CN116918546A - Corn high-yield planting and fertilizing method - Google Patents
Corn high-yield planting and fertilizing method Download PDFInfo
- Publication number
- CN116918546A CN116918546A CN202310840449.8A CN202310840449A CN116918546A CN 116918546 A CN116918546 A CN 116918546A CN 202310840449 A CN202310840449 A CN 202310840449A CN 116918546 A CN116918546 A CN 116918546A
- Authority
- CN
- China
- Prior art keywords
- soil fertility
- growth state
- corn
- training
- feature vector
- 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
- 240000008042 Zea mays Species 0.000 title claims abstract description 54
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 title claims abstract description 53
- 235000002017 Zea mays subsp mays Nutrition 0.000 title claims abstract description 53
- 235000005822 corn Nutrition 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000035558 fertility Effects 0.000 claims abstract description 125
- 239000002689 soil Substances 0.000 claims abstract description 106
- 230000012010 growth Effects 0.000 claims abstract description 92
- 241001057636 Dracaena deremensis Species 0.000 claims abstract description 56
- 239000003337 fertilizer Substances 0.000 claims abstract description 30
- 239000013598 vector Substances 0.000 claims description 120
- 238000012549 training Methods 0.000 claims description 63
- 238000013527 convolutional neural network Methods 0.000 claims description 37
- 230000008635 plant growth Effects 0.000 claims description 33
- 239000011159 matrix material Substances 0.000 claims description 24
- 238000009826 distribution Methods 0.000 claims description 17
- 238000005457 optimization Methods 0.000 claims description 12
- 238000012546 transfer Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 239000002356 single layer Substances 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 25
- 238000010586 diagram Methods 0.000 description 10
- 239000010410 layer Substances 0.000 description 10
- 241000196324 Embryophyta Species 0.000 description 9
- 230000004720 fertilization Effects 0.000 description 8
- 238000007726 management method Methods 0.000 description 6
- 235000015097 nutrients Nutrition 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 4
- 241000607479 Yersinia pestis Species 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- 241000238631 Hexapoda Species 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 208000002720 Malnutrition Diseases 0.000 description 1
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 1
- ZLMJMSJWJFRBEC-UHFFFAOYSA-N Potassium Chemical compound [K] ZLMJMSJWJFRBEC-UHFFFAOYSA-N 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 235000013339 cereals Nutrition 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000003967 crop rotation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000003973 irrigation Methods 0.000 description 1
- 230000002262 irrigation Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000001071 malnutrition Effects 0.000 description 1
- 235000000824 malnutrition Nutrition 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 235000021049 nutrient content Nutrition 0.000 description 1
- 208000015380 nutritional deficiency disease Diseases 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 229910052698 phosphorus Inorganic materials 0.000 description 1
- 239000011574 phosphorus Substances 0.000 description 1
- 230000001863 plant nutrition Effects 0.000 description 1
- 239000011591 potassium Substances 0.000 description 1
- 229910052700 potassium Inorganic materials 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000004383 yellowing Methods 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01C—PLANTING; SOWING; FERTILISING
- A01C21/00—Methods of fertilising, sowing or planting
- A01C21/007—Determining fertilization requirements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
- G06V10/765—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects using rules for classification or partitioning the feature space
-
- 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
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Soil Sciences (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Environmental Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Discloses a high-yield corn planting and fertilizing method. Firstly, acquiring a growth state image and a soil fertility state image of a monitored corn plant, then extracting growth state characteristic information in the growth state image of the monitored corn plant, then extracting fertility state characteristic information in the soil fertility state image, and finally, generating a recommended fertilizer based on the growth state characteristic information and the fertility state characteristic information. Thus, the type of the suitable fertilizer can be intelligently recommended based on an artificial intelligence algorithm, and the yield and quality of corn are improved to the greatest extent.
Description
Technical Field
The present disclosure relates to the field of smart agriculture, and more particularly, to a corn high-yield planting and fertilizing method.
Background
Corn is an important grain crop whose yield and quality are affected by fertilization management. The traditional fertilization management method relies on manual observation and experience judgment, and has the defects of strong subjectivity, low efficiency, poor precision and the like. The development of technologies such as the Internet of things, artificial intelligence and intelligent control brings revolutionary changes to the agricultural planting industry, and intelligent agriculture follows.
The intelligent agriculture is an agriculture mode for improving the agricultural production efficiency and sustainability by utilizing advanced technology and innovative method, combines technologies such as Internet of things, big data analysis, artificial intelligence, machine learning and the like, and tools such as sensors, automation equipment, unmanned aerial vehicles and the like, and provides accurate agricultural management and decision support for farmers.
Therefore, a corn high-yield planting and fertilizing scheme in a smart agriculture background is expected.
Disclosure of Invention
In view of this, the present disclosure proposes a corn high-yield planting and fertilizing method that can collect growth state information of a monitored corn plant and soil fertility state information of planting the monitored corn plant using a sensor and intelligently recommend a suitable fertilizer type based on an artificial intelligence algorithm to maximize yield and quality of corn.
According to an aspect of the present disclosure, there is provided a corn high-yield planting and fertilizing method, including:
acquiring a growth state image and a soil fertility state image of a monitored corn plant;
extracting growth state characteristic information in the growth state image of the monitored corn plant;
extracting fertility state characteristic information in the soil fertility state image; and
and generating recommended fertilizer application based on the growth state characteristic information and the fertility state characteristic information.
According to an embodiment of the present disclosure, a growth state image and a soil fertility state image of a monitored corn plant are first acquired, then growth state characteristic information in the growth state image of the monitored corn plant is extracted, then fertility state characteristic information in the soil fertility state image is extracted, and finally, a recommended application fertilizer is generated based on the growth state characteristic information and the fertility state characteristic information. Thus, the type of the suitable fertilizer can be intelligently recommended based on an artificial intelligence algorithm, and the yield and quality of corn are improved to the greatest extent.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 shows a flow chart of a corn high yield planting and fertilizing method according to an embodiment of the present disclosure.
Fig. 2 shows a schematic architecture diagram of a corn high yield planting and fertilizing method according to an embodiment of the present disclosure.
Fig. 3 shows a flowchart of sub-step S110 of a corn high yield planting and fertilizing method according to an embodiment of the present disclosure.
Fig. 4 shows a flowchart of sub-step S140 of the corn high yield planting fertilizing method according to an embodiment of the present disclosure.
Fig. 5 shows a flowchart of training steps further included in a corn high yield planting and fertilizing method according to an embodiment of the present disclosure.
Fig. 6 illustrates a block diagram of a corn high yield planting and fertilizing system according to an embodiment of the present disclosure.
Fig. 7 illustrates an application scenario diagram of a corn high-yield planting and fertilizing method according to an embodiment of the present disclosure.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, which can be made by one of ordinary skill in the art without undue burden based on the embodiments of the present disclosure, are also within the scope of the present disclosure.
As used in this disclosure and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following detailed description in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order not to obscure the present disclosure.
Aiming at the technical problems, the technical concept of the disclosure is to collect the growth state information of the monitored corn plants and the soil fertility state information of the monitored corn plants by using the sensors under the background of intelligent agriculture, and intelligently recommend the types of proper fertilizers based on an artificial intelligence algorithm so as to improve the yield and quality of corn to the greatest extent.
Fig. 1 shows a flow chart of a corn high yield planting and fertilizing method according to an embodiment of the present disclosure. Fig. 2 shows a schematic architecture diagram of a corn high yield planting and fertilizing method according to an embodiment of the present disclosure. As shown in fig. 1 and 2, the corn high-yield planting and fertilizing method according to the embodiment of the present disclosure includes the steps of: s110, acquiring a growth state image and a soil fertility state image of a monitored corn plant; s120, extracting growth state characteristic information in the growth state image of the monitored corn plant; s130, extracting fertility state characteristic information in the soil fertility state image; and S140, generating recommended fertilizer based on the growth state characteristic information and the fertility state characteristic information.
In particular, the growth status and soil fertility status information of corn plants can provide important clues as to the nutrient elements required by the corn plants and the nutrient content in the soil. For example, if a maize plant exhibits symptoms of malnutrition, such as yellowing of leaves, slow growth, etc., it may be inferred that the plant may lack a particular nutrient element, such as nitrogen, phosphorus, or potassium. The soil fertility status information can provide information about the content and proportion of various nutrients in the soil, further guiding the formulation of fertilization schemes. That is, the growth status of the corn plant may reflect the most needed nutrient elements of the current corn plant, while the soil fertility status may reflect whether the nutrients of the soil are sufficient for the current corn plant growth, in such a way that the selection of fertilizer is made intelligently.
Based on this, in the technical scheme of the present disclosure, first, a growth state image of a monitored corn plant collected by a first camera and a soil fertility state image for planting the monitored corn plant collected by a second camera are acquired. Accordingly, as shown in fig. 3, acquiring a growth state image and a soil fertility state image of a monitored corn plant includes: s111, receiving a growth state image of the monitored corn plant acquired by a first camera; and S112, receiving the soil fertility status image collected by the second camera and used for planting the monitored corn plants. It should be appreciated that the first camera is used to capture a growth status image of the monitored corn plant, while the second camera is used to capture a soil fertility status image for planting the monitored corn plant, the following factors are considered when selecting the camera: 1. image definition, in order to obtain accurate growth state and soil fertility state images, the camera should have high definition, and can capture details and fine changes; 2. color reducibility, the camera can accurately restore the color in the image so as to accurately analyze the growth state and the soil fertility state of plants; 3. the illumination adaptability, because the illumination condition of the outdoor environment can be changed continuously, the camera has good illumination adaptability so as to ensure that clear images can be obtained under different illumination conditions; 4. the durability, considering that the camera is to be installed in a field environment, should have durability and dustproof and waterproof functions to ensure long-term stable operation. For example, a high definition camera, a color camera, and an industrial-grade camera may be used as the first camera and the second camera.
And then, passing the growth state image of the monitored corn plant through a growth state feature extractor based on a first convolutional neural network model to obtain a plant growth state feature vector, and passing the soil fertility state image through a soil fertility state feature extractor based on a second convolutional neural network model to obtain a soil fertility state feature vector. That is, the growth state feature extractor and the soil fertility state feature extractor are constructed using a first convolutional neural network model and a second convolutional neural network model, respectively, to capture implicit correlation patterns in the growth state image and the soil fertility state image of the monitored corn plant.
Correspondingly, extracting growth state characteristic information in the growth state image of the monitored corn plant comprises the following steps: and extracting plant growth state feature vectors from the growth state image of the monitored corn plant as the growth state feature information by using a growth state feature extractor based on a first convolutional neural network model. Further, extracting plant growth state feature vectors from the growth state image of the monitored corn plant as the growth state feature information using a growth state feature extractor based on a first convolutional neural network model, comprising: and performing convolution kernel-based feature filtering on the growth state image of the monitored corn plant by using the growth state feature extractor based on the first convolution neural network model to obtain the plant growth state feature vector. It should be appreciated that the first convolutional neural network model is a deep learning model for extracting growth state feature information in a corn plant growth state image. A convolutional neural network (Convolutional Neural Network, CNN) is a neural network structure that is dedicated to processing image data. In the corn high-yield planting and fertilizing method, a first convolution neural network model is used as a growth state feature extractor, and plant growth state feature vectors are extracted by carrying out convolution operation and feature filtering on the growth state images of the monitored corn plants. These feature vectors can capture implicit correlation patterns in the growth state image, such as growth rate of plants, color and shape of leaves, etc. Such characteristic information is useful for assessing the health and growth status of plants and can be used to generate recommended fertilization protocols. By using the convolutional neural network model, features in the image data can be automatically learned and extracted, and the complex process of manually designing the feature extractor is avoided.
Correspondingly, extracting the fertility status characteristic information in the soil fertility status image comprises the following steps: and extracting a soil fertility state feature vector from the soil fertility state image as the fertility state feature information by using a soil fertility state feature extractor based on a second convolutional neural network model. Further, extracting a soil fertility state feature vector from the soil fertility state image as the fertility state feature information using a soil fertility state feature extractor based on a second convolutional neural network model, comprising: and performing feature filtering on the soil fertility state image based on a convolution kernel by using the soil fertility state feature extractor based on the second convolution neural network model so as to obtain the soil fertility state feature vector. It should be understood that the second convolutional neural network model is a deep learning model for extracting fertility status characteristic information in the soil fertility status image, and is also a neural network structure dedicated to processing image data, similar to the first convolutional neural network model. In the corn high-yield planting and fertilizing method, a second convolution neural network model is used as a soil fertility state feature extractor, and the soil fertility state feature vector is extracted by carrying out convolution operation and feature filtering on the soil fertility state image. These feature vectors may capture fertility conditions in the soil fertility status image, such as soil texture, fertilizer distribution, etc. Such characteristic information is useful for assessing soil fertility and proper fertilization levels, and can be used to generate recommended fertilization protocols.
Then, a type label recommending fertilizer application is generated. In one specific example of the present disclosure, a specific implementation of generating a type tag that recommends application of fertilizer is: fusing the plant growth state feature vector and the soil fertility state feature vector to obtain a classification feature vector; and then the classification feature vector passes through a classifier to obtain a classification result, wherein the classification result is used for indicating the type label of the recommended fertilizer application.
Accordingly, as shown in fig. 4, generating a recommended fertilizer based on the growth state characteristic information and the fertility state characteristic information includes: s141, fusing the plant growth state feature vector and the soil fertility state feature vector to obtain a classification feature vector; and S142, passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used for representing the type label of the recommended fertilizer application. It should be appreciated that fusing plant growth state feature vectors with soil fertility state feature vectors can yield a comprehensive classification feature vector for more accurately describing plant growth and soil fertility. This classification feature vector may be used to: the fertilization recommendation can obtain a feature vector for more comprehensively describing plant nutrition requirements by fusing the feature vector of the plant growth state and the soil fertility state, and based on the comprehensive feature vector, intelligent fertilization recommendation can be carried out to determine proper fertilizer types and application amounts so as to meet the plant growth requirements; the plant health information can be provided by comprehensively considering the characteristic vectors of the plant growth state and the soil fertility state, and can be used for predicting and monitoring the possible plant diseases and insect pests threats, and by timely taking measures, the occurrence of the plant diseases and insect pests can be effectively prevented and controlled, and the health and the yield of the plant are protected; the farmland management decision, the characteristic vector of the plant growth state and the soil fertility state is synthesized, a decision basis of farmland management can be provided, for example, proper irrigation strategies, pest control measures and crop rotation plans can be determined according to the growth condition of plants and the fertility condition of soil, so that the production benefit and sustainable development of farmlands are optimized. The feature vector fusing the plant growth state and the soil fertility state can provide more comprehensive and accurate farmland information, help to make scientific decisions and management, and improve the production benefit and the sustainability of farmlands.
More specifically, in an embodiment of the present disclosure, fusing the plant growth state feature vector and the soil fertility state feature vector to obtain a classification feature vector includes: fusing the plant growth state feature vector and the soil fertility state feature vector using a cascading function to obtain the classification feature vector; wherein the cascading function is:
f(X i ,X j )＝Relu(W f [θ(X i ),φ(X j )])
wherein W is f ,θ(X i ) And phi (X) j ) All representing the point convolution of the input, relu as the activation function, []Representing the splicing operation, X i X is the characteristic value of each position in the plant growth state characteristic vector j And the characteristic value of each position in the characteristic vector of the soil fertility state is the characteristic value of each position in the characteristic vector of the soil fertility state. In particular, the application of the cascading function can enable the network model to have certain logic reasoning capability, and the association relation between the plant growth state characteristic vector and the soil fertility state characteristic vector is mined.
More specifically, the classification feature vector is further passed through a classifier to obtain a classification result, wherein the classification result is used for indicating the type label of the recommended fertilizer application, and the classification result comprises: performing full-connection coding on the classification feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
It should be appreciated that the role of the classifier is to learn the classification rules and classifier using a given class, known training data, and then classify (or predict) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
It should be noted that the fully connected layer (Fully Connected Layer) is a common hierarchical structure in the neural network, also called a dense connected layer or fully connected layer, and is used for multiplying the input feature vector with the weight matrix, adding the bias vector, and then performing nonlinear transformation through the activation function to generate the output vector, where each neuron of the fully connected layer is connected with all neurons of the previous layer, so that a complex relationship between the input features can be captured. Full-join encoding (Fully Connected Encoding) refers to a process of encoding a classification feature vector through a full-join layer, in which the classification feature vector is multiplied by a weight matrix of the full-join layer, and added with a bias vector, and then nonlinear transformation is performed through an activation function to generate an encoded classification feature vector, where the full-join encoding can map the original classification feature vector to a higher-dimensional feature space to better represent the relationship between features. The Softmax classification function is a commonly used classification function for mapping an input vector onto a probability distribution, the Softmax function can convert each element of the input vector into a probability value between 0 and 1, and the sum of all probability values is equal to 1, in a classification task, the Softmax function is commonly used for converting a feature vector after full-connection coding into an output vector representing different class probabilities, and the probability that the input vector belongs to different classes can be determined by performing probability interpretation on the output vector, so that a classification result is obtained. In other words, the full-connection layer is used for carrying out full-connection coding on the classified feature vectors, and mapping the classified feature vectors into a higher-dimensional feature space; and generating output vectors representing different category probabilities by the feature vectors after full-connection coding through a Softmax classification function, so as to obtain a classification result. These steps are used to determine the type label of the recommended fertilizer during the recommended application of the fertilizer.
Further, the corn high-yield planting and fertilizing method further comprises the training steps of: and training the growth state feature extractor based on the first convolutional neural network model, the soil fertility state feature extractor based on the second convolutional neural network model and the classifier. It will be appreciated that the training step has a role in the corn high yield planting and fertilizing method by using known training data to optimize and adjust the parameters of the convolutional neural network model and classifier so that they can better adapt to the actual growth state and soil fertility state images and accurately classify. Specifically, the training step functions include: the optimized feature extractor can learn more effective feature representation methods by training the growth state feature extractor based on the first convolutional neural network model and the soil fertility state feature extractor based on the second convolutional neural network model, so that the optimized feature extractor can extract more informative features from the growth state image and the soil fertility state image to help the classifier to classify better; the classifier is optimized, by training the classifier, how to accurately classify different growth states and soil fertility states according to the extracted features, the training process of the classifier usually involves optimizing a loss function, the classification result of the classifier on training data is consistent with a real label as much as possible, and by the optimizing process, the classifier can learn how to correlate the features with corresponding fertilizer types, so that accurate classification of recommended application fertilizers is realized. Through the training step, the performance and accuracy of a convolutional neural network model and a classifier used in the corn high-yield planting and fertilizing method can be improved, and the trained model and classifier can be better adapted to different growth states and soil fertility states, so that a more accurate and reliable recommended fertilizing strategy is provided, and the yield and quality of corn are improved.
In one specific example, as shown in fig. 5, the training step includes: s210, training data is obtained, wherein the training data comprises training growth state images of monitored corn plants, training soil fertility state images for planting the monitored corn plants, and true values of type tags for recommending fertilizer application; s220, passing the training growth state image of the monitored corn plant through the growth state feature extractor based on the first convolutional neural network model to obtain a training plant growth state feature vector; s230, passing the training soil fertility state image through the soil fertility state feature extractor based on the second convolutional neural network model to obtain a training soil fertility state feature vector; s240, fusing the training plant growth state feature vector and the training soil fertility state feature vector by using a cascading function to obtain a training classification feature vector; s250, the training classification feature vectors pass through a classifier to obtain a classification loss function value; and S260, training the growth state feature extractor based on the first convolutional neural network model, the soil fertility state feature extractor based on the second convolutional neural network model and the classifier according to the classification loss function value, wherein in each round of iteration of training, feature transfer optimization iteration based on feature distribution cross-domain attention is performed on a weight matrix of the classifier.
In the technical scheme of the disclosure, considering that the training classification feature vector is obtained by fusing the training plant growth state feature vector and the training soil fertility state feature vector by using a cascading function, and the training plant growth state feature vector and the training soil fertility state feature vector express the image semantic features of the growth state image and the soil fertility state image of the monitored corn plant respectively, the training classification feature vector still has distribution diversity corresponding to the image semantic features of the growth state image and the soil fertility state image of the monitored corn plant when cascading by using the cascading function.
In this way, when the training classification feature vector is classified by the classifier, the weight matrix of the classifier needs to be adaptively optimized for the training classification feature vector in consideration of the distribution transferability difference of the diversified feature distribution in the domain transfer process of classification, so as to improve the training effect of the training classification feature vector for classification training by the classifier, namely, improve the classification speed and the accuracy of the obtained classification result. Therefore, the applicant of the present disclosure performs feature transfer optimization based on feature distribution cross-domain attention on the weight matrix M in the iterative process of the weight matrix of each classifier.
Correspondingly, in each round of iteration of the training, performing feature transfer optimization iteration based on feature distribution cross-domain attention on the weight matrix of the classifier according to the following optimization formula; wherein, the optimization formula is:
wherein M is the weight matrix of the classifier, and the scale of M is M multiplied by M, V 1 To V m Is the M row vectors of the weight matrix M, I.I 2 Representing the two norms of the feature vector (Σ) j m i,j Is a row vector obtained by arranging the summation value of each row vector of the weight matrix M, and cov 1 (. Cndot.) and cov 2 (. Cndot.) all represent a single-layer convolution operation,representing matrix multiplication (.) T Representing the transpose operation, M' represents the weight matrix of the classifier after iteration.
Here, the feature transfer optimization based on the attention of feature distribution cross-domain and the like is based on the different representations of feature distribution of the training classification feature vector in a feature space domain and a classification target domain, and based on the cross-domain diversity feature representation of a weight matrix M of the classifier relative to the training classification feature vector to be classified, the transferability of cross-domain gaps of good transfer feature distribution in the diversity feature distribution is enhanced by giving attention to spatial structured feature distribution of the weight matrix M through convolution operation, and meanwhile, negative transfer (negative transfer) of bad transfer feature distribution is restrained, so that unsupervised domain transfer adaptive optimization of the weight matrix M is realized based on the distribution structure of the weight matrix M relative to the training classification feature vector, and the training effect of classification training of the training classification feature vector through the classifier is improved.
It is worth mentioning that the two norms of a feature vector refer to the square root of the sum of the squares of each element of the vector. Mathematically, given an n-dimensional vector x= (x 1, x2,..once., xn), its two-norm (also known as euclidean norm) is calculated as follows:
||x||2＝sqrt(x1^2+x2^2+...+xn^2)
where sqrt represents a square root operation. The two norms can be used for measuring the length or the size of the vector, and can also be used for regularization, similarity measurement, clustering and other tasks of the vector. In the field of machine learning and data analysis, a two-norm is often used as a regularization term for optimization algorithms to control the complexity of the model and prevent overfitting.
It is worth mentioning that single-layer convolution operation refers to a basic operation in a convolutional neural network for extracting features from input data, which typically consists of a filter (also called a convolution kernel) and an input data tensor. In a single layer convolution operation, the filter is a small matrix of weights that extracts features by sliding window operations on the input data. Specifically, each element of the filter is multiplied by an element of a corresponding position of the input data, and then all the multiplication results are added to obtain a single output value. By sliding the filters over the input data, an output profile can be obtained in which each element corresponds to the result of a convolution operation of one filter with the input data. By stacking and combining multiple single-layer convolution operations, a more complex convolutional neural network model can be constructed, thereby achieving higher level feature extraction and pattern recognition capabilities.
In summary, according to the corn high-yield planting and fertilizing method disclosed by the embodiment of the disclosure, the growth state information of the monitored corn plants and the soil fertility state information of the monitored corn plants can be collected by using the sensor, and the type of the proper fertilizer is intelligently recommended based on an artificial intelligence algorithm so as to improve the yield and quality of corn to the greatest extent.
Fig. 6 shows a block diagram of a corn high yield planting and fertilizing system 100 in accordance with an embodiment of the present disclosure. As shown in fig. 6, a corn high yield planting and fertilizing system 100 according to an embodiment of the present disclosure includes: an image acquisition module 110 for acquiring a growth status image and a soil fertility status image of the monitored corn plants; a growth state feature information extraction module 120 for extracting growth state feature information in a growth state image of the monitored corn plant; a fertility status feature information extraction module 130, configured to extract fertility status feature information in the soil fertility status image; and a recommended-application fertilizer generation module 140 for generating a recommended-application fertilizer based on the growth status characteristic information and the fertility status characteristic information.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described corn high-yield planting and fertilizing system 100 have been described in detail in the above description of the corn high-yield planting and fertilizing method with reference to fig. 1 to 5, and thus, repetitive descriptions thereof will be omitted.
As described above, the corn high-yield planting and fertilizing system 100 according to the embodiment of the present disclosure may be implemented in various wireless terminals, such as a server or the like having a corn high-yield planting and fertilizing algorithm. In one possible implementation, corn high yield planting and fertilizing system 100 according to embodiments of the present disclosure may be integrated into a wireless terminal as one software module and/or hardware module. For example, the corn high yield planting and fertilizing system 100 can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the corn high yield planting and fertilizing system 100 can also be one of many hardware modules of the wireless terminal.
Alternatively, in another example, the corn high yield planting and fertilizing system 100 and the wireless terminal may be separate devices, and the corn high yield planting and fertilizing system 100 may be connected to the wireless terminal through a wired and/or wireless network and transmit interactive information in accordance with a agreed data format.
Fig. 7 illustrates an application scenario diagram of a corn high-yield planting and fertilizing method according to an embodiment of the present disclosure. As shown in fig. 7, in this application scenario, first, a growth state image of a monitored corn plant (e.g., D1 illustrated in fig. 7) and a soil fertility state image (e.g., D2 illustrated in fig. 7) are acquired, and then the growth state image of the monitored corn plant and the soil fertility state image are input into a server (e.g., S illustrated in fig. 7) where a corn high-yield planting and fertilizing algorithm is deployed, wherein the server is capable of processing the growth state image of the monitored corn plant and the soil fertility state image using the corn high-yield planting and fertilizing algorithm to obtain a classification result for a type tag representing the recommended application of fertilizer.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (10)
1. The corn high-yield planting and fertilizing method is characterized by comprising the following steps of:
acquiring a growth state image and a soil fertility state image of a monitored corn plant;
extracting growth state characteristic information in the growth state image of the monitored corn plant;
extracting fertility state characteristic information in the soil fertility state image; and
and generating recommended fertilizer application based on the growth state characteristic information and the fertility state characteristic information.
2. The method of high yield corn planting and fertilizing according to claim 1, wherein obtaining a growth status image and a soil fertility status image of the monitored corn plant comprises:
receiving a growth state image of the monitored corn plant acquired by a first camera; and
and receiving the soil fertility status image acquired by the second camera for planting the monitored corn plants.
3. The corn high yield planting and fertilizing method according to claim 2, wherein extracting growth state characteristic information in the growth state image of the monitored corn plant comprises:
and extracting plant growth state feature vectors from the growth state image of the monitored corn plant as the growth state feature information by using a growth state feature extractor based on a first convolutional neural network model.
4. The corn high yield planting and fertilizing method according to claim 3, characterized in that extracting plant growth state feature vectors from the growth state image of the monitored corn plant as the growth state feature information using a growth state feature extractor based on a first convolutional neural network model, comprising:
and performing convolution kernel-based feature filtering on the growth state image of the monitored corn plant by using the growth state feature extractor based on the first convolution neural network model to obtain the plant growth state feature vector.
5. The method of high-yield corn planting and fertilizing according to claim 4, wherein extracting fertility status characteristic information in the soil fertility status image comprises:
and extracting a soil fertility state feature vector from the soil fertility state image as the fertility state feature information by using a soil fertility state feature extractor based on a second convolutional neural network model.
6. The corn high-yield planting and fertilizing method according to claim 5, characterized in that extracting a soil fertility status feature vector from the soil fertility status image as the fertility status feature information using a soil fertility status feature extractor based on a second convolutional neural network model, comprising:
and performing feature filtering on the soil fertility state image based on a convolution kernel by using the soil fertility state feature extractor based on the second convolution neural network model so as to obtain the soil fertility state feature vector.
7. The method of high yield corn planting and fertilizing according to claim 6, wherein generating recommended application fertilizer based on said growth status characteristic information and said fertility status characteristic information comprises:
fusing the plant growth state feature vector and the soil fertility state feature vector to obtain a classification feature vector; and
and passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used for indicating the type label of the recommended fertilizer application.
8. The method of claim 7, wherein fusing the plant growth state feature vector and the soil fertility state feature vector to obtain a classification feature vector comprises:
fusing the plant growth state feature vector and the soil fertility state feature vector using a cascading function to obtain the classification feature vector;
wherein the cascading function is:
f(X i ,X j )＝Relu(W f [θ(X i ),φ(X j )])
wherein W is f ,θ(X i ) And phi (X) j ) All representing the point convolution of the input, relu as the activation function, []Representing the splicing operation, X i X is the characteristic value of each position in the plant growth state characteristic vector j And the characteristic value of each position in the characteristic vector of the soil fertility state is the characteristic value of each position in the characteristic vector of the soil fertility state.
9. The corn high yield planting and fertilizing method according to claim 8, further comprising a training step of: training the growth state feature extractor based on the first convolutional neural network model, the soil fertility state feature extractor based on the second convolutional neural network model and the classifier;
wherein the training step comprises:
acquiring training data, wherein the training data comprises a training growth state image of a monitored corn plant, a training soil fertility state image for planting the monitored corn plant, and a true value of a type label recommended to apply fertilizer;
passing the training growth state image of the monitored corn plant through the growth state feature extractor based on the first convolutional neural network model to obtain a training plant growth state feature vector;
passing the training soil fertility state image through the soil fertility state feature extractor based on the second convolutional neural network model to obtain a training soil fertility state feature vector;
fusing the training plant growth state feature vector and the training soil fertility state feature vector by using a cascading function to obtain a training classification feature vector;
the training classification feature vector passes through a classifier to obtain a classification loss function value; and
training the growth state feature extractor based on the first convolutional neural network model, the soil fertility state feature extractor based on the second convolutional neural network model and the classifier by using the classification loss function values, wherein in each round of iteration of training, feature transfer optimization iteration based on feature distribution cross-domain attention is performed on a weight matrix of the classifier.
10. The corn high-yield planting and fertilizing method according to claim 9, characterized in that in each iteration of the training, the feature transfer optimization iteration based on feature distribution cross-domain attention is performed on the weight matrix of the classifier according to the following optimization formula;
wherein, the optimization formula is:
wherein M is the weight matrix of the classifier, and the scale of M is M multiplied by M, V 1 To V m Is the M row vectors of the weight matrix M, I.I 2 Representing the two norms of the feature vector (Σ) j m i,j Is a row vector obtained by arranging the summation value of each row vector of the weight matrix M, and cov 1 (. Cndot.) and cov 2 (. Cndot.) all represent a single layer convolution operationThe process is carried out in such a way that,representing matrix multiplication (.) T Representing the transpose operation, M' represents the weight matrix of the classifier after iteration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310840449.8A CN116918546A (en) | 2023-07-11 | 2023-07-11 | Corn high-yield planting and fertilizing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310840449.8A CN116918546A (en) | 2023-07-11 | 2023-07-11 | Corn high-yield planting and fertilizing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116918546A true CN116918546A (en) | 2023-10-24 |
Family
ID=88385587
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310840449.8A Pending CN116918546A (en) | 2023-07-11 | 2023-07-11 | Corn high-yield planting and fertilizing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116918546A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117743975A (en) * | 2024-02-21 | 2024-03-22 | 君研生物科技(山西)有限公司 | Hillside cultivated land soil environment improvement method |
| CN117882546A (en) * | 2024-03-13 | 2024-04-16 | 山西诚鼎伟业科技有限责任公司 | Intelligent planting method for agricultural operation robot |
-
2023
- 2023-07-11 CN CN202310840449.8A patent/CN116918546A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117743975A (en) * | 2024-02-21 | 2024-03-22 | 君研生物科技(山西)有限公司 | Hillside cultivated land soil environment improvement method |
| CN117882546A (en) * | 2024-03-13 | 2024-04-16 | 山西诚鼎伟业科技有限责任公司 | Intelligent planting method for agricultural operation robot |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Akkem et al. | Smart farming using artificial intelligence: A review | |
| CN116918546A (en) | Corn high-yield planting and fertilizing method | |
| Mishra et al. | Weed density estimation in soya bean crop using deep convolutional neural networks in smart agriculture | |
| Kamath et al. | Classification of paddy crop and weeds using semantic segmentation | |
| CN117540908A (en) | Agricultural resource integration method and system based on big data | |
| De Saa et al. | Comparison between arima and deep learning models for temperature forecasting | |
| CN117114913A (en) | Intelligent agricultural data acquisition system based on big data | |
| Hasan et al. | Maize diseases image identification and classification by combining cnn with bi-directional long short-term memory model | |
| CN111325259A (en) | Remote sensing image classification method based on deep learning and binary coding | |
| CN116612435B (en) | Corn high-yield cultivation method | |
| Anitha Mary et al. | Scope and recent trends of artificial intelligence in Indian agriculture | |
| Chikkamath et al. | Benchmarking of Novel Convolutional Neural Network Models for Automatic Butterfly Identification | |
| CN117136765A (en) | Greenhouse control system and method based on intelligent agriculture | |
| CN116739868B (en) | Afforestation management system and method based on artificial intelligence | |
| Lydia et al. | Multi-label classification using deep convolutional neural network | |
| Gokulnath et al. | Regularized deep clustering approach for effective categorization of maize diseases | |
| Ahmed et al. | Detection and classification of leaf disease using deep learning for a greenhouses’ robot | |
| Anandhan et al. | Comprehensive study: machine learning & deep learning algorithms for paddy crops | |
| Berger et al. | Combining machine learning and simulation modelling for better predictions of crop yield and farmer income | |
| Rubini et al. | A comprehensive literature survey for deep learning approaches to agricultural applications | |
| Kour et al. | Machine Learning‐Based Hybrid Model for Wheat Yield Prediction | |
| Sathish et al. | A Review on Smart Farming based on Soil and Environmental Factors with Deep Learning Techniques | |
| Patil et al. | Crop Recommendation in Precision Agriculture Using Machine Learning Techniques | |
| Rohilla et al. | Automatic Image Segmentation and Feature Extraction of Potato Leaf Disease Using GLCM and HoG Features | |
| Kaur et al. | Applications of Machine Learning In Plant Disease Detection |
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 |