership of density under dry conditions The values ​​were 0.92 and 0.82 respectively; for citrus fruits that need to be sorted by moisture content. ,in Let be the numbering of citrus fruits during dry season sorting, and let be the numbering of citrus fruits obtained. density After the value is 0.793 (unit: grams per cubic centimeter), among which, The citrus sorting equipment uses an online weighing method to measure the weight of citrus fruits. The weight, Three-dimensional laser scanning is used to measure citrus fruits in citrus sorting equipment. The volume of the citrus fruit can determine its size. Regarding the membership function value when the density is under normal conditions It is 0.883, which shows that Then proceed to the second screening;

[0067] Step 4: For citrus fruits requiring moisture separation, after entering the second screening, a joint judgment can be made based on the membership functions of the citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio. If both conditions are met, the fruit is considered to be in normal condition and can be classified as Grade 1 or above. If either condition is within the membership range of moisture separation or sugar-acid ratio separation, the fruit is considered to be in a significantly moisture-deficient condition and can be classified as not eligible for further grade. In other cases, the fruit is considered to be in a slightly moisture-deficient condition and can be classified as Grade 2. The specific details are as follows:

[0068] During the secondary screening stage of citrus sorting, specifically the water shortage detection stage, a joint judgment can be made using the membership functions of citrus fruit regarding whether its titratable acidity is within a normal range and the membership function of citrus fruit regarding whether its sugar-acid ratio is within a normal range. The membership ranges for citrus fruit regarding whether its titratable acidity is within a normal range and the membership ranges for citrus fruit regarding whether its titratable acidity is within a water shortage range can be set as follows: [ t 2 , 1 , 1 ] and [ 0 , t 2 , 2 ] It is possible to set the membership range for citrus fruits with a normal sugar-acid ratio and the membership range for citrus fruits with a low sugar-acid ratio. [ t 3 , 1 , 1 ] and [ 0 , t 3 , 2 ] The lower limit of membership can be set when the titratable acid is under normal conditions. The sugar-acid ratio is at the lower limit of normal membership. Both are 0.92, and the upper limit of membership for titratable acids in a dehydrated state. The upper limit of membership of sugar-acid ratio in dry conditions All are 0.82;

[0069] For citrus fruits that need to be sorted by dryness During the secondary screening, spectral data after spectral preprocessing can be obtained using a hyperspectral imager on the citrus sorting equipment. In order to obtain citrus fruits Titratable acid values ​​obtained from a predictive model of titratable acidity in citrus fruits It was 0.43 (unit: %), and citrus fruit Sugar-acid ratio values ​​obtained from a predictive model of sugar-acid ratio in citrus fruits The value is 24.7; this indicates that the citrus fruit... Membership function values ​​of titratable acids under normal conditions And the membership function value with respect to the normal sugar-acid ratio. The values ​​are 0.939 and 0.862 respectively; therefore, it can be determined that... m B ˜ ( f ^ h ) ∈ [ t 2 , 1 , 1 ] ,but m C ˜ ( ϕ ^ h ) ∉ [ t 3 , 1 , 1 ] and m C ˜ ( ϕ ^ h ) ∉ [ 0 , t 3 , 2 ] Therefore, it can be determined that the citrus fruit It is in a state of slight drought and can be classified as a second-class product.

[0070] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to a preferred embodiment, it is not intended to limit the present invention. Any simple modifications, equivalent changes and alterations made by those skilled in the art to the above embodiments based on the technical content of the present invention shall fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for separating citrus fruits by moisture content for internal quality testing, characterized in that, Includes the following steps: Step 1: Set the membership function of citrus fruit density under normal conditions for the initial screening of citrus fruit moisture selection; Step 2: Set the membership functions of citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio for the secondary screening of citrus fruit with dry water content. Step 3: For citrus fruits that need to be sorted by moisture content, the initial screening can be made by judging the citrus fruit's density as a normal condition. If the citrus fruit's density is within the normal range, it is judged as normal and can be classified as first-class or above. If the citrus fruit's density is within the dry condition range, it is judged as obviously dry and can be classified as not eligible for grading. Other cases proceed to the second screening in Step 4. Step 4: For citrus fruits that need to be sorted by moisture content, after entering the second screening, a joint judgment can be made by using the membership function of the citrus fruit with respect to normal titratable acidity and the membership function of the citrus fruit with respect to normal sugar-acid ratio. If both the citrus fruit and the citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio are within the membership range, it is judged to be in normal condition and can be classified as Grade 1 or above. If the citrus fruit is in a state of moisture content with respect to titratable acidity or a state of moisture content with respect to sugar-acid ratio, it is judged to be in a state of obvious moisture content and can be classified as not eligible for grading. In other cases, it is judged to be in a state of slight moisture content and can be classified as Grade 2.

2. The method for separating citrus fruits by moisture content based on intrinsic quality testing according to claim 1, characterized in that, The specific details of step 1 are as follows: For a certain type of citrus that requires dry sorting, the selection... A citrus fruit sample set will be constructed by identifying citrus fruits that are in good condition and ready for market but may be dehydrated, and by assigning numbers to these fruits. ,in, The number of samples of this type of citrus used in the drought test, and the citrus fruit number. satisfy and ; by obtaining the Construct a citrus fruit density dataset based on the density of different citrus fruits. Among them, for citrus fruits density It is possible The result is given after measurement and calculation. For the citrus fruit The mass measured by an electronic scale For the citrus fruit The volume measured by the displacement method; by obtaining the... Constructing a spectral dataset of citrus fruits using spectral data from different citrus fruits. Among them, for citrus fruits spectral data Citrus fruits can be observed using a hyperspectral imager The results are obtained after detection and spectral preprocessing; by the... After cutting citrus fruits along the equatorial plane, it is manually determined whether they show signs of dehydration. Citrus fruits that are visually confirmed to be free of dehydration are categorized as the "normal" subset. Citrus fruits that are visually identified as being in a state of water shortage are categorized into the subset of citrus fruits in a state of water shortage. and according to the and Construct separate datasets of citrus fruit density under normal conditions. Data set on citrus fruit density under drought conditions We define the domain of discourse for citrus fruit density as being expressible through fuzzy mathematics. ,in, To find the minimum value of a set, To find the function that maximizes a set, we use fuzzy subsets. To indicate "citrus fruit density is at a normal level" to distinguish it from "citrus fruit density is in a low-water condition," the following settings are configured. membership function for: ； in, As the independent variable, To find the function of the average of a set, To find a function of the population standard deviation of a set, For the The power exponent can be used to set the boundary membership value of citrus fruits under normal conditions. Then, the aforementioned Determined as .

3. The method for separating citrus fruits by moisture content based on intrinsic quality testing according to claim 1, characterized in that, The specific details of step 2 are as follows: The and After juicing individual citrus fruits cut in half, titratable acid and sugar-acid ratio were measured to construct a dataset of titratable acid from citrus fruits under normal conditions. Data set of titratable acidity data for citrus fruits under drought conditions Data set of sugar-acid ratio of citrus fruits under normal conditions And data set on the sugar-acid ratio of citrus fruits under drought conditions. ,in, For the citrus fruit Titratable acid values ​​obtained through quality analysis. This is to determine the sugar-acid ratio of citrus fruits through quality analysis; a setting can be used to... The citrus fruit in the text corresponds to the After filtering for specific long wavelengths from the spectral data, principal component analysis was used to process them separately. and Therefore, predictive models for titratable acidity and sugar-acid ratio of citrus fruits were established, respectively. Using the union notation, the predictive model for titratable acidity in citrus fruits can be represented as follows: The predictive model for the sugar-acid ratio of citrus fruits can be expressed as follows: , Citrus fruit Titratable acid values ​​were obtained using a predictive model of titratable acidity in citrus fruits. Citrus fruit Titratable acid values ​​obtained from a predictive model of the sugar-acid ratio in citrus fruits. and These are the constant terms in the prediction models for titratable acid and sugar-acid ratio of citrus fruit, respectively. This is the set of wavelengths used in the predictive model of titratable acidity for citrus fruits. This is the set of wavelengths used in the predictive model of sugar-acid ratio for citrus fruits. and The respective and Wavelength number in wavelength Weighting coefficients in the predictive model of titratable acidity in citrus fruit. wavelength Weighting coefficients in the prediction model of sugar-acid ratio for citrus fruits. Citrus fruit Regarding wavelength In the The values ​​corresponding to the mid-spectral data, Citrus fruit Regarding wavelength In the The values ​​corresponding to the spectral data; We can define the domain of discourse concerning titratable acids of citrus using fuzzy mathematics. Through fuzzy subsets To indicate that "citrus fruit is in a normal condition regarding titratable acidity" to distinguish it from "citrus fruit is in a low-water condition regarding titratable acidity," the following settings are configured. membership function for: ； in, As the independent variable, For the The power exponent can be used to describe Determined as ; We can define the domain of discourse concerning titratable acids of citrus using fuzzy mathematics. Through fuzzy subsets To indicate that "the sugar-acid ratio of citrus fruit is in a normal state" to distinguish it from "the sugar-acid ratio of citrus fruit is in a low-water state," the following settings are configured. membership function for: ； in, As the independent variable, For the The power exponent can be used to describe Determined as .

4. The method for separating citrus fruits by moisture content based on intrinsic quality testing according to claim 1, characterized in that, The specific details of step 3 are as follows: During the citrus sorting process, in the dehydration detection and sorting stage, a joint judgment can be made using the membership functions of citrus fruit with respect to normal titratable acidity and normal sugar-acid ratio. The membership ranges for citrus fruit with respect to normal density and citrus fruit with respect to dehydration density can be set as follows: and ,in, and These are the lower limit of membership for fruits with normal density and the upper limit of membership for fruits with low water content; for citrus fruits that need to be sorted by water content. ,in This refers to the numbering of citrus fruits during the dry season sorting process. (The last sentence appears to be incomplete and possibly refers to a separate context.) density for Afterwards, among them, The citrus sorting equipment uses an online weighing method to measure the weight of citrus fruits. The weight, Three-dimensional laser scanning is used to measure citrus fruits in citrus sorting equipment. The volume of the citrus fruit can determine its size. Regarding the membership function value when the density is under normal conditions for: ； If the setting is Then determine the citrus fruit This is the normal condition and can be classified as first-class or above. Then determine the citrus fruit If the water level is clearly low and the level can be classified as not eligible for further classification, then... Then proceed to the secondary screening in step 4.

5. A method for separating citrus fruits by moisture content based on intrinsic quality testing, as described in claim 1, characterized in that, The specific details of step 4 are as follows: During the secondary screening stage of citrus sorting, specifically the water shortage detection stage, a joint judgment can be made using the membership functions of citrus fruit regarding whether its titratable acidity is within a normal range and the membership functions of citrus fruit regarding whether its sugar-acid ratio is within a normal range. The membership ranges for citrus fruit regarding whether its titratable acidity is within a normal range and the membership ranges for citrus fruit regarding whether its titratable acidity is within a water shortage range can be set as follows: and It is possible to set the membership range for citrus fruits with a normal sugar-acid ratio and the membership range for citrus fruits with a low sugar-acid ratio. and ,in, and These are the lower limit of membership for titratable acid under normal conditions and the upper limit of membership for titratable acid under low water conditions, respectively. and These represent the lower limit of membership for a sugar-acid ratio under normal conditions and the upper limit of membership for a sugar-acid ratio under low water conditions, respectively. For citrus fruits that need to be sorted by dryness During the secondary screening, spectral data after spectral preprocessing can be obtained using a hyperspectral imager on the citrus sorting equipment. This can identify citrus fruits. Membership function values ​​of titratable acids under normal conditions And the membership function value with respect to the normal sugar-acid ratio. They are respectively: ； ； in, Citrus fruit Titratable acid values ​​obtained from a predictive model of titratable acidity in citrus fruits can be obtained through [further details needed]. Give, Citrus fruit Regarding wavelength In the The corresponding value in Citrus fruit The sugar-acid ratio values ​​obtained from the prediction model of sugar-acid ratio in citrus fruits can be specifically obtained through... Give, Citrus fruit Regarding wavelength In the The corresponding value in; If the setting is and Then determine the citrus fruit This is the normal condition and can be classified as first-class or above. or Then determine the citrus fruit If the water level is clearly low, the grade can be classified as "not eligible for grading"; otherwise, the citrus fruit will be judged as "unsuitable for grading". It is in a state of slight drought and can be classified as a second-class product.

Images