CN115169728A - Soil fertility prediction method based on simplified neural network - Google Patents

Soil fertility prediction method based on simplified neural network Download PDF

Info

Publication number
CN115169728A
CN115169728A CN202210898792.3A CN202210898792A CN115169728A CN 115169728 A CN115169728 A CN 115169728A CN 202210898792 A CN202210898792 A CN 202210898792A CN 115169728 A CN115169728 A CN 115169728A
Authority
CN
China
Prior art keywords
model
neural network
organic matter
soil
matter content
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
Application number
CN202210898792.3A
Other languages
Chinese (zh)
Inventor
李明新
张树旺
程立亮
李绍俊
庞景秋
齐井春
陈兴钰
崔放
李忆平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changchun Jiacheng Information Technology Co ltd
Original Assignee
Changchun Jiacheng Information Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Changchun Jiacheng Information Technology Co ltd filed Critical Changchun Jiacheng Information Technology Co ltd
Priority to CN202210898792.3A priority Critical patent/CN115169728A/en
Publication of CN115169728A publication Critical patent/CN115169728A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/17—Terrestrial scenes taken from planes or by drones

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Multimedia (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Artificial Intelligence (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Software Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Game Theory and Decision Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • Medical Informatics (AREA)
  • Computational Linguistics (AREA)
  • Development Economics (AREA)
  • Biophysics (AREA)
  • Molecular Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Remote Sensing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a soil fertility prediction method based on a simplified neural network, which comprises the following prediction processes: constructing a simplified neural network model based on organic matter content; embedding the model into the drone; flying by an unmanned aerial vehicle close to the ground, and shooting an image of a position to be predicted; obtaining the prediction result of the organic matter content of the tested soil; determining the fertility information of the soil by combining the soil organic matter content prediction result calculated by the model with the improved Pearson coefficient; and (4) obtaining the final soil fertility grade of the position to be predicted by distinguishing the fertility grades. The method can acquire accurate soil fertility prediction information, is further beneficial to accurately judging the crops suitable for being planted in the land, can perform dynamic and accurate judgment on the soil fertility before sowing by agricultural workers so as to suggest and recommend the planted crops, and can realize the maximum utilization rate of the land.

Description

Soil fertility prediction method based on simplified neural network
Technical Field
The invention relates to a soil fertility prediction method, in particular to a soil fertility prediction method based on a simplified neural network.
Background
At present, most soil fertility prediction researches are carried out on-site sampling, a spectrometer is used for obtaining soil reflectivity indoors, the method eliminates the influence of vegetation and surface sediments, the accuracy is high, but the obtained spectral data information is punctiform, and the dynamic monitoring of the organic matter content in the area is difficult to realize. Meanwhile, the existing soil fertility research depends on different satellite images, but the satellite images have the problems of long revisit period, pixel mixing, weather condition and other factors, so that the requirements for accurately analyzing soil nutrient information cannot be met. Gradually, a low-altitude unmanned aerial vehicle mounted hyperspectral camera is developed for image acquisition, the mode is higher along with the cost, but the image acquired by the hyperspectral camera can easily acquire remote sensing data with high spatial resolution (centimeter level) and multiple wave bands (from visible light to near infrared), so that the balance of cost and usability is achieved; the neural network model can quickly and accurately obtain prediction information, and if the neural network model is combined with soil fertility prediction, more reliable support is provided for dynamic monitoring and accurate analysis of soil fertility.
Disclosure of Invention
In order to solve the defects of the technology, the invention provides a soil fertility prediction method based on a simplified neural network.
In order to solve the technical problems, the invention adopts the technical scheme that: a soil fertility prediction method based on a simplified neural network comprises the following prediction processes:
step one, constructing a simplified neural network model based on organic matter content;
step two, embedding the simplified neural network model into the unmanned aerial vehicle;
thirdly, flying by the unmanned aerial vehicle embedded with the simplified neural network in a ground-attached manner, and shooting an image of a position to be predicted;
step four, obtaining the prediction result of the organic matter content of the tested soil through a simplified neural network;
fifthly, determining the fertility information of the soil by combining the soil organic matter content prediction result calculated by the model with the improved Pearson coefficient;
and step six, obtaining the final soil fertility grade of the position to be predicted through the differentiation of the fertility grade.
Further, in the step one, the construction process of the simplified neural network model is as follows: the method comprises the steps of data acquisition, splitting of a training set and a testing set, definition of a neural network model, model training and model evaluation, and lightweight model tuning, so that a simplified neural network suitable for soil fertility prediction is obtained.
Further, in the process of model construction, the defined neural network model is a model of a CNN model and a support vector machine.
Further, in the model construction process, the CNN model uses a residual error network to optimize a convolution kernel, a support vector machine finds a maximum interval hyperplane, and then the SMO algorithm of the model is optimized to solve the dual problem of convex quadratic programming.
Furthermore, in the process of model construction, model accuracy comparison is carried out through Pearson correlation coefficient analysis, and model evaluation is achieved; pearson correlation coefficient analysis is shown below:
Figure BDA0003770076940000021
in the formula, X i Representing the organic matter content of the ith sample;
Figure BDA0003770076940000022
represents the average of organic matter of all samples; y is i Represents the spectral value of the ith sample,
Figure BDA0003770076940000023
represents the average of the spectral values of all samples.
Further, in the third step, when the unmanned aerial vehicle is used for shooting the position to be predicted, the speed of the unmanned aerial vehicle is controlled, so that the unmanned aerial vehicle can be correctly predicted by the simplified neural network model; and determining the ground flying speed of the unmanned aerial vehicle by calculating the approximate time predicted by the model.
Further, in step five, the accuracy of the prediction is judged according to the improved pearson coefficient, as shown in the following formula:
Figure BDA0003770076940000031
in the formula, ρ L，Q The accuracy of the prediction; l is the actual organic matter content, which is obtained in advance in the model construction stage; q is the organic matter content predicted by the model, and N is the sample number.
And further, in the sixth step, grading the organic matter content according to the organic matter condition in the farmland soil, wherein the grading is divided into three grades of excellent grade, good grade and medium grade, the medium grade is when the organic matter content is less than 6g/kg, the quantity is when the organic matter content is between 6 and 10g/kg, and the excellent grade is when the organic matter content is between 10 and 15 g/kg.
The invention discloses a soil fertility prediction method based on a simplified neural network, which can acquire accurate soil fertility prediction information, is further beneficial to accurately judging crops suitable for being planted in the soil, can perform dynamic and accurate judgment on the soil fertility before sowing by agricultural workers so as to suggest and recommend the planted crops, and can realize the maximum utilization rate of the soil.
Drawings
FIG. 1 is a schematic view of the soil fertility prediction process of the present invention.
Fig. 2 is a schematic view of a network structure of a simplified neural network of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and the detailed description.
The invention discloses a soil fertility prediction method based on a simplified neural network, which mainly comprises the following prediction steps as shown in figure 1:
step one, constructing a simplified neural network model based on organic matter content;
step two, embedding the simplified neural network model into the unmanned aerial vehicle;
thirdly, flying by the unmanned aerial vehicle embedded with the simplified neural network in a ground-attached manner, and shooting an image of a position to be predicted;
step four, obtaining the prediction result of the organic matter content of the tested soil through a simplified neural network;
fifthly, determining the fertility information of the soil by combining the soil organic matter content prediction result calculated by the model with the improved Pearson coefficient;
and step six, obtaining the final soil fertility grade of the position to be predicted through the differentiation of the fertility grade.
The construction process of the simplified neural network model can be summarized as follows: the method comprises the steps of data acquisition, splitting of a training set and a testing set, definition of a neural network model, model training and model evaluation, and lightweight model tuning, so that a simplified neural network suitable for soil fertility prediction is obtained.
Data required by model construction are mainly provided by existing actually-measured soil organic matter content data and photo information shot by an unmanned aerial vehicle; through cross validation, dividing data into n parts, and using one part as a test set and the other n-1 parts as a training set in sequence;
the neural network is a wide parallel interconnection network consisting of simple units with adaptability, and the organization of the neural network can simulate the interaction reaction of a biological nervous system to real-world objects; the invention adopts a CNN model combined with a support vector machine to define a neural network;
the support vector machine is a machine learning method based on statistical theory, and has strong mathematical basis and theoretical support. The support vector machine is based on the identification advantages of small samples, nonlinearity and high latitude modes of a VC theory and a structural risk minimization theory, the generalization capability of the model is improved by seeking for minimizing the structural risk, and the problem is finally converted into secondary optimization in the operation process, so that global optimization is realized, and a better statistical result is obtained under the condition of using fewer known samples.
The key of the support vector machine is the selection of a kernel function, the kernel function calculates an internal product in a low-dimensional space, the feature of the low-dimensional space is not directly mapped to a high-dimensional space, and the problem of linear inseparability in the low-dimensional space is solved while dimensionality disasters and subsequent processing complexity are avoided. According to functional theory, as long as one kernel Function meets the condition of Mercer, the kernel Function corresponds to an inner product in a certain transformation space, the kernel Function meeting the condition of Mercer comprises a polynomial Function, a linear Function, a Radial Basis Function (RBF), a Sigmoid Function and the like, and different types of support vector machines can be constructed based on different kernel functions.
The method mainly adopts a CNN model and a model of a support vector machine to predict, the CNN model uses a residual error network to optimize a convolution kernel, and the support vector machine mainly finds the farthest distance from various sample points to a hyperplane, namely finds the hyperplane with the largest interval, and then optimizes the SMO algorithm of the model to solve the dual problem of convex quadratic programming.
Then, model accuracy comparison is carried out through Pearson correlation coefficient analysis, and model evaluation is achieved; the method comprises the steps that Pearson correlation coefficient analysis (Pearson correlation coefficient) is carried out on the actually measured soil organic matter content of a sampling point and the spectrum reflectivity of a spectrum set, the Pearson correlation coefficient is a statistical method used for reflecting the correlation degree of two data variables, the r value is between-1 and 1, the larger the absolute value is, the higher the correlation between the two variables is, the smaller the absolute value is, and the smaller the correlation is; as shown in the following formula:
Figure BDA0003770076940000051
in the formula, X i Expressing the organic matter content of the ith sample;
Figure BDA0003770076940000052
represents the average value of organic matter of all samples; y is i Represents the spectral value of the ith sample,
Figure BDA0003770076940000053
represents the average of the spectral values of all samples.
The size of the model is mainly cut, and a heuristic algorithm of the model is optimized to obtain the simplified neural network model, as shown in fig. 2, in the figure, the input picture format is 32 × 32, and the simplified neural network model is sequentially cut into 28 × 28, 14 × 14, 10 × 10 and 5 × 5 pictures, stacked by 120 layers and stacked by 84 layers and output.
After obtaining the simplified neural network, embedding the simplified neural network into the unmanned aerial vehicle; the neural network operates in a code form, and the operating environment of the neural network is built based on the raspberry pi, so that the neural network is embedded in the unmanned aerial vehicle; the corresponding operation page mainly realizes a framing range for collecting images, after an instruction is sent, the raspberry group controls the hyperspectral camera to shoot the images in the framing range, the simplified neural network is used for predicting the soil organic matter content of the corresponding area of the shot images, and finally, the page can display the organic matter content predicted by the model.
Flying by the unmanned aerial vehicle embedded with the simplified neural network close to the ground to shoot an image of a position to be predicted; the method comprises the steps of determining specific position information of a predicted position by using a geographic positioning system on the unmanned aerial vehicle, and further acquiring specific land information (mainly comprising an area to be measured) and land distribution conditions to be predicted (mainly comprising judging whether measurement areas are all land and marking out buildings and water bodies).
When the unmanned aerial vehicle is used for shooting a position to be predicted, the speed of the unmanned aerial vehicle is controlled, so that the unmanned aerial vehicle can be correctly predicted by a simplified neural network model; and determining the ground-attached flight speed of the unmanned aerial vehicle by calculating the approximate time predicted by the model.
Further, inputting the shot prediction position image into a simplified neural network to obtain a prediction result of the organic matter content of the soil to be detected; and a support vector machine regression model constructed by first-order differential transformation is used for constructing an estimation model of the soil organic matter content, the soil fertility is researched and analyzed by detecting the organic matter content in the soil, and the soil organic matter content is obtained by calculation according to the spectral information of the image.
And determining the fertility information of the soil by combining the soil organic matter content prediction result calculated by the model with the improved pearson coefficient, and judging the prediction accuracy according to the pearson coefficient, wherein the prediction accuracy is shown as the following formula:
Figure BDA0003770076940000061
in the formula, ρ L，Q Is the accuracy of the prediction; l is the actual organic matter content, which is obtained in advance in the model construction stage; q is the organic matter content predicted by the model, and N is the sample number.
And finally, obtaining the soil fertility grade of the position to be predicted by distinguishing the fertility grades.
Whether the soil used for ploughing is suitable for the growth of crops is one of the key factors for determining the potential productivity of farmlands. In the process of improving farmland soil, reasonably arranging the farmland soil as a planting type, a management measure and the like, the remote sensing technology can effectively monitor soil fertility conditions and obtain soil fertility charts, and corresponding basis is provided for farmland management. The soil fertility monitoring comprises the soil particle size, the soil texture, the clay content and the like, influences the soil moisture conservation and nutrient substance migration of soil structural parameters, and can be applied to evaluating the utilization of fertilizers and the irrigation and drainage capability of soil.
The method comprises the steps of grading the organic matter content according to the organic matter condition of the farmland soil, dividing the organic matter content of the soil into a good grade, a good grade and a medium grade through a threshold value, specifically dividing the threshold value into three grades of less than 6g/kg,6-10g/kg and 10-15g/kg, effectively dividing and classifying the soil fertility through the division of the three grades, and further dividing the soil fertility into the good grade, the good grade and the medium grade.
The organic matter content of the soil is measured by heating potassium dichromate through an oil bath; under the condition of heating, oxidizing soil organic carbon by using excessive potassium dichromate-sulfuric acid solution, titrating redundant potassium dichromate by using ferrous sulfate solution, calculating the amount of the organic carbon by using the consumed potassium dichromate amount according to an oxidation correction coefficient, and multiplying the calculated amount by a conversion coefficient 1.724 obtained according to a related reaction chemical formula to obtain the content of soil organic matters.
Figure BDA0003770076940000071
Wherein, M is the molarity of the standard ferric salt solution, vo is the volume (ml) of the low ferric salt solution used for blank titration, V is the volume (ml) of the low ferric salt solution used for sample titration, W is the dried soil weight = the water content coefficient, and 1.724 is the equivalent organic matter gram per gram of carbon.
Therefore, for the soil fertility prediction method based on the simplified neural network disclosed by the invention, the soil fertility prediction is carried out based on the simplified neural network model and the unmanned aerial vehicle, the improved Pearson coefficient can specifically analyze the soil fertility prediction accuracy, and the accurate soil fertility prediction information can be obtained, so that the method is helpful for accurately judging the crops suitable for being planted in the soil; by the soil fertility prediction method, dynamic and accurate judgment of soil fertility can be performed before agricultural workers sow the soil so as to suggest and recommend planted crops, the maximum utilization rate of the soil can be realized, meanwhile, the soil fertility information can be rapidly and accurately acquired, and advance prediction of soil change can be guaranteed.
The above embodiments are not intended to limit the present invention, and the present invention is not limited to the above examples, and those skilled in the art may make variations, modifications, additions or substitutions within the technical scope of the present invention.

Claims (8)

1. A soil fertility prediction method based on a simplified neural network is characterized by comprising the following steps: the method comprises the following prediction processes:
step one, constructing a simplified neural network model based on organic matter content;
step two, embedding the simplified neural network model into the unmanned aerial vehicle;
thirdly, flying by an unmanned aerial vehicle embedded with a simplified neural network in a ground-contacting manner, and shooting an image of a position to be predicted;
step four, obtaining the prediction result of the organic matter content of the tested soil through a simplified neural network;
fifthly, determining the fertility information of the soil by combining the soil organic matter content prediction result calculated by the model with the improved Pearson coefficient;
and step six, obtaining the final soil fertility grade of the position to be predicted through the differentiation of the fertility grade.
2. The soil fertility prediction method based on the simplified neural network as claimed in claim 1, wherein: in the first step, the construction process of the simplified neural network model comprises the following steps: the method comprises the steps of data acquisition, splitting of a training set and a testing set, definition of a neural network model, model training and model evaluation, and lightweight model tuning, so that a simplified neural network suitable for soil fertility prediction is obtained.
3. The soil fertility prediction method based on the simplified neural network as claimed in claim 2, wherein: in the process of constructing the model, the defined neural network model is a CNN model and a model of a support vector machine.
4. The soil fertility prediction method based on the simplified neural network as claimed in claim 3, wherein: in the model construction process, the CNN model uses a residual error network to optimize a convolution kernel, a support vector machine finds a maximum interval hyperplane, and then the SMO algorithm of the model is optimized to solve the dual problem of convex quadratic programming.
5. The soil fertility prediction method based on the simplified neural network as claimed in claim 4, wherein: in the process of model construction, model accuracy comparison is carried out through Pearson correlation coefficient analysis, and model evaluation is achieved; pearson correlation coefficient analysis is shown below:
Figure FDA0003770076930000021
in the formula, X i Representing the organic matter content of the ith sample;
Figure FDA0003770076930000022
represents the average of organic matter of all samples; y is i Represents the spectral value of the ith sample,
Figure FDA0003770076930000023
represents the average of the spectral values of all samples.
6. The soil fertility prediction method based on the simplified neural network as claimed in claim 5, wherein: in the third step, when the unmanned aerial vehicle is used for shooting the position to be predicted, the speed of the unmanned aerial vehicle is controlled, so that the unmanned aerial vehicle can be correctly predicted by the simplified neural network model; and determining the ground flying speed of the unmanned aerial vehicle by calculating the approximate time predicted by the model.
7. The soil fertility prediction method based on the simplified neural network as claimed in claim 6, wherein: in the fifth step, the accuracy of prediction is judged according to the improved Pearson coefficient, which is shown as the following formula:
Figure FDA0003770076930000024
in the formula, ρ L，Q The accuracy of the prediction; l is the actual organic matter content, which is obtained in advance in the model construction stage; q is the organic matter content predicted by the model, and N is the sample number.
8. The soil fertility prediction method based on the simplified neural network as claimed in claim 7, wherein: and step six, grading the organic matter content according to the organic matter condition in the cultivated land soil, and dividing the organic matter content into three grades, namely a good grade, a medium grade and a medium grade, wherein the organic matter content is less than 6g/kg, the organic matter content is equal to the organic matter content between 6 and 10g/kg, and the organic matter content is good when the organic matter content is between 10 and 15 g/kg.
CN202210898792.3A 2022-07-28 2022-07-28 Soil fertility prediction method based on simplified neural network Pending CN115169728A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210898792.3A CN115169728A (en) 2022-07-28 2022-07-28 Soil fertility prediction method based on simplified neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210898792.3A CN115169728A (en) 2022-07-28 2022-07-28 Soil fertility prediction method based on simplified neural network

Publications (1)

Publication Number Publication Date
CN115169728A true CN115169728A (en) 2022-10-11

Family

ID=83478266

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210898792.3A Pending CN115169728A (en) 2022-07-28 2022-07-28 Soil fertility prediction method based on simplified neural network

Country Status (1)

Country Link
CN (1) CN115169728A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117114627A (en) * 2023-10-18 2023-11-24 日照市自然资源和规划局 land resource management system
CN117158173A (en) * 2023-10-10 2023-12-05 常熟市农业科技发展有限公司 Agricultural high-efficiency fertilization system and fertilization method based on neural network model

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117158173A (en) * 2023-10-10 2023-12-05 常熟市农业科技发展有限公司 Agricultural high-efficiency fertilization system and fertilization method based on neural network model
CN117114627A (en) * 2023-10-18 2023-11-24 日照市自然资源和规划局 land resource management system

Similar Documents

Publication Publication Date Title
CN110174359B (en) Aviation hyperspectral image soil heavy metal concentration assessment method based on Gaussian process regression
Whetton et al. Nonlinear parametric modelling to study how soil properties affect crop yields and NDVI
CN115169728A (en) Soil fertility prediction method based on simplified neural network
CN109508824A (en) A kind of detection of crop growth situation and yield predictor method
Natarajan et al. Prediction of soil pH from remote sensing data using gradient boosted regression analysis
CN114529097B (en) Multi-scale crop phenological period remote sensing dimensionality reduction prediction method
CN115860269B (en) Crop yield prediction method based on triple attention mechanism
CN115759524B (en) Soil productivity grade identification method based on remote sensing image vegetation index
CN113343783A (en) Intelligent crop identification and growth prediction method and system
CN114813651A (en) Remote sensing water quality inversion method combining difference learning rate and spectrum geometric characteristics
CN108764527B (en) Screening method for soil organic carbon library time-space dynamic prediction optimal environment variables
CN116148188A (en) Air-space-ground integrated lake water quality tracing method, system, equipment and storage medium
Bai et al. Optimal window size selection for spectral information extraction of sampling points from UAV multispectral images for soil moisture content inversion
CN117935081B (en) Cultivated land change monitoring method and system based on remote sensing satellite data
Aziz et al. Determine the Ph of soil by using neural network based on soil’s colour
Wang et al. Application of hyperspectral technology combined with bat algorithm-AdaBoost model in field soil nutrient prediction
Zhai et al. CatBoost algorithm for estimating maize above-ground biomass using unmanned aerial vehicle-based multi-source sensor data and SPAD values
Lou et al. Recognition of thaw slumps based on machine learning and UAVs: A case study in the Qilian Mountains, northeastern Qinghai-Tibet Plateau
CN116403048B (en) Crop growth estimation model construction method based on multi-mode data fusion
CN110321528B (en) Hyperspectral image soil heavy metal concentration assessment method based on semi-supervised geospatial regression analysis
Feng et al. Prediction of cotton yield based on soil texture, weather conditions and UAV imagery using deep learning
WO2023131949A1 (en) A versatile crop yield estimator
Li et al. Early drought plant stress detection with bi-directional long-term memory networks
Amankulova et al. Integrating the Sentinel-1, Sentinel-2 and topographic data into soybean yield modelling using machine learning
Devaser et al. An approach to analyse the agriculture acreage and estimate production

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