CN113008890B - Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system - Google Patents
Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system Download PDFInfo
- Publication number
- CN113008890B CN113008890B CN202110258200.7A CN202110258200A CN113008890B CN 113008890 B CN113008890 B CN 113008890B CN 202110258200 A CN202110258200 A CN 202110258200A CN 113008890 B CN113008890 B CN 113008890B
- Authority
- CN
- China
- Prior art keywords
- hyperspectral
- nitrogen content
- data
- monitoring
- sensitive
- 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
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 title claims abstract description 300
- 229910052757 nitrogen Inorganic materials 0.000 title claims abstract description 150
- 238000012544 monitoring process Methods 0.000 title claims abstract description 137
- 229920000742 Cotton Polymers 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 45
- 241000219146 Gossypium Species 0.000 claims abstract description 76
- 238000012549 training Methods 0.000 claims abstract description 43
- 238000012417 linear regression Methods 0.000 claims abstract description 25
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 18
- 238000001228 spectrum Methods 0.000 claims abstract description 7
- 238000009826 distribution Methods 0.000 claims description 33
- 239000002689 soil Substances 0.000 claims description 27
- 238000002310 reflectometry Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 9
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 9
- 238000010521 absorption reaction Methods 0.000 claims description 6
- 230000035945 sensitivity Effects 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 3
- 230000001419 dependent effect Effects 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000004590 computer program Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 4
- 230000003595 spectral effect Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 239000000618 nitrogen fertilizer Substances 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 235000016709 nutrition Nutrition 0.000 description 1
- 230000035764 nutrition Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
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
- G01N21/84—Systems specially adapted for particular applications
-
- 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
- G01N21/84—Systems specially adapted for particular applications
- G01N2021/8466—Investigation of vegetal material, e.g. leaves, plants, fruits
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application discloses a cotton leaf nitrogen content monitoring method and system based on unmanned aerial vehicle hyperspectrum, which relate to the technical field of crop biochemical component monitoring and have the technical scheme that: classifying sensitive wavelength bands to obtain a crest sensitive set and a trough sensitive set; acquiring hyperspectral image information of a target area and the nitrogen content of corresponding cotton leaves; decomposing the mixed pixel by a mixed spectrum decomposition method to obtain a multi-hyperspectral data set; intercepting a first multi-hyperspectral feature discrete set and a second multi-hyperspectral feature discrete set from the multi-hyperspectral data set; obtaining a first nitrogen content monitoring model and a second nitrogen content monitoring model through a linear regression training model; and respectively inputting the estimated wavelength and the multispectral characteristic estimation into a first nitrogen content monitoring model and a second nitrogen content monitoring model to be matched, and preferentially selecting the monitoring models according to the matching result to carry out nitrogen content estimation monitoring on the target cotton leaves. The application can effectively improve the accuracy of the monitoring result.
Description
Technical Field
The application relates to the technical field of crop biochemical component monitoring, in particular to a cotton leaf nitrogen content monitoring method and system based on unmanned aerial vehicle hyperspectrum.
Background
The nitrogen is a main nutrient element required by most crops, such as cotton, wheat, rice and the like, the nitrogen content directly influences the growth and development of the crops and the formation of yield and quality, the nitrogen content of the crop leaves is detected rapidly and accurately in real time, the nitrogen is beneficial to guiding the application of nitrogen fertilizer scientifically and reasonably, and the nitrogen fertilizer has important practical significance and application value in reducing environmental pollution caused by excessive nitrogen application and improving the yield and quality of the crops. In recent years, with the rapid development of hyperspectral technology, nondestructive detection technology is applied to perform nondestructive monitoring estimation on biochemical components of crops including nitrogen, and the nondestructive detection technology is an important method for researching the growth vigor and nutrition diagnosis of crops in the field.
At present, two methods for detecting the nitrogen content of crops by hyperspectral technology are mainly available. First, various spectral vegetation indexes generated by using reflectivity combinations of sensitive wavelengths; and secondly, a response model is directly built by applying spectral characteristic variables and the nitrogen content of crops, so that the spectral nondestructive estimation of the nitrogen content is realized. However, the hyperspectral image is not only the hyperspectral data of the crop leaves are obtained singly during imaging, but also the hyperspectral data of various other ground objects such as soil, moisture, weeds and the like are included, and the effect of the other ground objects on the monitoring of the nitrogen content of the crop is considered in the method for detecting the nitrogen content of the crop through the hyperspectral technology, so that a larger error exists in the result of monitoring the nitrogen content of the actual crop.
Therefore, how to research and design an accurate and reasonable cotton leaf nitrogen content monitoring method and system based on unmanned aerial vehicle hyperspectrum is a problem which needs to be solved at present.
Disclosure of Invention
The application aims to solve the problem of low accuracy of monitoring crop content by using the existing hyperspectral technology, and provides a cotton leaf nitrogen content monitoring method and system based on hyperspectral of an unmanned plane.
The technical aim of the application is realized by the following technical scheme:
in a first aspect, a method for monitoring nitrogen content of cotton leaves based on hyperspectral of unmanned aerial vehicle is provided, which comprises the following steps:
s101: acquiring hyperspectral reflectivity data of nitrogen in cotton leaves, intercepting a sensitive wavelength band from the hyperspectral reflectivity data, and classifying the sensitive wavelength band according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set;
s102: acquiring hyperspectral image information of a target area and the nitrogen content of corresponding cotton leaves;
s103: dividing hyperspectral image information into a plurality of mixed pixels, decomposing the mixed pixels by a mixed spectrum decomposition method, and classifying and integrating decomposition results to obtain a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data;
s104: intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the peak sensitivity set, wherein the first multi-hyperspectral feature group of the plurality of wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitive set, the second multi-hyperspectral feature sets of the multiple wavelengths forming a second multi-hyperspectral feature discrete set of the same hybrid pixel;
s105: inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a first nitrogen content monitoring model; inputting a second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a second nitrogen content monitoring model;
s106: acquiring hyperspectral data information of a target cotton leaf, and randomly selecting an estimated wavelength and a multispectral characteristic estimation group under a corresponding wavelength from the hyperspectral data information;
s107: and respectively inputting the estimated wavelength and the multispectral characteristic estimation into a first nitrogen content monitoring model and a second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation and monitoring on the target cotton leaf to obtain an estimation and monitoring result.
Further, the classifying process of the crest sensitive set and the trough sensitive set specifically comprises the following steps:
smoothing the hyperspectral reflectivity data to establish a hyperspectral reflectivity curve;
slope solving is carried out on the hyperspectral reflectivity curve, and calculus processing is carried out on slopes in continuous wave bands in the hyperspectral reflectivity curve to obtain wave band slopes;
the slope of the N-th wave band is simultaneously larger than the slope of the N-1 th wave band and the slope of the (n+1) -th wave band, and then the N-th wave band is used as a first sensitive wavelength band, and a plurality of first sensitive wavelength bands form a wave crest sensitive set;
and screening the N-th wave band slope with the N-th wave band slope smaller than the N-1-th wave band slope and the N+1-th wave band slope, and taking the N-th wave band as a second sensitive wavelength band, wherein a plurality of second sensitive wavelength bands form a trough sensitive set.
Further, the specific process of decomposing the hyperspectral image information to obtain the multispectral data set is as follows:
respectively extracting hyperspectral data in each mixed pixel in hyperspectral image information to obtain hyperspectral distribution data;
dividing hyperspectral distribution data into soil distribution data, water distribution data and leaf distribution data according to soil, water and leaf ground object categories;
the soil distribution data, the moisture distribution data and the leaf distribution data are integrated into a multi-hyperspectral data set which is continuously distributed and simultaneously contains the soil distribution data, the moisture distribution data and the leaf distribution data under a single wavelength.
Further, the establishing process of the first nitrogen content monitoring model and the second nitrogen content monitoring model specifically comprises the following steps:
carrying out regression training on the multi-hyperspectral feature groups and the nitrogen content of cotton leaves under different wavelengths in the multi-hyperspectral feature discrete set through a linear regression training model to obtain a quaternary linear regression training model;
the quaternary linear regression training model takes soil, moisture and leaf hyperspectral data as independent variables and takes nitrogen content as dependent variables;
and carrying out optimization correction training on the quaternary linear regression training models under different multiple mixed pixels to obtain a corresponding first nitrogen content monitoring model or a second nitrogen content monitoring model.
Further, the matching process of the first nitrogen content monitoring model and the second nitrogen content monitoring model specifically comprises the following steps:
randomly determining an estimated wavelength, and selecting a multi-hyperspectral characteristic estimation group of a corresponding wavelength from hyperspectral data information of a target cotton blade according to the estimated wavelength;
respectively inputting the multispectral characteristic estimation models into a first nitrogen content monitoring model and a second nitrogen content monitoring model to calculate the matching degree, and respectively obtaining a first matching degree and a second matching degree;
and carrying out nitrogen content estimation monitoring on the target cotton leaf by using a monitoring model corresponding to the maximum value in the first matching degree and the second matching degree to obtain a nitrogen content monitoring result in the target cotton leaf.
Further, the calculating process of the first matching degree and the second matching degree specifically includes:
respectively calculating the correlation between soil, water and leaf hyperspectral features in the multispectral feature estimation group to obtain three correlation coefficients;
calculating a first uniformity correlation of the multi-hyperspectral feature estimation group according to the three correlation coefficients;
and taking the ratio of the first uniformity correlation of the multi-hyperspectral characteristic estimation group to the second uniformity correlation calculated by the monitoring model as the matching degree.
Furthermore, the hyperspectral data are collected in a low-altitude mode within a height of 10-50m through the unmanned aerial vehicle loaded with the MVD, and the collection wave band range is 400-1200nm.
In a second aspect, a cotton leaf nitrogen content monitoring system based on unmanned aerial vehicle hyperspectrum is provided, comprising:
the data classification module is used for acquiring hyperspectral reflectivity data of nitrogen in cotton leaves, intercepting a sensitive wavelength band from the hyperspectral reflectivity data, and classifying the sensitive wavelength band according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set;
the data acquisition module is used for acquiring hyperspectral image information of the target area and the nitrogen content of the corresponding cotton leaf;
the data decomposition module is used for dividing the hyperspectral image information into a plurality of mixed pixels, decomposing the mixed pixels by a mixed spectrum decomposition method, and classifying and integrating decomposition results to obtain a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data;
the data interception module is used for intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the crest sensitive set, and the first multi-hyperspectral feature group of the multiple wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitive set, the second multi-hyperspectral feature sets of the multiple wavelengths forming a second multi-hyperspectral feature discrete set of the same hybrid pixel;
the model construction module is used for inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into the linear regression training model for training to obtain a first nitrogen content monitoring model; inputting a second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a second nitrogen content monitoring model;
the data acquisition module is used for acquiring hyperspectral data information of the target cotton leaf and randomly selecting an estimated wavelength and a multispectral characteristic estimation group under the corresponding wavelength from the hyperspectral data information;
the monitoring module is used for respectively inputting the estimated wavelength and the multispectral characteristic estimation module into the first nitrogen content monitoring model and the second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation monitoring on the target cotton leaf so as to obtain an estimation monitoring result.
Compared with the prior art, the application has the following beneficial effects:
1. according to the method, the peaks and the troughs of the sensitive wavelength bands of the cotton leaf nitrogen are divided, the corresponding hyperspectral characteristic groups are intercepted, a first nitrogen content monitoring model and a second nitrogen content monitoring model are established, and the cotton leaf nitrogen content is independently monitored through the first nitrogen content monitoring model and the second nitrogen content monitoring model, so that the integral error range of the monitoring result can be effectively reduced, and the accuracy of monitoring the cotton leaf nitrogen content is improved;
2. according to the application, the first nitrogen content monitoring model or the second nitrogen content monitoring model established by comprehensively considering the association relation between soil, moisture and leaf series ground objects and cotton leaf nitrogen content can be more in line with the actual situation of cotton leaf nitrogen content;
3. and the matching degree is calculated, and a proper first nitrogen content monitoring model or second nitrogen content monitoring model is selected for monitoring, so that accurate positioning monitoring is realized.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:
FIG. 1 is a flow chart in an embodiment of the application;
fig. 2 is a system architecture diagram in an embodiment of the application.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present application, the present application will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present application and the descriptions thereof are for illustrating the present application only and are not to be construed as limiting the present application.
Example 1: the cotton leaf nitrogen content monitoring method based on the hyperspectral of the unmanned aerial vehicle is shown in fig. 1, and is specifically realized by the following steps.
S101: the hyperspectral reflectivity data of the nitrogen in the cotton leaf is obtained, sensitive wavelength bands are intercepted from the hyperspectral reflectivity data, and the sensitive wavelength bands are classified according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set.
The classification processing process of the crest sensitive set and the trough sensitive set specifically comprises the following steps: smoothing the hyperspectral reflectivity data to establish a hyperspectral reflectivity curve; slope solving is carried out on the hyperspectral reflectivity curve, and calculus processing is carried out on slopes in continuous wave bands in the hyperspectral reflectivity curve to obtain wave band slopes; the slope of the N-th wave band is simultaneously larger than the slope of the N-1 th wave band and the slope of the (n+1) -th wave band, and then the N-th wave band is used as a first sensitive wavelength band, and a plurality of first sensitive wavelength bands form a wave crest sensitive set; and screening the N-th wave band slope with the N-th wave band slope smaller than the N-1-th wave band slope and the N+1-th wave band slope, and taking the N-th wave band as a second sensitive wavelength band, wherein a plurality of second sensitive wavelength bands form a trough sensitive set.
S102: hyperspectral image information of a target area and corresponding nitrogen content of cotton leaves are obtained.
S103: the hyperspectral image information is divided into a plurality of mixed pixels, the mixed pixels are decomposed by a mixed spectrum decomposition method, and a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data is obtained after classification and integration of decomposition treatment results.
The specific process for obtaining the multi-hyperspectral data set by the hyperspectral image information decomposition processing is as follows: respectively extracting hyperspectral data in each mixed pixel in hyperspectral image information to obtain hyperspectral distribution data; dividing hyperspectral distribution data into soil distribution data, water distribution data and leaf distribution data according to soil, water and leaf ground object categories; the soil distribution data, the moisture distribution data and the leaf distribution data are integrated into a multi-hyperspectral data set which is continuously distributed and simultaneously contains the soil distribution data, the moisture distribution data and the leaf distribution data under a single wavelength.
S104: intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the peak sensitivity set, wherein the first multi-hyperspectral feature group of the plurality of wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitivity set, wherein the second multi-hyperspectral feature sets of the multiple wavelengths form a second multi-hyperspectral feature discrete set of the same mixed pixel.
S105: inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a first nitrogen content monitoring model; and inputting the second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels into a linear regression training model for training to obtain a second nitrogen content monitoring model.
The building process of the first nitrogen content monitoring model and the second nitrogen content monitoring model specifically comprises the following steps: carrying out regression training on the multi-hyperspectral feature groups and the nitrogen content of cotton leaves under different wavelengths in the multi-hyperspectral feature discrete set through a linear regression training model to obtain a quaternary linear regression training model; the quaternary linear regression training model takes soil, moisture and leaf hyperspectral data as independent variables and takes nitrogen content as dependent variables; and carrying out optimization correction training on the quaternary linear regression training models under different multiple mixed pixels to obtain a corresponding first nitrogen content monitoring model or a second nitrogen content monitoring model.
S106: and acquiring hyperspectral data information of the target cotton leaf, and randomly selecting an estimated wavelength and a multispectral characteristic estimated group under the corresponding wavelength from the hyperspectral data information.
S107: and respectively inputting the estimated wavelength and the multispectral characteristic estimation into a first nitrogen content monitoring model and a second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation and monitoring on the target cotton leaf to obtain an estimation and monitoring result.
The matching process of the first nitrogen content monitoring model and the second nitrogen content monitoring model specifically comprises the following steps: randomly determining an estimated wavelength, and selecting a multi-hyperspectral characteristic estimation group of a corresponding wavelength from hyperspectral data information of a target cotton blade according to the estimated wavelength; respectively inputting the multispectral characteristic estimation models into a first nitrogen content monitoring model and a second nitrogen content monitoring model to calculate the matching degree, and respectively obtaining a first matching degree and a second matching degree; and carrying out nitrogen content estimation monitoring on the target cotton leaf by using a monitoring model corresponding to the maximum value in the first matching degree and the second matching degree to obtain a nitrogen content monitoring result in the target cotton leaf.
In addition, the calculation process of the first matching degree and the second matching degree specifically comprises the following steps: respectively calculating the correlation between soil, water and leaf hyperspectral features in the multispectral feature estimation group to obtain three correlation coefficients; calculating a first uniformity correlation of the multi-hyperspectral feature estimation group according to the three correlation coefficients; and taking the ratio of the first uniformity correlation of the multi-hyperspectral characteristic estimation group to the second uniformity correlation calculated by the monitoring model as the matching degree.
In this embodiment, the hyperspectral data is collected in a low altitude of 10-50m by the unmanned aerial vehicle loaded with MVD, and the collection band range is 400-1200nm.
Example 2: the cotton leaf nitrogen content monitoring system based on the hyperspectrum of the unmanned aerial vehicle comprises a data classification module, a data acquisition module, a data decomposition module, a data interception module, a model construction module, a data acquisition module and a monitoring module as shown in fig. 2.
The data classification module is used for acquiring hyperspectral reflectivity data of nitrogen in cotton leaves, intercepting a sensitive wavelength band from the hyperspectral reflectivity data, and classifying the sensitive wavelength band according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set.
And the data acquisition module is used for acquiring hyperspectral image information of the target area and the nitrogen content of the corresponding cotton leaf.
The data decomposition module is used for dividing the hyperspectral image information into a plurality of mixed pixels, decomposing the mixed pixels by a mixed spectrum decomposition method, and classifying and integrating decomposition results to obtain a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data.
The data interception module is used for intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the crest sensitive set, and the first multi-hyperspectral feature group of the multiple wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitivity set, wherein the second multi-hyperspectral feature sets of the multiple wavelengths form a second multi-hyperspectral feature discrete set of the same mixed pixel.
The model construction module is used for inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into the linear regression training model for training to obtain a first nitrogen content monitoring model; and inputting the second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels into a linear regression training model for training to obtain a second nitrogen content monitoring model.
The data acquisition module is used for acquiring hyperspectral data information of the target cotton leaf and randomly selecting an estimated wavelength and a multispectral characteristic estimated group under the corresponding wavelength from the hyperspectral data information.
The monitoring module is used for respectively inputting the estimated wavelength and the multispectral characteristic estimation module into the first nitrogen content monitoring model and the second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation monitoring on the target cotton leaf so as to obtain an estimation monitoring result.
Working principle: according to the method, the peaks and the troughs of the sensitive wavelength bands of the cotton leaf nitrogen are divided, the corresponding hyperspectral characteristic groups are intercepted, a first nitrogen content monitoring model and a second nitrogen content monitoring model are established, and the cotton leaf nitrogen content is independently monitored through the first nitrogen content monitoring model and the second nitrogen content monitoring model, so that the integral error range of the monitoring result can be effectively reduced, and the accuracy of monitoring the cotton leaf nitrogen content is improved; in addition, the first nitrogen content monitoring model or the second nitrogen content monitoring model established by comprehensively considering the association relation between soil, moisture and leaf series ground objects and cotton leaf nitrogen content can be more in line with the actual situation of cotton leaf nitrogen content; in addition, the matching degree is calculated, and a proper first nitrogen content monitoring model or a proper second nitrogen content monitoring model is selected for monitoring, so that accurate positioning monitoring is realized.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing detailed description of the application has been presented for purposes of illustration and description, and it should be understood that the application is not limited to the particular embodiments disclosed, but is intended to cover all modifications, equivalents, alternatives, and improvements within the spirit and principles of the application.
Claims (8)
1. The cotton leaf nitrogen content monitoring method based on the hyperspectrum of the unmanned aerial vehicle is characterized by comprising the following steps of:
s101: acquiring hyperspectral reflectivity data of nitrogen in cotton leaves, intercepting a sensitive wavelength band from the hyperspectral reflectivity data, and classifying the sensitive wavelength band according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set;
s102: acquiring hyperspectral image information of a target area and the nitrogen content of corresponding cotton leaves;
s103: dividing hyperspectral image information into a plurality of mixed pixels, decomposing the mixed pixels by a mixed spectrum decomposition method, and classifying and integrating decomposition results to obtain a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data;
s104: intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the peak sensitivity set, wherein the first multi-hyperspectral feature group of the plurality of wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitive set, the second multi-hyperspectral feature sets of the multiple wavelengths forming a second multi-hyperspectral feature discrete set of the same hybrid pixel;
s105: inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a first nitrogen content monitoring model; inputting a second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a second nitrogen content monitoring model;
s106: acquiring hyperspectral data information of a target cotton leaf, and randomly selecting an estimated wavelength and a multispectral characteristic estimation group under a corresponding wavelength from the hyperspectral data information;
s107: and respectively inputting the estimated wavelength and the multispectral characteristic estimation into a first nitrogen content monitoring model and a second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation and monitoring on the target cotton leaf to obtain an estimation and monitoring result.
2. The method for monitoring nitrogen content of cotton leaf based on hyperspectrum of unmanned aerial vehicle according to claim 1, wherein the classifying treatment process of the crest sensitive set and the trough sensitive set is specifically as follows:
smoothing the hyperspectral reflectivity data to establish a hyperspectral reflectivity curve;
slope solving is carried out on the hyperspectral reflectivity curve, and calculus processing is carried out on slopes in continuous wave bands in the hyperspectral reflectivity curve to obtain wave band slopes;
the slope of the N-th wave band is simultaneously larger than the slope of the N-1 th wave band and the slope of the (n+1) -th wave band, and then the N-th wave band is used as a first sensitive wavelength band, and a plurality of first sensitive wavelength bands form a wave crest sensitive set;
and screening the N-th wave band slope with the N-th wave band slope smaller than the N-1-th wave band slope and the N+1-th wave band slope, and taking the N-th wave band as a second sensitive wavelength band, wherein a plurality of second sensitive wavelength bands form a trough sensitive set.
3. The method for monitoring nitrogen content of cotton leaf based on hyperspectral of unmanned aerial vehicle according to claim 1, wherein the specific process of decomposing hyperspectral image information to obtain a multispectral data set is as follows:
respectively extracting hyperspectral data in each mixed pixel in hyperspectral image information to obtain hyperspectral distribution data;
dividing hyperspectral distribution data into soil distribution data, water distribution data and leaf distribution data according to soil, water and leaf ground object categories;
the soil distribution data, the moisture distribution data and the leaf distribution data are integrated into a multi-hyperspectral data set which is continuously distributed and simultaneously contains the soil distribution data, the moisture distribution data and the leaf distribution data under a single wavelength.
4. The method for monitoring nitrogen content of cotton leaf based on hyperspectral of unmanned aerial vehicle according to claim 1, wherein the establishing process of the first nitrogen content monitoring model and the second nitrogen content monitoring model is specifically as follows:
carrying out regression training on the multi-hyperspectral feature groups and the nitrogen content of cotton leaves under different wavelengths in the multi-hyperspectral feature discrete set through a linear regression training model to obtain a quaternary linear regression training model;
the quaternary linear regression training model takes soil, moisture and leaf hyperspectral data as independent variables and takes nitrogen content as dependent variables;
and carrying out optimization correction training on the quaternary linear regression training models under different multiple mixed pixels to obtain a corresponding first nitrogen content monitoring model or a second nitrogen content monitoring model.
5. The method for monitoring nitrogen content of cotton leaf based on hyperspectral of unmanned aerial vehicle according to claim 1, wherein the matching process of the first nitrogen content monitoring model and the second nitrogen content monitoring model is specifically as follows:
randomly determining an estimated wavelength, and selecting a multi-hyperspectral characteristic estimation group of a corresponding wavelength from hyperspectral data information of a target cotton blade according to the estimated wavelength;
respectively inputting the multispectral characteristic estimation models into a first nitrogen content monitoring model and a second nitrogen content monitoring model to calculate the matching degree, and respectively obtaining a first matching degree and a second matching degree;
and carrying out nitrogen content estimation monitoring on the target cotton leaf by using a monitoring model corresponding to the maximum value in the first matching degree and the second matching degree to obtain a nitrogen content monitoring result in the target cotton leaf.
6. The method for monitoring nitrogen content of cotton leaf based on hyperspectral of unmanned aerial vehicle according to claim 5, wherein the calculation process of the first matching degree and the second matching degree is specifically as follows:
respectively calculating the correlation between soil, water and leaf hyperspectral features in the multispectral feature estimation group to obtain three correlation coefficients;
calculating a first uniformity correlation of the multi-hyperspectral feature estimation group according to the three correlation coefficients;
and taking the ratio of the first uniformity correlation of the multi-hyperspectral characteristic estimation group to the second uniformity correlation calculated by the monitoring model as the matching degree.
7. The method for monitoring nitrogen content of cotton leaf based on hyperspectral of unmanned aerial vehicle according to claim 1, wherein the hyperspectral data is acquired in a low-altitude mode within 10-50m by unmanned aerial vehicle loaded with MVD, and the acquisition band range is 400-1200nm.
8. Cotton leaf nitrogen content monitoring system based on unmanned aerial vehicle hyperspectrum, characterized by includes:
the data classification module is used for acquiring hyperspectral reflectivity data of nitrogen in cotton leaves, intercepting a sensitive wavelength band from the hyperspectral reflectivity data, and classifying the sensitive wavelength band according to the characteristic quantity of reflection peaks and the characteristic quantity of absorption valleys to obtain a peak sensitive set and a trough sensitive set;
the data acquisition module is used for acquiring hyperspectral image information of the target area and the nitrogen content of the corresponding cotton leaf;
the data decomposition module is used for dividing the hyperspectral image information into a plurality of mixed pixels, decomposing the mixed pixels by a mixed spectrum decomposition method, and classifying and integrating decomposition results to obtain a multispectral data set consisting of soil hyperspectral data, moisture hyperspectral data and blade hyperspectral data;
the data interception module is used for intercepting a first multi-hyperspectral feature group of corresponding wavelength from the multi-hyperspectral data group according to the crest sensitive set, and the first multi-hyperspectral feature group of the multiple wavelengths forms a first multi-hyperspectral feature discrete set of the same mixed pixel; and intercepting a second multi-hyperspectral feature set of corresponding wavelengths from the multi-hyperspectral data set according to the trough sensitive set, the second multi-hyperspectral feature sets of the multiple wavelengths forming a second multi-hyperspectral feature discrete set of the same hybrid pixel;
the model construction module is used for inputting a first multi-hyperspectral characteristic discrete set of a plurality of mixed pixels and the nitrogen content of cotton leaves into the linear regression training model for training to obtain a first nitrogen content monitoring model; inputting a second multi-hyperspectral characteristic discrete set of the plurality of mixed pixels and the nitrogen content of cotton leaves into a linear regression training model for training to obtain a second nitrogen content monitoring model;
the data acquisition module is used for acquiring hyperspectral data information of the target cotton leaf and randomly selecting an estimated wavelength and a multispectral characteristic estimation group under the corresponding wavelength from the hyperspectral data information;
the monitoring module is used for respectively inputting the estimated wavelength and the multispectral characteristic estimation module into the first nitrogen content monitoring model and the second nitrogen content monitoring model for matching, and preferentially selecting the monitoring model according to the matching result to carry out nitrogen content estimation monitoring on the target cotton leaf so as to obtain an estimation monitoring result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110258200.7A CN113008890B (en) | 2021-03-09 | 2021-03-09 | Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110258200.7A CN113008890B (en) | 2021-03-09 | 2021-03-09 | Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113008890A CN113008890A (en) | 2021-06-22 |
CN113008890B true CN113008890B (en) | 2023-11-03 |
Family
ID=76403578
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110258200.7A Active CN113008890B (en) | 2021-03-09 | 2021-03-09 | Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113008890B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107064053A (en) * | 2017-02-23 | 2017-08-18 | 浙江大学 | Method based on polyphenol content in near-infrared hyperspectral technique Non-Destructive Testing red bayberry |
CN109187441A (en) * | 2018-08-27 | 2019-01-11 | 中国水利水电科学研究院 | The construction method of summer corn nitrogen content monitoring model based on canopy spectrum information |
CN110160967A (en) * | 2019-04-16 | 2019-08-23 | 安徽大学 | A kind of total nitrogen content evaluation method of crop canopies blade |
CN111161362A (en) * | 2020-01-15 | 2020-05-15 | 昆山小茶智能科技有限公司 | Tea tree growth state spectral image identification method |
-
2021
- 2021-03-09 CN CN202110258200.7A patent/CN113008890B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107064053A (en) * | 2017-02-23 | 2017-08-18 | 浙江大学 | Method based on polyphenol content in near-infrared hyperspectral technique Non-Destructive Testing red bayberry |
CN109187441A (en) * | 2018-08-27 | 2019-01-11 | 中国水利水电科学研究院 | The construction method of summer corn nitrogen content monitoring model based on canopy spectrum information |
CN110160967A (en) * | 2019-04-16 | 2019-08-23 | 安徽大学 | A kind of total nitrogen content evaluation method of crop canopies blade |
CN111161362A (en) * | 2020-01-15 | 2020-05-15 | 昆山小茶智能科技有限公司 | Tea tree growth state spectral image identification method |
Non-Patent Citations (4)
Title |
---|
Estimation of Nitrogen Content on Apple Tree Canopy through Red-Edge Parameters from Fractional-Order Differential Operators using Hyperspectral Reflectance;Yufeng Peng et al.;《Journal of the Indian Society of Remote Sensing》;第49卷(第2期);377-392 * |
基于高光谱的梭梭幼苗生境土壤氮素含量估算模型研究;王泽;轩俊伟;任财;麻浩;罗艳丽;邓策;梁燕;范亚楠;;广东农业科学(第01期);68-76 * |
基于高光谱的甜菜冠层氮素遥感估算研究;李哲;农机化研究(第6期);210-214 * |
基于高光谱的苹果花氮素含量预测模型研究;朱西存;赵庚星;王凌;董芳;雷彤;战兵;;光谱学与光谱分析(第02期);416-420 * |
Also Published As
Publication number | Publication date |
---|---|
CN113008890A (en) | 2021-06-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Li et al. | Modern imaging techniques in plant nutrition analysis: A review | |
CA2773113C (en) | Computer-automated spatial analysis | |
CN111738066B (en) | Grid late rice sheath blight disease habitat evaluation method integrating multisource remote sensing information | |
Tits et al. | The potential and limitations of a clustering approach for the improved efficiency of multiple endmember spectral mixture analysis in plant production system monitoring | |
CN107122739A (en) | The agricultural output assessment model of VI time-serial positions is reconstructed based on Extreme mathematical modelings | |
CN110647932B (en) | Planting crop structure remote sensing image classification method and device | |
CN116227758B (en) | Agricultural product maturity prediction method and system based on remote sensing technology and deep learning | |
CN113657158A (en) | Google Earth Engine-based large-scale soybean planting region extraction algorithm | |
CN102156886A (en) | Statistical data and remote sensing image data based regional fertilizer application spatialization method | |
Lyu et al. | UAV time-series imagery with novel machine learning to estimate heading dates of rice accessions for breeding | |
Sahoo et al. | Drone remote sensing of wheat N using hyperspectral sensor and machine learning | |
Li et al. | Soybean leaf estimation based on RGB images and machine learning methods | |
CN113008890B (en) | Unmanned aerial vehicle hyperspectral-based cotton leaf nitrogen content monitoring method and system | |
Wang et al. | Construction of complex features for predicting soil total nitrogen content based on convolution operations | |
CN111751295A (en) | Modeling method and application of wheat powdery mildew severity detection model based on imaging hyperspectral data | |
CN116994126A (en) | Crop leaf area index obtaining method and device based on canopy reflectivity spectrum | |
CN115508292B (en) | Machine learning-based hyperspectral detection and visualization method for nitrogen content of soil profile | |
Jinglei et al. | Distance-based separability criterion of ROI in classification of farmland hyper-spectral images | |
Yerzhanova et al. | Analysis of the spectral properties of wheat growth in different vegetation periods | |
CN115266720A (en) | Multi-scale spectral imaging corn growth period nutrient information detection method | |
CN114663761A (en) | Crop growth condition determining method, device, equipment and storage medium | |
CN113836784A (en) | Apple identification system and method based on information fusion technology | |
CN113537604A (en) | Crop yield prediction method and device by coupling process model and deep learning | |
CN117173122B (en) | Lightweight ViT-based image leaf density determination method and device | |
Hu et al. | Convolutional Neural Network-Based Estimation of Nitrogen Content in Regenerating Rice Leaves |
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 |