CN107290299B - Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode - Google Patents
Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode Download PDFInfo
- Publication number
- CN107290299B CN107290299B CN201710271814.2A CN201710271814A CN107290299B CN 107290299 B CN107290299 B CN 107290299B CN 201710271814 A CN201710271814 A CN 201710271814A CN 107290299 B CN107290299 B CN 107290299B
- Authority
- CN
- China
- Prior art keywords
- acidity
- peach
- peaches
- sugar degree
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 235000006040 Prunus persica var persica Nutrition 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 38
- 240000006413 Prunus persica var. persica Species 0.000 title 1
- 244000144730 Amygdalus persica Species 0.000 claims abstract description 44
- 240000005809 Prunus persica Species 0.000 claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 25
- 238000013528 artificial neural network Methods 0.000 claims abstract description 24
- 230000003595 spectral effect Effects 0.000 claims abstract description 15
- 238000000513 principal component analysis Methods 0.000 claims abstract description 11
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 230000002068 genetic effect Effects 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims abstract description 5
- 238000001514 detection method Methods 0.000 claims description 20
- 238000001228 spectrum Methods 0.000 claims description 12
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 9
- 239000012153 distilled water Substances 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012847 principal component analysis method Methods 0.000 claims description 4
- 238000007635 classification algorithm Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 235000011389 fruit/vegetable juice Nutrition 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 230000007935 neutral effect Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000009499 grossing Methods 0.000 claims description 2
- 230000035772 mutation Effects 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000005406 washing Methods 0.000 claims description 2
- 230000003044 adaptive effect Effects 0.000 claims 2
- 238000002329 infrared spectrum Methods 0.000 abstract description 5
- 239000003153 chemical reaction reagent Substances 0.000 abstract description 2
- 238000005457 optimization Methods 0.000 abstract 1
- 235000013399 edible fruits Nutrition 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 6
- 239000013598 vector Substances 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 235000013305 food Nutrition 0.000 description 4
- 230000006378 damage Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 230000001066 destructive effect Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241000238631 Hexapoda Species 0.000 description 1
- 238000004497 NIR spectroscopy Methods 0.000 description 1
- 235000006029 Prunus persica var nucipersica Nutrition 0.000 description 1
- 244000017714 Prunus persica var. nucipersica Species 0.000 description 1
- 235000004789 Rosa xanthina Nutrition 0.000 description 1
- 241000220222 Rosaceae Species 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009658 destructive testing Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000009987 spinning Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Theoretical Computer Science (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a method for detecting the sugar degree and the acidity of peaches in a real-time and nondestructive manner. The method comprises the steps of collecting a near infrared spectrum of a peach sample, and respectively detecting the sugar degree and the acidity of the peach through a handheld refractometer and a pen-type pH meter; and then, performing dimensionality reduction and optimization on the spectral data of the peaches through principal component analysis, an isometric mapping algorithm and a genetic algorithm, finally classifying through a neural network method, and determining an optimal prediction model after multiple verifications. The prediction model based on the neural network established by the invention can well predict the sugar degree and the acidity of peaches. Compared with the traditional method for detecting the sugar acidity of the peach by damaging the sample, the method can simultaneously detect two indexes of the sugar acidity in real time, and has the advantages of high speed, low cost, no destructiveness, no need of chemical reagents, no pollution and the like.
Description
Technical Field
The invention belongs to the technical field of food detection, and particularly relates to a method for detecting the sugar degree and acidity of peaches in a real-time and nondestructive manner.
Background
China is a large world-wide fruit production country, the fruit industry plays an important role in national economy, and fruit export is an important component of foreign trade in China. The peach native China belongs to small deciduous trees of Rosaceae, the cultivation history has more than 4000 years so far, and the peach yield and the peach consumption of China are at the top of the world. However, agricultural products in China are low in commercialization, and are often in a disadvantage in international competition. Therefore, the detection and classification technology of agricultural products becomes an important research direction in the scientific research field of China.
The traditional method for detecting the internal quality of the fruit mainly determines the sugar content, acidity and sugar-acid ratio of the fruit by a chemical analysis mode to judge the quality of the fruit. For example, the sugar content of peaches can be measured by an Abbe refractometer according to the food hygiene monitoring rules of chemistry (GB/T5009.1-2003). The acidity of peach can be determined by pH meter according to food hygiene monitoring method (GB/T5009.1-2003) and determination method of acidity in food (GB/T12456-90). However, these methods all require damage to the peaches, and belong to destructive testing; moreover, they are tedious in sample preparation and long in detection time, and cannot meet the actual requirements of rapid grading and sorting of fruits.
The near infrared spectrum can quickly reflect the structure and composition information of a detected object, and is suitable for nondestructive detection of internal components of fruits and other agricultural products. Near infrared spectrum detection is a non-destructive detection technology which has no pretreatment, high detection speed, wide detection range and high accuracy and is concerned by people. In recent years, near-infrared detection technology is rapidly developing, and the technology tends to be miniaturized and convenient on the basis of keeping the original advantages. Near infrared spectroscopy has been used primarily to detect the brix of a small number of species of peaches, but the source of the subject samples studied is limited. For peaches of different producing areas on the market, the prediction of the sugar degree and acidity still needs to establish a more effective method.
Disclosure of Invention
The invention aims to provide a method for detecting the sugar degree and the acidity of peaches in a real-time and nondestructive manner so as to solve the defects caused by the prior art.
In order to solve the technical problems, the invention provides the following technical scheme:
a method for rapidly and nondestructively detecting the sugar degree and the acidity of peaches comprises the following steps:
1) performing spectrum collection on a peach sample by using a near-infrared spectrometer to obtain an original spectrum, wherein the wavelength range of the spectrometer is 500-1900 nm;
2) measuring the sugar degree and the acidity of the peach sample to be used as an observed value of an analysis and prediction model;
3) smoothing and filtering the original spectrum by adopting a data moving average filtering method;
4) constructing a prediction model of the peach sugar degree, and firstly performing dimensionality reduction on collected peach spectral data by a mixed principal component analysis method, an arithmetic mean mapping method and a genetic algorithm; carrying out BP neural network classification on the data subjected to dimensionality reduction, including training the network by using training set data and testing the network prediction capability by using a test set, and carrying out multiple verification to finally determine a prediction model with the optimal peach sugar degree;
5) constructing a prediction model of peach acidity, performing dimensionality reduction on collected peach spectral data by a principal component analysis and equidistant mapping method, performing BP neural network classification on the dimensionality reduced data, performing verification for multiple times, and finally determining the prediction model with the optimal peach acidity;
6) collecting spectral information of the peach to be tested, preprocessing the spectral information according to the step 3), and substituting the preprocessed spectral data into the prediction model established in the step 4)5) to obtain predicted values of the sugar degree and the acidity of the sample to be tested.
As a preferred technical scheme of the invention, the number of the samples in the step 1) is not less than 50, preferably 50-300.
As a preferred technical scheme of the invention, in the step 1), the near infrared spectrometer randomly selects 5-20 points on the ring equatorial plane of each peach to perform spectral scanning by adopting a diffuse reflection mode, and preferably 15 points are selected.
As a preferred technical scheme of the invention, the method for measuring the sugar degree of peaches in the step 2) comprises the following steps: firstly, the prism of the brix meter is cleaned by distilled water and wiped to remove water, the correction and zero setting are carried out, then the peach with the measured spectrum is cut by a clean knife and juice is squeezed out on the mirror surface of the dioptric prism, the measurement is carried out for three times continuously, and the record and the average value are obtained.
As a preferred technical scheme of the invention, the method for measuring the acidity of peaches in the step 2) comprises the following steps: washing the front end of the hand-held pH meter by distilled water until the pH is neutral; the moisture on the pH meter was spun off, inserted onto the equatorial plane of the peach ring, the pH was measured at 6 positions, the data were recorded and averaged.
As a preferred technical scheme of the invention, the original 2047-dimensional data is reduced to the lowest dimension by the three methods of the mixed principal component analysis method, the arithmetic mean mapping and the genetic algorithm in the step 4) 5).
As a preferred technical scheme of the invention, the peach classification algorithm based on the BP network in the step 4)5) comprises three steps of constructing the BP neural network, training the BP neural network and classifying the BP neural network.
As a preferred embodiment of the present invention, the genetic algorithm described in step 4) generates a group of more environment-adaptive individuals by random selection, crossover and mutation operations starting from any initial population, and finally converges to a group of most environment-adaptive individuals.
As a preferred technical scheme of the invention, the error percentage in the step 6) within 10 percent is determined as accurate prediction. Has the advantages that: compared with the prior art, the method can perform real-time, rapid, accurate and nondestructive detection on the sugar degree and the acidity of the peach through near infrared spectrum analysis. The traditional method needs to destroy peaches, is destructive detection, needs to detect the sugar degree and the acidity respectively, and does not have a proper method to detect the variety of peaches. The method has the advantages of no destruction, high speed, low cost, no need of pretreatment, no need of chemical reagents, no pollution and the like, and can predict the sugar degree and the acidity of the peach.
Drawings
FIG. 1 is a basic calculation flow of genetic algorithm;
FIG. 2 is a BP neural network algorithm flow;
FIG. 3 shows the PCA-ISOMAP-GA-BP method for predicting the sugar degree and acidity of peaches: (A) the actual (triangles) and predicted (squares) brix values for the test set were tested. (B) Percentage error of brix prediction. (C) The actual (triangle) and predicted (square) acidity values of the test set were tested. (D) Percentage error of acidity prediction.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
Example 1
Selecting two kinds of peaches without damage to surfaces and diseases and insect pests: the peach and nectarine are respectively used as 30 as materials, and each peach is numbered, and the total number is 60.
The near infrared spectrum data are collected in a diffuse reflection mode, 20 points are selected on the surfaces which are near the equatorial plane of each peach ring and avoid the edge angles of the side surfaces and are 90 degrees to each other for spectrum scanning, and the collection spectrum area is 340-1021 nm and is 2047 points in total. The spectral data collected at the sampling point of each sample is used for subsequent data analysis.
The brix of each peach sample was measured using a hand-held digital refractometer. Firstly, the prism of the brix meter is cleaned by distilled water and wiped to remove water, the correction and zero setting are carried out, then the peach with the measured spectrum is cut by a clean knife and juice is squeezed out on the mirror surface of the dioptric prism, the measurement is carried out twice continuously, and the record and the average value are obtained.
The acidity of each peach sample was measured using a conical pen pH meter. The hand-held pH meter front end was rinsed with distilled water until pH was shown to be neutral. And (3) spinning water on a pH meter, inserting the water on the equatorial plane of the peach ring, taking the pH values at 6 points, recording data and averaging the data.
Due to the fact that the spectral data dimension of the peaches is large, the predictive variables are mutually related, multiple co-linearity can cause instability of a solution space, and therefore results can be discontinuous. Therefore, we first perform dimension reduction on the collected spectral data of peaches. We chose the PCA (principal component analysis) method for dimensionality reduction. The method not only reduces the dimension of high-dimensional data, but also removes noise through dimension reduction and discovers the mode in the data. The PCA method "combines" the basic elements of the attributes by creating an alternative, smaller set of variables, allowing the raw data to be projected into the smaller set. And (4) according to the principal component analysis result, the original 2047-dimensional data is reduced to the lowest dimension.
The basic process for PCA is as follows:
(a) the input data is normalized so that each attribute falls within the same interval. This step helps to ensure that attributes with a larger domain of definition do not dominate attributes with a smaller domain of definition.
(b) PCA calculates k orthonormal vectors as the basis for normalizing the input data. These are unit vectors, each perpendicular to the other vectors. These vectors are called principal components. The input data is a linear combination of principal components.
(c) The principal components are arranged in descending order of "importance" or intensity. The principal component essentially acts as a new coordinate system for the data, providing important information about the variance.
(d) Because the principal components are sorted in descending order according to "importance", the data can be reduced by removing the weaker components (i.e., those with smaller variances). Using the strongest principal component, a good approximation of the metadata can be reconstructed.
And optimizing the data from the protocol to the ten-dimensional data after the principal component analysis by using an isometric mapping (ISOMAP) algorithm so as to achieve the aim of further improving the model prediction accuracy.
The isometric mapping (ISOMAP) algorithm is a nonlinear flow type learning algorithm with global characteristics based on a domain graph and a classical MDS algorithm. The algorithm is an unsupervised algorithm, because the multidimensional scaling algorithm is used for dimension reduction, the ISOMAP algorithm can well reserve the global characteristic of data, and the overall structure of the data obtained after dimension reduction is closer to the original data. The IOSMAP algorithm has the main idea that the global manifold geodesic distance between data points is estimated by using the local domain distance, and the data dimensionality reduction is realized by establishing the peer-to-peer relationship between the geodesic distance between original data and the spatial distance between dimensionality reduced data.
The algorithm flow is described as follows:
(1) a neighbor graph G is constructed. Calculating point xiAnd point xjEuropean style of ChineseDistance D (x)i,xj) If point xjFalls at point xiAs the centre of a circle, as a circle of radius, or point xjIs a point xiOne of the K neighbor points of (1), then x isiAnd xjEdges are connected and Euclidean distances are used as the weight of the edges. Otherwise, not connect, and the weight is recorded as infinity.
(2) The shortest path is calculated according to the Floyd algorithm. For xiAnd xjIf there is an edge between them, initialize its shortest path D (x)i，xj)＝dG(xi，xj). For m ═ 1,2, …, n, calculations
dG(xi，xj)＝{dG(xi，xj)，dG(xi，xm)+dG(xm，xj)} (1)
The shortest path distance matrix D can be obtained from the formula (1)G＝{dG(xi，xj)}。
(3) And solving the dimensionality reduction embedding. Establishing shortest path distance matrix D by using classical MDS algorithmGDistance matrix D between the data and the dimension reduction dataY＝{dG(yi，yj)＝||yi-yjThe peer-to-peer relationship of | | is obtained by minimizing the cost function, so as to obtain the global low-dimensional coordinates:
in the formula: matrix operation operator tau (D) — HSH/2, squared distance matrix
Centralized matrix
WhereinijIs Kronecker delta operator.
We assume τ (D)G) Maximum d eigenvalues λ of1，λ2，...，λdThe corresponding feature vector is u1，u2，…udWhen Y is diag { lambda ═ d1 1/2，λ2 1/2，...，λd 1/2}[u1，u2，…ud]TEquation (2) reaches a global minimum.
And optimizing and selecting variables meeting the conditions by using an isometric mapping algorithm as input, and classifying the BP neural network. The BP neural network is a multilayer feedforward neural network and is mainly characterized by signal forward transmission and error backward propagation. In forward pass, the input signal is processed layer by layer from the input layer through the hidden layer to the output layer. The neuronal state of each layer only affects the neuronal state of the next layer. If the expected output cannot be obtained by the output layer, the backward propagation is carried out, and the network weight and the threshold are adjusted according to the prediction error, so that the predicted output of the BP neural network continuously approaches the expected output. The BP neural network can be considered as a non-linear function, and the network input value and the predicted value are respectively an independent variable and a dependent variable of the function. When the number of input nodes is n and the number of output nodes is m, the BP neural network expresses the function mapping relation from n independent variables to m dependent variables. Before the BP neural network prediction, firstly, a network is trained, and the network has associative memory and prediction capabilities through training. The peach classification algorithm based on the BP network comprises three steps of BP neural network construction, BP neural network training and BP neural network classification, and the algorithm flow is shown in figure 2.
And constructing a PCA-ISOMAP-GA-BP model for predicting the sugar degree and the acidity of peaches. We first perform dimensionality reduction on the collected mean spectral data of the peaches. High dimensional data is projected into a lower dimensional space, i.e. the overall properties of the object are represented by several main variables. Processing and optimizing an original data set by mixing three methods of PCA, ISOMAP and GA, and reducing the data dimension to 8 dimensions in the sugar degree prediction; in acidity prediction, the data dimension is reduced to 6 dimensions. And respectively establishing a sugar degree and acidity prediction model for the processed data through a BP network model. And 5 groups of data are selected from each variety data to be used as a test set to test the classification capability of the network, and other data are used as a training set to train the network.
As shown in FIG. 3, we analyzed the predicted effect of the PCA-ISOMAP-GA-BP model on the brix and acidity. In fig. 3, (a) the triangle is the actual brix value, and the square is the predicted brix value. (B) Is the percentage error of brix prediction. (C) The actual (triangle) and predicted (square) acidity values of the test set were tested. (D) Percentage error of acidity prediction. In the sugar degree prediction model, the predicted value is basically the same as the actual value, the error percentage is small, and the average absolute percentage error is 7.55%. In acidity prediction, most predicted values are uniformly distributed around the actual value, and the average absolute percentage error is 4.79%. The original 2047-dimensional data are respectively reduced to 8-6 dimensions by the hybrid algorithm, the hybrid algorithm is the minimum dimension of all models, and the characteristics that the hybrid model still has the original data structure in a smaller dimension and the prediction effect is good are fully reflected. The PCA-ISOMAP-GA-BP model established by the method can well predict the sugar degree and the acidity of peaches.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (9)
1. A real-time nondestructive detection method for sugar degree and acidity of peaches is characterized by comprising the following steps: the method comprises the following specific steps:
firstly, carrying out spectrum collection on a peach sample through a near-infrared spectrometer to obtain an original spectrum, wherein the wavelength range of the near-infrared spectrometer to 1900nm is 500-;
secondly, measuring the sugar degree and the acidity of the peach sample to be used as an observed value of an analysis and prediction model;
thirdly, smoothing and filtering the original spectrum by adopting a data moving average filtering method;
fourthly, constructing a prediction model of the peach sugar degree, and firstly performing dimensionality reduction on the collected peach spectral data by a mixed principal component analysis method, an arithmetic mean square mapping method and a genetic algorithm; carrying out BP neural network classification on the data subjected to dimensionality reduction, including training the network by using training set data and testing the network prediction capability by using a test set, and carrying out multiple verification to finally determine a prediction model with the optimal peach sugar degree;
fifthly, constructing a prediction model of the acidity of the peach, performing dimensionality reduction on the collected peach spectral data through a principal component analysis and isometric mapping method, performing BP neural network classification on the dimensionality-reduced data, performing verification for multiple times, and finally determining the prediction model with the optimal acidity of the peach;
and sixthly, collecting the spectral information of the peach to be tested, preprocessing the peach according to the third step, and substituting the preprocessed spectral data into the prediction models established in the fourth step and the fifth step to obtain the predicted values of the sugar degree and the acidity of the sample to be tested.
2. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: the number of samples in the first step is not less than 50.
3. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: in the first step, the near infrared spectrometer adopts a diffuse reflection mode, and 5-20 points on the annular equatorial plane of each peach are randomly selected for spectrum scanning.
4. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: the method for measuring the sugar degree of peach in the second step comprises the following steps: firstly, the prism of the brix meter is cleaned by distilled water and wiped to remove water, the correction and zero setting are carried out, then the peach with the measured spectrum is cut by a clean knife and juice is squeezed out on the mirror surface of the dioptric prism, the measurement is carried out for three times continuously, and the record and the average value are obtained.
5. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: the measuring method of the acidity of peaches in the second step comprises the following steps: washing the front end of the hand-held pH meter by distilled water until the pH is neutral; the moisture on the pH meter was spun off, inserted onto the equatorial plane of the peach ring, the pH was measured at 6 positions, the data were recorded and averaged.
6. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: and in the fourth step, the original 2047-dimensional data is reduced to the lowest dimension by a mixed principal component analysis method, an arithmetic of arithmetic.
7. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: and the BP neural network classification algorithm in the fourth step and the fifth step comprises the three steps of BP neural network construction, BP neural network training and BP neural network classification.
8. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: the genetic algorithm in the fourth step starts from any initial population, generates a group of individuals more adaptive to the environment through random selection, crossover and mutation operations, and finally converges to a group of individuals most adaptive to the environment.
9. The real-time nondestructive detection method for sugar degree and acidity of peaches as claimed in claim 1, wherein: and in the sixth step, the error percentage within 10 percent is determined as accurate prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710271814.2A CN107290299B (en) | 2017-04-24 | 2017-04-24 | Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710271814.2A CN107290299B (en) | 2017-04-24 | 2017-04-24 | Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107290299A CN107290299A (en) | 2017-10-24 |
| CN107290299B true CN107290299B (en) | 2020-08-11 |
Family
ID=60094902
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710271814.2A Active CN107290299B (en) | 2017-04-24 | 2017-04-24 | Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107290299B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108593596A (en) * | 2018-05-31 | 2018-09-28 | 福建医科大学 | The method that Normal juice content in coconut juice is quickly detected based on near-infrared spectrum technique |
| CN109540836A (en) * | 2018-11-30 | 2019-03-29 | 济南大学 | Near infrared spectrum pol detection method and system based on BP artificial neural network |
| CN111474137A (en) * | 2020-04-27 | 2020-07-31 | 无锡雪浪数制科技有限公司 | Citrus sugar degree nondestructive testing method based on convolutional neural network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101638402A (en) * | 2008-07-30 | 2010-02-03 | 天津天士力现代中药资源有限公司 | Online quality monitoring method for salvianolic acid B production |
| CN102830087A (en) * | 2011-09-26 | 2012-12-19 | 武汉工业学院 | Method for quickly identifying food waste oils based on near infrared spectroscopy |
| CN105527244A (en) * | 2015-10-26 | 2016-04-27 | 沈阳农业大学 | Near infrared spectrum-based Hanfu apple quality nondestructive test method |
| CN106769970A (en) * | 2016-12-07 | 2017-05-31 | 甘肃萃英大农科技有限公司 | A kind of method of the kind, pol and acidity of Fast nondestructive evaluation apple |
-
2017
- 2017-04-24 CN CN201710271814.2A patent/CN107290299B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101638402A (en) * | 2008-07-30 | 2010-02-03 | 天津天士力现代中药资源有限公司 | Online quality monitoring method for salvianolic acid B production |
| CN102830087A (en) * | 2011-09-26 | 2012-12-19 | 武汉工业学院 | Method for quickly identifying food waste oils based on near infrared spectroscopy |
| CN105527244A (en) * | 2015-10-26 | 2016-04-27 | 沈阳农业大学 | Near infrared spectrum-based Hanfu apple quality nondestructive test method |
| CN106769970A (en) * | 2016-12-07 | 2017-05-31 | 甘肃萃英大农科技有限公司 | A kind of method of the kind, pol and acidity of Fast nondestructive evaluation apple |
Non-Patent Citations (5)
| Title |
|---|
| 基于高光谱图像技术的农产品品质无损检测;田有文等;《农机化研究》;20091031(第10期);220-223 * |
| 近红外光谱数据建模方法的研究进展;王凤花等;《农业工程》;20110531;第1卷(第1期);56-61 * |
| 高光谱图像技术在水果无损检测中的研究进展;吴龙国等;《激光与红外》;20130930;第43卷(第9期);990-996 * |
| 高光谱成像在农产品无损检测中的应用;欧阳爱国等;《广东农业科学》;20151231(第23期);164-171 * |
| 黄酒糖度近红外光谱降维算法的研究;蒋巧勇等;《中国计量学院学报》;20150331;第26卷(第1期);50-54 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107290299A (en) | 2017-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106769970A (en) | A kind of method of the kind, pol and acidity of Fast nondestructive evaluation apple | |
| CN108875913B (en) | Tricholoma matsutake rapid nondestructive testing system and method based on convolutional neural network | |
| WO2018010352A1 (en) | Qualitative and quantitative combined method for constructing near infrared quantitative model | |
| CN107290299B (en) | Method for detecting sugar degree and acidity of peaches in real time in nondestructive mode | |
| CN107202761A (en) | The portable detection equipment and detection method of a kind of quick detection fruit internal quality | |
| Gill et al. | A review of automatic fruit classification using soft computing techniques | |
| CN110826552A (en) | Grape nondestructive automatic detection device and method based on deep learning | |
| WO2023108514A1 (en) | Spectral peak automatic detection and reconstruction method and system in tea near-infrared spectral analysis | |
| CN107976417B (en) | Crude oil type identification method based on infrared spectrum | |
| CN103743705A (en) | Rapid detection method for sorghum halepense and similar species | |
| Wang et al. | Grading detection of “Red Fuji” apple in Luochuan based on machine vision and near-infrared spectroscopy | |
| CN114414521A (en) | Milk main component measuring method based on infrared multispectral sensor | |
| CN113408616A (en) | Spectrum classification method based on PCA-UVE-ELM | |
| CN112069911A (en) | Fruit and vegetable quality detection method based on multispectral image information and TLMD-WOA-SIFT | |
| Tripathi et al. | Identification of mango variety using near infrared spectroscopy | |
| Gomes et al. | Towards robust Machine Learning models for grape ripeness assessment | |
| Shen et al. | Nondestructive detection for forecasting the level of acidity and sweetness of apple based on NIR spectroscopy | |
| Wei et al. | Polysaccharide determination and habitat classification for fresh Dendrobiums with hyperspectral imagery and modified RBFNN | |
| Lv et al. | Identification of less-ripen, ripen, and over-ripen grapes during harvest time based on visible and near-infrared (Vis-NIR) spectroscopy | |
| Shen et al. | Feasibility of non-destructive internal quality analysis of pears by using near-infrared diffuse reflectance spectroscopy | |
| Manessa et al. | Study of rubber leaf disease using hyperspectral reflectance | |
| Lu et al. | Rapid detecting soluble solid content of pears based on near-infrared spectroscopy | |
| Lin et al. | Nondestructive detection of mango soluble solid content in hyperspectral imaging based on multi-combinatorial feature wavelength selection | |
| Lyu et al. | In-situ and non-destructive grape quality discrimination via field spectroradiometer and machine learning models | |
| Chen et al. | Hyperspectral Imaging and Convolutional Learning for Sugariness Prediction: Example of Syzygium Samarangense |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |