CN115479902A - Chlorophyll content determination model construction method based on wavelength and system thereof - Google Patents
Chlorophyll content determination model construction method based on wavelength and system thereof Download PDFInfo
- Publication number
- CN115479902A CN115479902A CN202211100290.8A CN202211100290A CN115479902A CN 115479902 A CN115479902 A CN 115479902A CN 202211100290 A CN202211100290 A CN 202211100290A CN 115479902 A CN115479902 A CN 115479902A
- Authority
- CN
- China
- Prior art keywords
- chlorophyll content
- wavelength
- detected
- leaf sample
- model
- 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.)
- Pending
Links
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
-
- 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
- G01N21/55—Specular reflectivity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Pathology (AREA)
- Software Systems (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Immunology (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a method for constructing a chlorophyll content measurement model based on wavelength, which comprises the following steps of: acquiring a spectral data set of a leaf sample and a corresponding chlorophyll content value; preprocessing a spectral data set of a leaf sample to obtain a preprocessed wavelength data set; carrying out variable screening on the preprocessed wavelength data set to obtain a target wavelength data set as a characteristic wavelength data set; constructing a model by using the characteristic wavelength data set, comparing the chlorophyll content predicted value obtained by using the model with the corresponding chlorophyll content value, and optimizing the model to obtain a constructed chlorophyll content measuring model; through establishing the chlorophyll content determination model based on the leaf measurement spectrum and the wavelength thereof, the spectrum or the wavelength of the leaf sample is processed, and the chlorophyll content of the leaf sample is conveniently determined.
Description
Technical Field
The invention relates to the technical field of crop observation, in particular to a method and a system for constructing a chlorophyll content determination model based on wavelength.
Background
Chlorophyll, the major pigment of plants for photosynthesis, is located in thylakoid membranes of plant cells and plays a central role in the light absorption process of photosynthesis. Therefore, the chlorophyll content and the change condition thereof can well reflect the photosynthesis capability, the growth health condition and the like of the plant, and have important significance in the aspects of monitoring the growth vigor of the crops, the plant diseases and insect pests, the crop yield, predicting the maturity period of the crops and the like.
At present, three methods, namely a chemical analysis method, a computer vision method and a chlorophyll meter method, are mainly used for measuring the chlorophyll content of plant leaves. The chemical analysis method is high in accuracy, but the preparation of the chlorophyll solution requires grinding of leaves, is time-consuming and labor-consuming, damages the tissue of the leaves, and is only suitable for laboratory measurement. The computer vision method can be applied to a vehicle-mounted farmland diagnosis system, is a rapid and non-contact measurement method, but the realization of the method needs the participation of an upper computer and the storage of a large amount of databases, and the whole system has large volume and is inconvenient to carry. The chlorophyll meter method is a method for obtaining the chlorophyll content of plant leaves by reading the relevant information of transmitted light or reflected light of the plant leaves. The chlorophyll meter method has the advantages of real-time and quick performance, no damage to leaves and the like. The most representative of the existing portable chlorophyll meters is SPAD502 of Meinenda, japan. The SPAD502 has an LED (light emitting Diode) therein, which emits light with wavelengths of 650nm and 940nm, and chlorophyll has a high absorbance for infrared light of 650nm and a low absorbance for infrared light of 940 nm. The working principle of the SPAD502 is to measure the absorbance of the leaves at 650nm by using 940nm infrared light as a reference. The SPAD502 measures only with light of two wavelength bands, and the obtained absorbance information is limited, so that the measurement accuracy is low.
Disclosure of Invention
According to the method, the chlorophyll content of the leaf sample is conveniently measured by constructing the chlorophyll content measuring model based on the leaf measuring spectrum and the wavelength of the spectrum, and processing the spectrum or the wavelength of the leaf sample, the measuring result is high in accuracy and efficiency, the chlorophyll content in a large range can be effectively measured, time and labor are saved, the rule hidden behind the spectrum data is mined from a deep level, and related scientific problems are solved.
The application discloses a chlorophyll content determination model construction method based on wavelength, including:
acquiring a spectral data set of a leaf sample and a corresponding chlorophyll content value;
preprocessing the spectral data set of the leaf sample to obtain a preprocessed wavelength data set;
carrying out variable screening on the preprocessed wavelength data set to obtain a target wavelength data set as a characteristic wavelength data set; the variable screening process comprises the following steps: constructing a model by adopting the preprocessed wavelength data set to obtain a constructed primary model; reserving points of the primary model with the weight of the regression coefficient absolute value greater than or equal to a first threshold as a new subset, and removing points with the weight less than or equal to the first threshold to obtain a secondary iteration subset after primary variable screening; establishing a second iteration model based on the second iteration subset to obtain a third iteration subset; after N cycles, respectively establishing N iteration models based on the N iteration subsets to respectively obtain N +1 iteration subsets; respectively calculating N groups of root mean square error values corresponding to the N times of iterative models, and determining an iterative subset corresponding to the minimum value of the root mean square error values as a characteristic wavelength data set;
and constructing a model by using the characteristic wavelength data set, comparing the chlorophyll content predicted value obtained by using the model with the corresponding chlorophyll content value, and optimizing the model to obtain the constructed chlorophyll content determination model.
The pretreatment comprises one or more of the following steps: first derivative, second derivative, baseline correction, multivariate scatter correction, variable normalization, reciprocal logarithm.
The pretreatment method comprises the following steps: and (5) standardizing the variables.
The spectral data set of the leaf sample is obtained by smoothing the original spectral data of the leaf sample: optionally, the smoothing method includes, but is not limited to, one of the following: savitzky-Golay.
The characteristic wavelength data set comprises one or more of the following: 510. 558, 559, 713, 1717, 1720, 1898, 2031, 2033, 2304;
performing model construction on the characteristic wavelength data by using a machine learning method to obtain a constructed chlorophyll content measuring model;
optionally, the machine learning method includes one or more of the following: partial least squares regression, least squares support vector machine, neural network, random forest method, linear regression, logistic regression, linear discriminant analysis, classification and regression tree, naive Bayes, KNN, learning vector quantization, support vector machine, lightGBM, extreme gradient boosting;
optionally, the machine learning method includes: partial Least Squares Regression (PLSR).
A method for wavelength-based chlorophyll content determination, comprising:
acquiring wavelength data of a leaf sample to be detected;
inputting the wavelength data of the leaf sample to be detected into the chlorophyll content measuring model for processing to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, the wavelength data of the leaf sample to be detected is obtained by preprocessing the spectral data of the leaf sample to be detected; according to different methods for preprocessing the spectral data of the leaf sample to be detected, inputting the wavelength data of the leaf sample to be detected into a corresponding chlorophyll content determination model to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, when the wavelength data of the leaf sample to be detected is subjected to variable standardization preprocessing, the wavelength data is input into a chlorophyll content determination model one to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to preprocessing of multivariate scattering correction, inputting the wavelength data into a chlorophyll content determination model II to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal preprocessing, inputting the wavelength data into a chlorophyll content determination model III to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to second derivative pretreatment, inputting the wavelength data into a chlorophyll content determination model V to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to baseline correction preprocessing, inputting the wavelength data into a chlorophyll content determination model VI to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal logarithm preprocessing, inputting the wavelength data into a chlorophyll content determination model seven to obtain a chlorophyll content value of the leaf sample to be detected;
and when the wavelength data of the leaf sample to be detected is subjected to the pretreatment of the first derivative, inputting the wavelength data into a chlorophyll content determination model eight to obtain the chlorophyll content value of the leaf sample to be detected.
A wavelength-based chlorophyll-content determining apparatus, the apparatus comprising: a memory and a processor;
the memory is to store program instructions;
the processor is configured to invoke program instructions that, when executed, are configured to perform the wavelength-based chlorophyll content determination method described above.
A wavelength-based chlorophyll content determination system, comprising:
the acquisition unit is used for acquiring wavelength data of a blade sample to be detected;
and the processing unit is used for inputting the wavelength data of the leaf sample to be detected into the chlorophyll content measuring model for processing to obtain the chlorophyll content value of the leaf sample to be detected.
A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the above-mentioned wavelength-based chlorophyll-content determining method.
The application has the following beneficial effects:
1. the application innovatively discloses a method for constructing a chlorophyll content measurement model based on wavelength, which comprises the steps of sequentially preprocessing a spectral data set of a leaf sample and screening variables to obtain a target wavelength data set, and modeling by using the target wavelength data set to obtain a constructed chlorophyll content measurement model; the law hidden behind the spectral data is deeply mined, and the accuracy and the depth of data analysis are greatly improved through deep analysis processing of a plurality of layers such as a preprocessing mode, a variable screening mode and the like; different pretreatment modes are adopted to obtain different chlorophyll content measuring models;
2. the method and the device have the advantages that the constructed chlorophyll content measuring model is innovatively provided for measuring the chlorophyll content in the leaf sample to be measured, different preprocessed wavelength data sets are obtained according to different spectral data preprocessing methods, different chlorophyll content measuring models are built, the method and the device are suitable for processing different varieties of leaf samples, the accuracy of chlorophyll content measuring values of different leaf samples is guaranteed, existing group data are fully utilized, and the method and the device are suitable for various crops.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for constructing a chlorophyll content measurement model based on wavelength according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method for measuring chlorophyll content based on wavelength according to an embodiment of the present invention;
FIG. 3 is a flow chart of the method for measuring chlorophyll content based on wavelength according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a wavelength-based chlorophyll content measuring apparatus provided by an embodiment of the present invention;
FIG. 5 is a schematic flow chart of a wavelength-based chlorophyll content determination system provided by an embodiment of the present invention;
FIG. 6 is a table of results of band screening of target wavelength data provided by an embodiment of the present invention;
fig. 7 is a table of modeling accuracy of a characteristic waveband PLSR of chlorophyll content in a pepper leaf provided by an embodiment of the present invention;
FIG. 8 is a table of modeling accuracy comparison of full-band PLSR of chlorophyll content in pepper leaves according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
In some of the flows described in the present specification and claims and in the above figures, a number of operations are included that occur in a particular order, but it should be clearly understood that these operations may be performed out of order or in parallel as they occur herein, with the order of the operations being indicated as 101, 102, etc. merely to distinguish between the various operations, and the order of the operations by themselves does not represent any order of performance. Additionally, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first", "second", etc. in this document are used for distinguishing different messages, devices, modules, etc., and do not represent a sequential order, nor limit the types of "first" and "second" to be different.
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow chart of a method for constructing a chlorophyll content measurement model based on wavelength according to an embodiment of the present invention, and specifically, the method includes the following steps:
101: acquiring a spectral data set of a leaf sample and a corresponding chlorophyll content value;
in one embodiment, the test area of the blade sample was located at the official demonstration base of capsicum research institute of agricultural sciences institute of Guizhou province, new Zengyuan (E107 ° 2'27 ", N27 ° 44' 5"), at an altitude of 835m, in a subtropical monsoon climate, at an annual average air temperature of 14.7 ℃, at an annual average precipitation of 1200 mm, for a frost free period of 270 days. The soil type is yellow soil, soil basic indexes are measured before soil preparation, and the specific physical and chemical properties are as follows: the pH value is 5.21, the organic matter is 11.36g/kg, the total nitrogen is 1.78g/kg, the available phosphorus is 29.68mg/kg, and the quick-acting potassium is 109.59mg/kg.
In one embodiment, the types of leaf samples include: guizhou pepper No. 8 (cayenne pepper), red spicy No. 18 (cayenne pepper), chili No. 101 (pod pepper), and Red Global (pod pepper); the fertilizer comprises the following components: urea (ammonium sulfate), calcium superphosphate, and potassium sulfate.
In one embodiment, the test area of the leaf sample is designed by adopting a two-factor cleavage area, the 4 pepper varieties are used as a main area, the nitrogen fertilizer application amount is used as a secondary area, and 4 different nitrogen fertilizer application amounts are respectively 0kg/hm 2 、200kg/hm 2 、350kg/hm 2 、500kg/hm 2 The fertilizer is applied in the base fertilizer and the initial flowering stage according to the base additional fertilizer 1:1. The total number of the tests is 16 horizontal combinations, the test is repeated for 3 times, the total number is 48 cells, and each cell has the area of 28.8m 2 . The fertilizing amount of each phosphate fertilizer and potassium fertilizer is consistent, and the phosphate fertilizer (P) 2 O 5 ) The application rate is 150kg/hm 2 The fertilizer is used as a base fertilizer and is applied at one time; potassium fertilizer (K) 2 O) application rate of 300kg/hm 2 The top dressing is applied in the initial flowering period according to the base top dressing 1:1. The organic fertilizer is applied in a basal mode at a time according to the dosage of 100 kg/mu. Wherein, the nitrogenous fertilizer is urea (46.4%), the phosphate fertilizer is calcium superphosphate (16%), and the potash fertilizer is potassium sulfate (50%). The fertilizer application design for the test area is shown in the following table:
note: organic fertilizer: 100 kg/mu =4.3 kg/cell, cell area 28.8m 2 。
The test adopts soft disc substrate seedling culture, seeding is carried out in 2021 year, 4 month and 9 days, ridging is carried out according to 1.2m compartment width before transplanting, and field planting is carried out according to the plant row spacing of 0.3 multiplied by 0.6m when pepper seedlings reach 8 true leaves. Protective rows are arranged around the carriage, and a 0.3m passageway is reserved between the carriages, so that field investigation and field management are facilitated.
In one embodiment, a leaf sample is co-collected in spectral dataset (spectral reflectance) chlorophyll content value (SPAD) data 3699 groups, outliers are identified by mean ± 3-fold standard deviation of leaf chlorophyll, 19 outliers are identified, and data 3680 groups are retained after outliers are deleted, wherein V1 (qian la No. 8) 698 group, V2 (hong la No. 18) 856 group, V3 (la sha No. 101) 580 group, and V4 (hong global) 1546 group.
In one embodiment, the sample division method in the present application employs an SPXY method. The method is recently composed of
And the like firstly propose that the method is developed on the basis of a KS method, and experiments prove that the SPXY method can be effectively used for establishing an NIR quantitative model. SPXY between samplesThe distance calculation takes into account both the x and y variables. The method has the advantage of effectively covering the multidimensional vector space, thereby improving the prediction capability of the built model. The sample ratio of the correction set to the prediction set is 3:1.
102: preprocessing a spectrum data set of the leaf sample to obtain a preprocessed wavelength data set;
in one embodiment, the pre-treatment comprises one or more of: first derivative (FR, step 7), second derivative (SR, step 7), baseline correction (BLR), multivariate Scatter Correction (MSCR), variable normalization (SNVR), reciprocal (1/R), reciprocal Logarithm (LR). The pretreatment method is preferably: variable normalization (SNVR).
In one embodiment, the spectral data set of the leaf sample is a spectral data set obtained by smoothing the raw spectral data of the leaf sample: optionally, the smoothing method includes, but is not limited to, one of the following: savitzky-Golay (smooth window 9); the Savitzky-Golay filtering fitting method is that according to the average trend of an NDVI time sequence curve, proper filtering parameters are determined, least square fitting in a sliding window is achieved through a polynomial, the least square fitting is used for data preprocessing, and background noise is reduced.
103: carrying out variable screening on the preprocessed wavelength data set to obtain a target wavelength data set as a characteristic wavelength data set; the variable screening process comprises the following steps: constructing a model by adopting the preprocessed wavelength data set to obtain a constructed primary model; reserving points of the primary model with the weight of the regression coefficient absolute value greater than or equal to a first threshold as a new subset, and removing points with the weight less than or equal to the first threshold to obtain a secondary iteration subset after primary variable screening; establishing a second iteration model based on the second iteration subset to obtain a third iteration subset; after N times of circulation, respectively establishing N times of iteration models based on the N times of iteration subsets to respectively obtain N +1 times of iteration subsets; respectively calculating N groups of root mean square error values corresponding to the N times of iterative models, and determining an iterative subset corresponding to the minimum value of the root mean square error values as a characteristic wavelength data set;
in one embodiment, the characteristic wavelength data set includes one or more of: 510. 558, 559, 713, 1717, 1720, 1898, 2031, 2033, 2304. The characteristic wavelength data has a unit of nm.
In one embodiment, N is a natural integer.
104: and (3) constructing a model by using the characteristic wavelength data set, comparing the chlorophyll content predicted value obtained by using the model with the corresponding chlorophyll content value, optimizing the model, and obtaining the constructed chlorophyll content determination model.
In one embodiment, a machine learning method is used for carrying out model construction on the characteristic wavelength data to obtain a constructed chlorophyll content determination model;
optionally, the machine learning method includes one or more of the following: partial least square method regression, least square support vector machine, neural network, random forest method, linear regression, logistic regression, linear discriminant analysis, classification and regression tree, naive Bayes, KNN, learning vector quantization, support vector machine, lightGBM, and extreme gradient lifting;
optionally, the machine learning method includes: partial Least Squares Regression (PLSR); in the modeling process, PLSR integrates the characteristics of principal component analysis, canonical correlation analysis and linear regression analysis, so that in the analysis result, besides providing a more reasonable regression model, the PLSR can also complete some research contents similar to the principal component analysis and the canonical correlation analysis at the same time, and provide some more abundant and deeper information.
RMSE is adopted for precision verification v 、R 2 v 、RMSE p 、R 2 p And RPD5 indices. When RPD<1.4 hours indicate that the model cannot predict the sample; when 1.4<RPD<2, the model can roughly estimate the sample, and the prediction capability of the model can be improved by improving a modeling method; when RPD is present>2, indicates that the model has excellent predictive ability. R 2 The closer the (coefficient of determination) is to 1, the smaller the rmse value (root mean square error), the larger the RPD value (relative analytical error), the better the model.
FIG. 2 isThe invention provides a baseThe method for measuring the chlorophyll content in the wavelength is a schematic flow chart, fig. 3 is a flow chart used in the method for measuring the chlorophyll content based on the wavelength, which is provided by the embodiment of the invention, and concretely, the method comprises the following steps:
201: acquiring wavelength data of a leaf sample to be detected;
in one embodiment, the wavelength data of the leaf sample to be detected is obtained by preprocessing the spectrum data of the leaf sample to be detected; according to different methods for preprocessing the spectral data of the leaf sample to be detected, inputting the wavelength data of the leaf sample to be detected into a corresponding chlorophyll content measuring model to obtain a chlorophyll content value of the leaf sample to be detected;
202: inputting the wavelength data of the leaf sample to be detected into the chlorophyll content measurement model for processing to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, the wavelength data of the leaf sample to be detected is obtained by preprocessing the spectral data of the leaf sample to be detected; according to different methods for preprocessing the spectral data of the leaf sample to be detected, inputting the wavelength data of the leaf sample to be detected into a corresponding chlorophyll content measuring model to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, when the wavelength data of the leaf sample to be detected is subjected to variable standardization preprocessing, inputting the wavelength data into a chlorophyll content determination model I to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the first chlorophyll content measuring model comprises one or more of the following data: 510. 558, 559, 713, 1717, 1720, 1898, 2031, 2033, 2304;
when the wavelength data of the leaf sample to be detected is preprocessed through multivariate scattering correction, the wavelength data is input into a second chlorophyll content measuring model to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the second chlorophyll content measuring model comprises one or more of the following data: 404. 503, 505, 507, 508, 510, 547, 548, 549, 550, 551, 553, 554, 555, 556, 557, 688, 689, 690, 693, 708, 709, 710, 711, 712, 713, 1503, 1710, 1711, 1712, 1713, 1714, 1715, 1716, 1717, 1718, 1719, 1720, 1721, 1925, 1926, 2304, 2305, 2306, 2307;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal preprocessing, inputting the wavelength data into a chlorophyll content determination model III to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the chlorophyll content determination model III comprises one or more of the following data: 440. 441, 442, 443, 479, 480, 482, 533, 663, 664, 665, 677, 678, 681, 2211, 2212, 2213, 2215, 2216, 2232, 2233, 2235, 2236, 2240, 2241, 2244, 2245, 2290, 2292, 2293, 2294, 2295, 2296, 2297, 2298, 2299, 2300, 2301, 2302, 2303, 2307, 2308, 2309, 2310, 2311;
when the wavelength data of the leaf sample to be detected is not preprocessed, namely the wavelength data obtained by preprocessing the spectral data (marked as R) of the leaf sample to be detected is input into a chlorophyll content determination model IV to obtain a chlorophyll content value of the leaf sample to be detected; the chlorophyll content determination model four is a negative control; the characteristic wavelength data forming the fourth chlorophyll content determination model comprises one or more of the following data: 525. 529, 540, 542, 544, 546, 547, 548, 557, 558, 559, 560, 561, 562, 563, 578, 579, 580, 685, 837, 945, 1143, 1258, 1333, 1383, 1384, 1385, 1402, 1403, 1404, 1405, 1406, 1407, 1435, 1436, 1438, 1578, 1579, 1580, 1584, 1585, 1648, 1649, 1657, 1658, 1692, 1693, 1766, 1767, 1813, 1814, 1815;
when the wavelength data of the leaf sample to be detected is preprocessed by the second derivative, inputting the wavelength data into a chlorophyll content determination model V to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the chlorophyll content determination model V comprises one or more of the following data: 408. 422, 426, 427, 512, 513, 514, 542, 543, 544, 545, 546, 547, 548, 558, 559, 560, 561, 562, 585, 609, 610, 656, 657, 688, 703, 748, 1044, 1152, 1200, 1201, 1202, 1203, 1407, 1408, 1410, 1411, 1659, 1660, 1661, 1689, 1690, 1691, 1692, 1694, 1703, 1704, 1705, 1706, 1707, 1708, 1709, 1725, 1726, 1727, 1793, 1794, 1795, 1804, 1805, 1806, 1861, 1865, 1885, 2236, 2237, 2238, 2239;
when the wavelength data of the leaf sample to be detected is subjected to baseline correction preprocessing, inputting the wavelength data into a chlorophyll content determination model VI to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the chlorophyll content determination model six comprises one or more of the following data: 404. 502, 504, 505, 506, 507, 540, 542, 558, 559, 560, 561, 562, 697, 711, 836, 838, 839, 854, 926, 928, 931, 1149, 1150, 1151, 1152, 1155, 1156, 1158, 1160, 1257, 1258, 1259, 1260, 1261, 1262, 1263, 1264, 1265, 1332, 1333, 1335, 1378, 1379, 1399, 1402, 1403, 1404, 1407, 1409, 1410, 1443, 1444, 1447, 1449, 1450, 1452, 1453, 1454, 1550, 1557, 1585, 1590, 1647, 1648, 1649, 1650, 1651, 1653, 1654, 1655, 1656, 1658, 1659, 1665, 1691, 1692, 2293, 94, 95, 17096, 17097, 2298, 221, 1693, 223, 16983, 2284, 16983, 2282, 16983, 2284, 16983, 223, 16983, 2282, 16983, 224, 16983, 221, 16983, 2282;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal logarithm preprocessing, inputting the wavelength data into a chlorophyll content determination model seven to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the chlorophyll content determination model seven comprises one or more of the following data: 504. 505, 506, 507, 508, 509, 556, 557, 558, 560, 561, 562, 563, 566, 567, 568, 709, 710, 712, 713, 766, 767, 1326, 1327, 1328, 1334, 1336, 1338, 1342, 1343, 1345, 1346, 1347, 1352, 1356, 1358, 1702, 1704, 1705, 1706, 1707, 1708, 1709, 1710, 1711, 1712, 1713, 1714, 1715, 1716, 1718, 1719, 1720, 1721, 1722, 1723, 1724, 1725, 1901, 1902, 2042, 2043 2044, 2045, 2046, 2047, 2048, 2049, 2050, 2052, 2053, 2220, 2221, 2228, 2229, 2231, 2233, 2234, 2235, 2237, 2238, 2241, 2242, 2243, 2284, 2285, 2287, 2288, 2289, 2290, 2291, 2292, 2294, 2295, 2296, 2297, 2298, 2299, 2300, 2301, 2302, 2303, 2304, 2305, 2306, 2307, 2308, 2309, 2310, 2311, 2312, 2378, 2379, 2380, 2381, 2386, 2387, 2391, 2392, 2393;
when the wavelength data of the leaf sample to be detected is subjected to pretreatment of a first-order derivative, inputting the wavelength data into a chlorophyll content determination model eight to obtain a chlorophyll content value of the leaf sample to be detected; the characteristic wavelength data forming the chlorophyll content determination model eight comprises one or more of the following data: 403. 422, 423, 426, 446, 451, 452, 482, 483, 484, 485, 486, 488, 508, 523, 525, 535, 536, 537, 538, 539, 540, 541, 552, 553, 554, 555, 556, 557, 558, 559, 567, 568, 569, 570, 571, 575, 576, 589, 590, 591, 592, 632, 633, 738, 739, 746, 793, 805, 846, 892, 893, 894, 981, 990, 991, 1047, 1048, 1049, 1050, 1125, 1160, 1161, 1165, 1166, 1192, 1193, 1195, 1198, 1225, 1228, 1230, 1246, 1303, 1304, 1307, 1308, 1375, 1389, 1390, 1416, 1415, 1416, 483 1417, 1418, 1503, 1505, 1507, 1591, 1592, 1593, 1594, 1638, 1662, 1663, 1665, 1666, 1667, 1668, 1670, 1671, 1672, 1673, 1674, 1675, 1676, 1705, 1706, 1707, 1708, 1709, 1710, 1758, 1760, 1762, 1764, 1787, 1789, 1790, 1792, 1793, 1806, 1807, 1808, 1858, 1860, 1862, 1870, 1871, 1872, 1936, 1937, 1986, 1987, 1988, 2008, 2009, 2010, 2020, 2023, 2025, 2029, 2035, 2037, 2079, 2110, 2115, 2116, 2118, 2121, 2337, 2282, 2284, 2285, 2311, 229, 2341, 2390, 2393.
FIG. 4 isThe embodiment of the invention provides wavelength-based chlorophyll content measuring equipment, which comprises: a memory and a processor;
the memory is used for storing program instructions;
the processor is configured to invoke program instructions that, when executed, are configured to perform the wavelength-based chlorophyll content determination method described above.
FIG. 5Is thatThe chlorophyll content measuring system based on the wavelength provided by the embodiment of the invention comprises:
an obtaining unit 301, configured to obtain wavelength data of a blade sample to be measured;
the processing unit 302 is configured to input the wavelength data of the leaf sample to be detected into the chlorophyll content measurement model for processing, so as to obtain a chlorophyll content value of the leaf sample to be detected;
a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the above-described wavelength-based chlorophyll-content determining method.
FIG. 6 is a table of results of band filtering for target wavelength data provided by an embodiment of the present invention;
the variable screening is respectively carried out on various varieties and various spectrum pretreatment methods, the number results of characteristic wave bands under different varieties and spectrum pretreatment methods are different, the small number is V3-R (4), V2-LR (8), V2-BLR (10) and full sample-SNVR (10), the compression ratio is high and reaches 99.80%, 99.60%, 99.50% and 99.50%, the large number is V4-R (210), full sample-FR (159), V4-LR (158), V1-R (138), V3-SNVR (138) and V3-LR (138), and the compression ratio is 89.51%, 92.05%, 92.10%, 93.10%, 93.10% and 93.10%, respectively. In general, the compressibility is high, reaching 89% or more. The full sample is the sum of 4 tested species, increasing the complexity of the sample.
FIG. 7 is a table of modeling accuracy comparison of characteristic bands PLSR of chlorophyll content of pepper leaves according to an embodiment of the present invention;
there are also some differences between varieties and between spectrum pretreatments. Overall, the model accuracy is also higher, RMSE v Between 3.56 and 6.59, R 2 v Between 0.6557-0.8697, RMSE p Between 2.03 and 4.75, R 2 p Between 0.6908-0.9242, the RPD is between 2.46-4.52, and the prediction set accuracy is also higher than the correction set accuracy.
From the viewpoint of variety, V4 (RMSE) is relatively good v ：4.46±0.57，R 2 v ：0.7925±0.0530，RMSE p ：2.58±0.35，R 2 p :0.8662 ± 0.0386, RPD:3.62 ± 0.51) and the full sample (RMSE) with the relatively least effect v ：5.57±0.59，R 2 v ：0.7804±0.0507，RMSE p ：3.43±0.46，R 2 p :0.8177 ± 0.0534, RPD:3.28 ± 0.38). Overall, full sample modeling also shows better accuracy.
The most effective from the viewpoint of the spectral preprocessing method is SR (RMSE) v ：4.84±0.75，R 2 v ：0.8260±0.0400，RMSE p ：2.92±0.55，R 2 p :0.8965 ± 0.0353, RPD:3.87 ± 0.49), the effect is still 1/R (RMSE) relatively least v ：6.10±0.97，R 2 v ：0.7257±0.0665，RMSE p ：4.00±0.72，R 2 p ：0.7905±0.0630，RPD：2.82±0.28)。
Fig. 8 is a table of modeling accuracy comparison of full-wave band PLSR of chlorophyll content of pepper leaves provided by an embodiment of the present invention.
The PLSR is used for modeling the 400-2400nm full-wave band, and the results of the modeling are different between varieties and between spectrum pretreatment. Overall higher model accuracy, RMSE v Between 4.18 and 7.67, R 2 v Between 0.6363-0.8561, RMSE p Between 2.28 and 5.62, R 2 p Between 0.6937-0.9096, the RPD is between 2.38-4.02, and the prediction set accuracy is higher than the correction set accuracy.
From the viewpoint of variety, V2 (RMSE) is relatively effective v ：5.91±0.69，R 2 v ：0.8298±0.0475，RMSE p ：3.98±0.63，R 2 p :0.8803 ± 0.0342, RPD: 3.52. + -. 0.43) and V4 (RMSE) v ：4.49±0.22，R 2 v ：0.7936±0.0200，RMSE p ：2.48±0.12，R 2 p :0.8725 ± 0.0182, RPD:3.72 ± 0.18), the relatively least effective is the full sample (RMSE) v ：5.80±0.56，R 2 v ：0.7618±0.0508，RMSE p ：3.53±0.48，R 2 p :0.8110 ± 0.0478, rpd:3.18 ± 0.36). This is due to the difference in varietiesThe full sample is the sum of 4 tested species, the complexity of the sample is increased, the modeling precision is slightly reduced, and the full sample modeling still shows better precision on the whole.
FD (RMSE) is relatively effective from the viewpoint of the spectrum pretreatment method v ：5.46±0.41，R 2 v ：0.7778±0.0511，RMSE p ：3.05±0.44，R 2 p :0.8773 ± 0.0319, RPD: 3.66. + -. 0.18) and SNVR (RMSE) v ：5.25±0.50，R 2 v ：0.7923±0.0406，RMSE p ：3.22±0.45，R 2 p :0.8811 ± 0.0177, RPD: 3.48. + -. 0.36) and the relatively least effective is 1/R (RMSE) v ：5.51±0.76，R 2 v ：0.7761±0.0461，RMSE p ：3.50±0.65，R 2 p ：0.8480±0.0355，RPD：3.23±0.40)。
The validation results of this validation example show that assigning an intrinsic weight to an indication can moderately improve the performance of the method relative to the default settings.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by hardware related to instructions of a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: read Only Memory (ROM), random Access Memory (RAM), magnetic or optical disks, and the like.
It will be understood by those skilled in the art that all or part of the steps in the method according to the above embodiments may be implemented by hardware that is instructed by a program, and the program may be stored in a computer-readable storage medium, such as a read-only memory, a magnetic or optical disk, and the like.
While the invention has been described in detail with reference to certain embodiments, it will be apparent to one skilled in the art that the invention may be practiced without these specific details.
Claims (10)
1. A construction method of a chlorophyll content measurement model based on wavelength comprises the following steps:
acquiring a spectral data set of a leaf sample and a corresponding chlorophyll content value;
preprocessing the spectral data set of the leaf sample to obtain a preprocessed wavelength data set; carrying out variable screening on the preprocessed wavelength data set to obtain a target wavelength data set as a characteristic wavelength data set; the variable screening process comprises the following steps: constructing a model by adopting the preprocessed wavelength data set to obtain a constructed primary model; reserving points of the primary model with the weight of the regression coefficient absolute value greater than or equal to a first threshold as a new subset, and removing points with the weight less than or equal to the first threshold to obtain a secondary iteration subset after primary variable screening; establishing a second iteration model based on the second iteration subset to obtain a third iteration subset; after N cycles, respectively establishing N iteration models based on the N iteration subsets to respectively obtain N +1 iteration subsets; respectively calculating N groups of root mean square error values corresponding to the N times of iterative models, and determining an iterative subset corresponding to the minimum value of the root mean square error values as a characteristic wavelength data set;
and constructing a model by using the characteristic wavelength data set, comparing the chlorophyll content predicted value obtained by using the model with the corresponding chlorophyll content value, and optimizing the model to obtain the constructed chlorophyll content determination model.
2. The method for constructing the model for measuring chlorophyll content on the basis of wavelength according to claim 1, wherein the pretreatment comprises one or more of the following steps: first derivative, second derivative, baseline correction, multivariate scatter correction, variable normalization, reciprocal logarithm.
3. The method for constructing the model for measuring chlorophyll content based on wavelength according to claim 2, wherein the pretreatment method comprises: and (6) normalizing the variables.
4. The method for constructing a model for measuring chlorophyll content based on wavelength according to claim 1, wherein the spectral data set of the leaf sample is obtained by smoothing raw spectral data of the leaf sample: optionally, the smoothing method includes, but is not limited to, one of the following: savitzky-Golay.
5. The method for constructing a wavelength-based chlorophyll content measurement model according to claim 1, wherein the characteristic wavelength data set comprises one or more of: 510. 558, 559, 713, 1717, 1720, 1898, 2031, 2033, 2304.
6. The method for constructing the chlorophyll content measurement model based on the wavelength according to claim 1, wherein the model construction is performed on the characteristic wavelength data by a machine learning method to obtain the constructed chlorophyll content measurement model;
optionally, the machine learning method includes one or more of the following: partial least square method regression, least square support vector machine, neural network, random forest method, linear regression, logistic regression, linear discriminant analysis, classification and regression tree, naive Bayes, KNN, learning vector quantization, support vector machine, lightGBM, and extreme gradient lifting;
optionally, the machine learning method includes: partial Least Squares Regression (PLSR).
7. A method for wavelength-based chlorophyll content determination, comprising:
acquiring wavelength data of a leaf sample to be detected;
inputting the wavelength data of the leaf sample to be detected into the chlorophyll content measuring model in claims 1-6 for processing to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, the wavelength data of the leaf sample to be detected is obtained by preprocessing the spectral data of the leaf sample to be detected; according to different methods for preprocessing the spectral data of the leaf sample to be detected, inputting the wavelength data of the leaf sample to be detected into a corresponding chlorophyll content determination model to obtain a chlorophyll content value of the leaf sample to be detected;
optionally, when the wavelength data of the to-be-detected leaf sample is subjected to variable standardization preprocessing, inputting the wavelength data into a chlorophyll content determination model I to obtain a chlorophyll content value of the to-be-detected leaf sample; when the wavelength data of the leaf sample to be detected is subjected to preprocessing of multivariate scattering correction, inputting the wavelength data into a chlorophyll content determination model II to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal preprocessing, inputting the wavelength data into a chlorophyll content determination model III to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is subjected to second derivative pretreatment, inputting the wavelength data into a chlorophyll content determination model V to obtain a chlorophyll content value of the leaf sample to be detected;
when the wavelength data of the leaf sample to be detected is preprocessed through baseline correction, the wavelength data is input into a chlorophyll content measuring model VI, and a chlorophyll content value of the leaf sample to be detected is obtained;
when the wavelength data of the leaf sample to be detected is subjected to reciprocal logarithm preprocessing, inputting the wavelength data into a chlorophyll content determination model seven to obtain a chlorophyll content value of the leaf sample to be detected;
and when the wavelength data of the leaf sample to be detected is subjected to the pretreatment of the first derivative, inputting the wavelength data into a chlorophyll content determination model eight to obtain the chlorophyll content value of the leaf sample to be detected.
8. A wavelength-based chlorophyll-content determining apparatus, the apparatus comprising: a memory and a processor;
the memory is to store program instructions;
the processor is configured to invoke program instructions that, when executed, are configured to perform the wavelength-based chlorophyll content determination method of claim 7.
9. A wavelength-based chlorophyll content determination system, comprising:
the acquisition unit is used for acquiring wavelength data of a blade sample to be detected;
the processing unit is used for inputting the wavelength data of the leaf sample to be detected into the chlorophyll content determination model in claims 1 to 6 for processing, so as to obtain the chlorophyll content value of the leaf sample to be detected.
10. A computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the wavelength-based chlorophyll content determining method according to claim 7 above.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211100290.8A CN115479902A (en) | 2022-09-08 | 2022-09-08 | Chlorophyll content determination model construction method based on wavelength and system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211100290.8A CN115479902A (en) | 2022-09-08 | 2022-09-08 | Chlorophyll content determination model construction method based on wavelength and system thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115479902A true CN115479902A (en) | 2022-12-16 |
Family
ID=84392901
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211100290.8A Pending CN115479902A (en) | 2022-09-08 | 2022-09-08 | Chlorophyll content determination model construction method based on wavelength and system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115479902A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116660206A (en) * | 2023-05-31 | 2023-08-29 | 浙江省农业科学院 | Crop yield estimation method and system |
| CN117451639A (en) * | 2023-12-21 | 2024-01-26 | 内蒙古工业大学 | Water chlorophyll concentration inversion method based on remote sensing data |
-
2022
- 2022-09-08 CN CN202211100290.8A patent/CN115479902A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116660206A (en) * | 2023-05-31 | 2023-08-29 | 浙江省农业科学院 | Crop yield estimation method and system |
| CN116660206B (en) * | 2023-05-31 | 2024-05-28 | 浙江省农业科学院 | Crop yield estimation method and system |
| CN117451639A (en) * | 2023-12-21 | 2024-01-26 | 内蒙古工业大学 | Water chlorophyll concentration inversion method based on remote sensing data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110082300B (en) | Method for establishing winter wheat canopy water content monitoring model based on spectral parameters | |
| Zhang et al. | Rapid determination of leaf water content using VIS/NIR spectroscopy analysis with wavelength selection | |
| CN109187441B (en) | Method for constructing summer corn nitrogen content monitoring model based on canopy spectral information | |
| Jain et al. | Use of hyperspectral data to assess the effects of different nitrogen applications on a potato crop | |
| CN110376167A (en) | Rice leaf nitrogen content monitoring method based on unmanned plane EO-1 hyperion | |
| Li et al. | Discriminating varieties of tea plant based on Vis/NIR spectral characteristics and using artificial neural networks | |
| CN115479902A (en) | Chlorophyll content determination model construction method based on wavelength and system thereof | |
| Christenson et al. | Predicting soybean relative maturity and seed yield using canopy reflectance | |
| Li et al. | Assessing leaf nitrogen concentration of winter oilseed rape with canopy hyperspectral technique considering a non-uniform vertical nitrogen distribution | |
| CN110082309B (en) | Method for establishing SPAD value comprehensive spectrum monitoring model of winter wheat canopy | |
| CN110069895B (en) | Method for establishing winter wheat nitrogen content full-growth period spectrum monitoring model | |
| Yu et al. | A study of nitrogen deficiency inversion in rice leaves based on the hyperspectral reflectance differential | |
| US20230078617A1 (en) | A method for assessing nitrogen nutritional status in plants by visible-to-shortwave infrared reflectance spectroscopy of carbohydrates | |
| Guo et al. | A robust method to estimate foliar phosphorus of rubber trees with hyperspectral reflectance | |
| Zaeen et al. | In‐season potato yield prediction with active optical sensors | |
| Wen et al. | Estimation of the vertically integrated leaf nitrogen content in maize using canopy hyperspectral red edge parameters | |
| Chivasa et al. | Phenology-based discrimination of maize (Zea mays L.) varieties using multitemporal hyperspectral data | |
| Peng et al. | Winter wheat canopy water content monitoring based on spectral transforms and “three-edge” parameters | |
| Yu et al. | An entirely new approach based on remote sensing data to calculate the nitrogen nutrition index of winter wheat | |
| Rubo et al. | Exploring hyperspectral reflectance indices for the estimation of water and nitrogen status of spinach | |
| Wang et al. | Prediction of N, P, and K contents in sugarcane leaves by VIS-NIR spectroscopy and modeling of NPK interaction effects | |
| CN115479903A (en) | Chlorophyll content measuring method and system based on wavelength | |
| Ong et al. | Visible and near-infrared spectroscopic determination of sugarcane chlorophyll content using a modified wavelength selection method for multivariate calibration | |
| Wang Xue et al. | Monitoring model for predicting maize grain moisture at the filling stage using NIRS and a small sample size. | |
| Li et al. | Spectrometric prediction of nitrogen content in different tissues of slash pine trees |
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 |