CN114140692A - Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning - Google Patents
Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning Download PDFInfo
- Publication number
- CN114140692A CN114140692A CN202111471473.6A CN202111471473A CN114140692A CN 114140692 A CN114140692 A CN 114140692A CN 202111471473 A CN202111471473 A CN 202111471473A CN 114140692 A CN114140692 A CN 114140692A
- Authority
- CN
- China
- Prior art keywords
- fresh corn
- deep learning
- aerial vehicle
- unmanned aerial
- fresh
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- 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
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning. The method identifies corn ears by using a deep learning technology, and constructs a sugar content and water content estimation model of fresh corn ears by combining vegetation indexes extracted by unmanned aerial vehicle remote sensing data and ground artificial sample measured values, so as to predict the maturity of fresh corn of a whole field plant. The random forest model established by the invention has stronger robustness and can be suitable for various field conditions. The invention has low requirement on the sensor, and greatly reduces the cost for purchasing the sensor by a user.
Description
Technical Field
The invention belongs to the field of agricultural automation, particularly relates to a fresh corn maturity prediction method, and particularly relates to a fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning.
Background
The maturity of the fresh corn can influence the taste and the nutritional value of the fresh corn, and is related to the market value of the fresh corn. The water content and the sugar content of the fresh corn are important reference standards reflecting the maturity of the fresh corn, so that a fresh corn maturity prediction model is constructed, and a decision can be provided for the picking period of the fresh corn.
By applying the unmanned aerial vehicle remote sensing technology, the sugar content and the water content of the fresh corn can be predicted with low cost, high efficiency and no damage. Few judgments are currently made about the harvest time of fresh corn, especially in terms of the extraction of the related characteristics of the female ear and the male ear of fresh corn, mainly including the male ear and mostly the estimated yield of the male ear. And the tassels of the fresh-eating corns are arranged on the upper parts of the corns, but the tassels are not sensitive to change, so that the maturity of the fresh-eating corns cannot be accurately reflected. In the former study, no specific solution is proposed for the harvest period of fresh-eating corn.
According to the method, a high-definition single-lens reflex camera is carried on a multi-rotor unmanned aerial vehicle platform, a high-resolution image of the fresh corn is shot, and then deep learning and machine learning are used for extraction and modeling respectively, so that the purpose of predicting the harvest time of the fresh corn is achieved.
Disclosure of Invention
Technical problem to be solved
In order to harvest fresh corn in a timely period and ensure the highest economic value and highest profit, the invention provides a fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning, and accurate prediction of the fresh corn harvesting period is realized.
(II) technical scheme
In order to solve the problems, the invention provides the following technical scheme, and provides a method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning, which is specifically as follows.
A fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning is characterized by comprising the following steps:
step S1, carrying a high-definition digital camera by using an unmanned aerial vehicle platform, and carrying out high-throughput and heading middle and later period data acquisition on the fresh-eating corn crop plants for acquiring continuously-changed dynamic phenotype data;
step S2, manually collecting a plurality of plant samples in the field to obtain measured values of the water content and the sugar content of the fresh corn;
step S3, precisely identifying the ear area of the fresh corn by using a deep learning neural network aiming at the image obtained in the step S1;
step S4, extracting DN values aiming at the ear areas identified in the step S3, carrying out radiometric calibration treatment, and converting the DN values into the surface reflectivity of the outer layer of the atmosphere;
step S5, calculating and obtaining a related color vegetation index based on the reflectivity result;
step S6, aiming at the sample object collected in the step S2, carrying out correlation analysis on the correlated color vegetation index and the artificially measured sugar content and water content, and selecting the color vegetation index with good correlation performance;
step S7, adopting a random forest method to construct a prediction model of the sugar content and the water content of the fresh corn ear by using the selected color vegetation index;
and step S8, based on the sugar content and water content prediction model of the fresh corn ear, using the related color vegetation index obtained in the step S5 as input to obtain a predicted value of the sugar content and the water content of the whole field plant, and evaluating the maturity of the fresh corn through the predicted value.
More specifically, the data collection of step S1 is a continuous large area collection covering the middle and late stages of the corn ear color change.
More specifically, in step S3, the fresh corn ear is subjected to image segmentation using the deep learning neural network U-Net.
More specifically, when a deep learning neural network U-Net is adopted to carry out image segmentation on fresh corn ears, 256 × 256 images are input, a convolution layer with a window shape of 3 × 3 is formed, and then a maximum pooling layer with a step of 2 and a window shape of 2 × 2 is formed; the convolutional layer keeps the height and width of the input unchanged, while the pooling layer halves them.
More specifically, the color vegetation index calculated in step S5 includes, but is not limited to, one or more of the following vegetation indexes: greenness vegetation index GVI, normalized yellowness vegetation index NDYI, normalized difference vegetation index NDI, green leaf area index GLA, hyper-red index EXR, R component R, G component G, and B component B.
More specifically, in step S7, a random forest algorithm in machine learning is used, the number of trees to be constructed is 100, and the number of input features when each node of the decision tree is split is 50.
More specifically, in step S8, the image data of any time and area in the middle and later stages of the ear obtained by the unmanned aerial vehicle is used to extract the relevant color vegetation index, so as to obtain the predicted values of the sugar content and the water content of the target fresh-eating corn ear.
(III) advantageous effects
The invention provides a method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning based on technical accumulation and research and development of an inventor in the field for years.
Compared with the prior art, the method has the following technical advantages: (1) the method comprises the steps of extracting the color vegetation index of the fresh corn ear area by using phenotype data acquired by an unmanned aerial vehicle, and constructing a prediction model of the sugar content and the water content of the fresh corn by combining a random forest method. (2) The fresh corn ear region is obtained on the high-definition image obtained by the unmanned aerial vehicle, the method is combined with the deep learning neural network method, the accurate and ideal result can be obtained, and compared with the method for delineating the ROI region through manual visual interpretation, the method saves time and cost and is higher in identification precision. (3) The requirement on the resolution of the image is not high, and a similar and ideal result can still be obtained under the appropriate resolution, so that the cost for purchasing the multispectral camera by a user can be saved for the remote sensing data acquisition by using the unmanned aerial vehicle.
Drawings
FIG. 1 is a flow chart of the method of the present invention
FIG. 2 shows regression analysis of fresh corn ear and tassel in terms of sugar content and water content
FIG. 3 is a comparison graph of predicted value and measured value of sugar content prediction model constructed by fresh corn ears
FIG. 4 is a comparison graph of the predicted value and the measured value of a water content prediction model constructed by fresh corn ears
Detailed Description
The invention provides a method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning, aiming at solving the technical problem, and the flow chart of the method is shown in figure 1.
A fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning is characterized by comprising the following steps:
step S1, carrying a high-definition digital camera by using an unmanned aerial vehicle platform, and carrying out high-throughput and heading middle and later period data acquisition on the fresh-eating corn crop plants for acquiring continuously-changed dynamic phenotype data;
step S2, manually collecting a plurality of plant samples in the field to obtain measured values of the water content and the sugar content of the fresh corn;
step S3, precisely identifying the ear area of the fresh corn by using a deep learning neural network aiming at the image obtained in the step S1;
step S4, extracting DN values aiming at the ear areas identified in the step S3, carrying out radiometric calibration treatment, and converting the DN values into the surface reflectivity of the outer layer of the atmosphere;
step S5, calculating and obtaining a related color vegetation index based on the reflectivity result;
step S6, aiming at the sample object collected in the step S2, carrying out correlation analysis on the correlated color vegetation index and the artificially measured sugar content and water content, and selecting the color vegetation index with good correlation performance;
step S7, adopting a random forest method to construct a prediction model of the sugar content and the water content of the fresh corn ear by using the selected color vegetation index;
and step S8, based on the sugar content and water content prediction model of the fresh corn ear, using the related color vegetation index obtained in the step S5 as input to obtain a predicted value of the sugar content and the water content of the whole field plant, and evaluating the maturity of the fresh corn through the predicted value.
More specifically, the data collection of step S1 is a continuous large area collection covering the middle and late stages of the corn ear color change.
More specifically, in step S3, the fresh corn ear is subjected to image segmentation using the deep learning neural network U-Net.
More specifically, when a deep learning neural network U-Net is adopted to carry out image segmentation on fresh corn ears, 256 × 256 images are input, a convolution layer with a window shape of 3 × 3 is formed, and then a maximum pooling layer with a step of 2 and a window shape of 2 × 2 is formed; the convolutional layer keeps the height and width of the input unchanged, while the pooling layer halves them.
More specifically, the color vegetation index calculated in step S5 includes, but is not limited to, one or more of the following vegetation indexes: greenness vegetation index GVI, normalized yellowness vegetation index NDYI, normalized difference vegetation index NDI, green leaf area index GLA, hyper-red index EXR, R component R, G component G, and B component B.
More specifically, in step S7, a random forest algorithm in machine learning is used, the number of trees to be constructed is 100, and the number of input features when each node of the decision tree is split is 50.
More specifically, in step S8, the image data of any time and area in the middle and later stages of the ear obtained by the unmanned aerial vehicle is used to extract the relevant color vegetation index, so as to obtain the predicted values of the sugar content and the water content of the target fresh-eating corn ear.
In order to better explain the technical scheme of the invention, the corn ear is taken as a specific application object, and the invention obtains the following results:
(1) the effect of constructing a model by using the pistils of the fresh corn is better than that of the stamens
Based on stamens and pistils extracted by the deep learning U-Net segmentation network method, the selected vegetation indexes are respectively calculated, and then regression analysis is carried out on the calculated vegetation indexes, the sugar content and the water content, so that the determining coefficient R of the stamens of the fresh-eating corn is found2Generally lower, no freshly eaten corn pistils perform well as shown in figure 2.
(2) Fresh corn maturity prediction effect
Based on the method provided by the invention, the optimal harvesting period of the fresh-eating corn is predicted by the sugar content and the water content of pistils of the fresh-eating corn respectively, as shown in fig. 3, the correlation coefficient R of a sugar content prediction model is 0.902, and the standard root mean square error NRMSE is 0.121; as shown in fig. 4, the correlation coefficient R of the moisture content prediction model is 0.776 and the standard root mean square error NRMSE is 0.116. It can be found that the present invention can stably predict the optimum maturation period of fresh-eating corn.
(3) Application of corn planting field in Hannan region of Wuhan city
Based on the operation flow of the method, the method is applied to the field of corn field areas (30.319 degrees N and 113.961 degrees E) planted by farmers in Deng-West village in Han, Wuhan. The method is developed from 1 day 6 months to 21 days 2021 years in 2021, unmanned aerial vehicle image data of the fresh corn ears at the middle and later periods are collected, and simultaneously, the sugar content and the water content of a target corresponding to the unmanned aerial vehicle are actually measured. A sugar content and water content prediction model of the fresh corn is constructed, predicted values of the sugar content and the water content of the fresh corn in a target area are successfully obtained, the maturity of the fresh corn is accurately reflected through the height of the predicted values, and the optimal harvesting period is further obtained.
The specific examples described in the application are only illustrative of the spirit of the invention. Various modifications, additions and substitutions of types may be made by those skilled in the art without departing from the spirit of the invention or exceeding the scope of the invention as defined in the accompanying claims.
Claims (7)
1. A fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning is characterized by comprising the following steps:
step S1, carrying a high-definition digital camera by using an unmanned aerial vehicle platform, and carrying out high-throughput and heading middle and later period data acquisition on the fresh-eating corn crop plants for acquiring continuously-changed dynamic phenotype data;
step S2, manually collecting a plurality of plant samples in the field to obtain measured values of the water content and the sugar content of the fresh corn;
step S3, precisely identifying the ear area of the fresh corn by using a deep learning neural network aiming at the image obtained in the step S1;
step S4, extracting DN values aiming at the ear areas identified in the step S3, carrying out radiometric calibration treatment, and converting the DN values into the surface reflectivity of the outer layer of the atmosphere;
step S5, calculating and obtaining a related color vegetation index based on the reflectivity result;
step S6, aiming at the sample object collected in the step S2, carrying out correlation analysis on the correlated color vegetation index and the artificially measured sugar content and water content, and selecting the color vegetation index with good correlation performance;
step S7, adopting a random forest method to construct a prediction model of the sugar content and the water content of the fresh corn ear by using the selected color vegetation index;
and step S8, based on the sugar content and water content prediction model of the fresh corn ear, using the related color vegetation index obtained in the step S5 as input to obtain a predicted value of the sugar content and the water content of the whole field plant, and evaluating the maturity of the fresh corn through the predicted value.
2. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 1, wherein the method comprises the following steps: the data collection of the step S1 is continuous large-area collection and covers the middle and later periods of the color change of the corn ear.
3. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 1, wherein the method comprises the following steps: in step S3, the fresh corn ears are subjected to image segmentation by adopting a deep learning neural network U-Net.
4. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 3, wherein the method comprises the following steps: when a deep learning neural network U-Net is adopted to carry out image segmentation on fresh corn ears, 256 multiplied by 256 images are input, a convolution layer with a window shape of 3 multiplied by 3 is connected, and then a maximum pooling layer with a step of 2 and a window shape of 2 multiplied by 2 is connected; the convolutional layer keeps the height and width of the input unchanged, while the pooling layer halves them.
5. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 1, wherein the method comprises the following steps: the color vegetation index calculated in step S5 includes, but is not limited to, one or more of the following vegetation indexes: greenness vegetation index GVI, normalized yellowness vegetation index NDYI, normalized difference vegetation index NDI, green leaf area index GLA, hyper-red index EXR, R component R, G component G, and B component B.
6. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 1, wherein the method comprises the following steps: in step S7, a random forest algorithm in machine learning is used, the number of trees constructed is 100, and the number of input features when each node of the decision tree is split is 50.
7. The method for predicting the maturity of fresh corn based on unmanned aerial vehicle remote sensing and deep learning according to claim 1, wherein the method comprises the following steps: in step S8, the image data of any time and area in the middle and late stages of the ear obtained by the unmanned aerial vehicle is used to extract the vegetation index of the relevant color, and the predicted values of the sugar content and the water content of the target fresh-eating corn ear are obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111471473.6A CN114140692A (en) | 2021-11-25 | 2021-11-25 | Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111471473.6A CN114140692A (en) | 2021-11-25 | 2021-11-25 | Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114140692A true CN114140692A (en) | 2022-03-04 |
Family
ID=80387924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111471473.6A Pending CN114140692A (en) | 2021-11-25 | 2021-11-25 | Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114140692A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115029405A (en) * | 2022-04-04 | 2022-09-09 | 国药肽谷有限公司 | Method for extracting xanthoceras sorbifolia bunge peptide |
| CN115187609A (en) * | 2022-09-14 | 2022-10-14 | 合肥安杰特光电科技有限公司 | Method and system for detecting rice yellow grains |
| CN116406564A (en) * | 2023-04-24 | 2023-07-11 | 山东常林机械集团股份有限公司 | Efficient corn harvester |
| CN117409403A (en) * | 2023-12-15 | 2024-01-16 | 南京农业大学三亚研究院 | Rice spike maturity estimation method based on deep learning |
-
2021
- 2021-11-25 CN CN202111471473.6A patent/CN114140692A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115029405A (en) * | 2022-04-04 | 2022-09-09 | 国药肽谷有限公司 | Method for extracting xanthoceras sorbifolia bunge peptide |
| CN115029405B (en) * | 2022-04-04 | 2023-08-22 | 国药肽谷有限公司 | Method for extracting shinyleaf yellowhorn peptide |
| CN115187609A (en) * | 2022-09-14 | 2022-10-14 | 合肥安杰特光电科技有限公司 | Method and system for detecting rice yellow grains |
| CN116406564A (en) * | 2023-04-24 | 2023-07-11 | 山东常林机械集团股份有限公司 | Efficient corn harvester |
| CN116406564B (en) * | 2023-04-24 | 2023-11-07 | 山东常林机械集团股份有限公司 | Efficient corn harvester |
| CN117409403A (en) * | 2023-12-15 | 2024-01-16 | 南京农业大学三亚研究院 | Rice spike maturity estimation method based on deep learning |
| CN117409403B (en) * | 2023-12-15 | 2024-03-19 | 南京农业大学三亚研究院 | Rice spike maturity estimation method based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114140692A (en) | Fresh corn maturity prediction method based on unmanned aerial vehicle remote sensing and deep learning | |
| EP4150515A1 (en) | System and method for crop monitoring | |
| CN116308866B (en) | Rice ear biomass estimation method and system based on canopy reflection spectrum | |
| CN111275567B (en) | SENP cotton yield estimation method based on unmanned aerial vehicle image and estimation model construction method | |
| CN112861712B (en) | Agricultural pest and disease monitoring method based on artificial intelligence and multi-temporal remote sensing | |
| CN114821321A (en) | Blade hyperspectral image classification and regression method based on multi-scale cascade convolution neural network | |
| CN113223040A (en) | Remote sensing-based banana yield estimation method and device, electronic equipment and storage medium | |
| CN114694020B (en) | Construction method of cotton aphid remote sensing prediction model | |
| CN115661641A (en) | Method and system for evaluating important phenotype parameters of tea garden in different growth stages based on multi-source unmanned aerial vehicle remote sensing | |
| CN117292267B (en) | Method and system for estimating rice aboveground biomass in segments based on weather information | |
| US11864552B2 (en) | Digital detection method and system for predicting drug resistance of transgenic maize | |
| Liang et al. | CanopyCAM–an edge-computing sensing unit for continuous measurement of canopy cover percentage of dry edible beans | |
| CN116482041B (en) | Rice heading period nondestructive rapid identification method and system based on reflection spectrum | |
| CN115035423B (en) | Hybrid rice parent and parent identification extraction method based on unmanned aerial vehicle remote sensing image | |
| CN115170969A (en) | Method and system for constructing overground biomass model of vertically-grown crop | |
| CN115063690A (en) | Vegetation classification method based on NDVI (normalized difference vegetation index) time sequence characteristics | |
| CN114821300A (en) | Method for judging plant protection operation time for preventing wheat scab based on multispectral remote sensing of unmanned aerial vehicle | |
| Poleshchenko et al. | Development of a System for Automated Control of Planting Density, Leaf Area Index and Crop Development Phases by UAV Photos | |
| CN115753758B (en) | Method and device for predicting cotton field opening rate based on unmanned aerial vehicle multispectral | |
| CN114913426A (en) | Swin-Transformer-based blade counting method | |
| Chandel et al. | Mapping apple canopy attributes using aerial multispectral imagery for precision crop inputs management | |
| Liang et al. | Development an edge-computing sensing unit for continuous measurement of canopy cover percentage of dry edible beans | |
| CN117011702A (en) | Tea garden identification method and device based on satellite images | |
| Treder et al. | Application of ImageJ Software in the assessment of flowering Intensity and growth Vigor of Pear Trees | |
| Ouyang | The Applications of UAV and Ground Digital Imagery Analysis for Monitoring Grapevine Canopy in Vineyard |
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 |