CN106442329A - Winter wheat leaf area index estimation method based on canopy image parameters - Google Patents
Winter wheat leaf area index estimation method based on canopy image parameters Download PDFInfo
- Publication number
- CN106442329A CN106442329A CN201610770439.1A CN201610770439A CN106442329A CN 106442329 A CN106442329 A CN 106442329A CN 201610770439 A CN201610770439 A CN 201610770439A CN 106442329 A CN106442329 A CN 106442329A
- Authority
- CN
- China
- Prior art keywords
- canopy
- area index
- image
- shooting
- winter wheat
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/28—Measuring arrangements characterised by the use of optical techniques for measuring areas
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
- G01N2021/1797—Remote sensing in landscape, e.g. crops
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a winter wheat leaf area index estimation method based on canopy image parameters, relating to the technical field of plants. The estimation method comprises the following steps: fixing a digital camera 1m away from the tops of wheat canopies to be vertical to the ground for shooting when the digital camera is used for shooting the images of the wheat canopies; aligning the viewing range of the camera with rectangular squares for shooting during shooting; storing the brightness values of image pixels in 14 bits, decoding the brightness values by dcraw v9.22 software and setting the following options: automatic white balance set during shooting is adopted; an AHD algorithm is adopted for demosaicing; the RGB values are converted to sRGB color space and are stored in the 16-bit TIFF; estimating the winter wheat leaf area index by utilizing a constructed artificial neural network model and using the canopy cover and canopy images R, G and B, which are obtained through image analysis, as input variables. The method is convenient for achieving accurate estimation, is convenient to use and simple and convenient to operate and saves time.
Description
Technical field
The present invention relates to plant technology field is and in particular to a kind of leaves of winter wheat area index based on canopy image parameter
Evaluation method.
Background technology
Digital camera, as a kind of new visible light ground remote sensing instrument, is studied in crop condition monitoring in recent years
To widely applying.Most results of study show the Canopy cover degrees being obtained by canopy graphical analysis and canopy image color index
It is in extremely significantly correlated with Crop leaf area index, and is in mainly nonlinear correlation, and utilize its correlation matching paddy rice, cotton
Leaf area index appraising model Deng crop.Based on crop canopies coverage come of age tomographic image colo r index crop growing state estimation
Model is mainly fitted by non-linear regression method.But when crop leaf area index or the upperground part biomass are larger, its
Estimation error significantly increases, and when this is larger with Crop leaf area index, overlapping the led to Canopy cover degrees of blade can not be regular
Ground reflection Crop leaf area index variation is relevant, and existing with the saturation appeared in the crop growing state estimation based on high-spectral data
As similar.The traditional statistical methods such as nonlinear regression are not enough to the explanation of this problem, and artificial neural network is asked to non-linear
The matching of topic has clear superiority, and existing majority is about the crop length based on high spectrum reflection data and Artificial Neural Network
The research report of gesture estimation.But grinding of the crop growing state estimation based on canopy image parameter is built using Artificial Neural Network
Study carefully report less；Its data in estimation is inaccurate.
Content of the invention
Present invention aims to the defect of prior art and deficiency, one kind is provided to be based on canopy image parameter and people
The leaves of winter wheat area index evaluation method of artificial neural networks model.
In order to solve the problems of background technology, a kind of winter wheat blade face based on canopy image parameter of the present invention
Long-pending index evaluation method, its evaluation method is：When shooting wheat canopy image with digital camera, camera is fixed on away from wheat
At the 1m of canopy top, perpendicular to the ground shot；The rectangle grid of homemade 72.5cm × 45cm is placed in field, and little
Wheat canopy keeps identical height；During shooting, the viewfinder range be aligned rectangle grid of camera is shot, shooting time is 11:
00-13:00；Camera is set to auto exposure mode, and using multiple spot auto-focusing, ISO is set as 400, and image resolution ratio sets
For 4,500,000 pixels, image is with the storage of RAW form；The brightness value of image slices vegetarian refreshments is to be stored with 14, and its decoding adopts
Dcraw v9.22 software, and set following option：White balance is using the AWB arranging when shooting；Demosaicing adopts
AHD algorithm；Rgb value is converted to sRGB color space and is preserved with 16 tiff formats；Using constructed artificial neural network,
Come of age tomographic image R, G, B as input variable with the Canopy cover degrees that graphical analysis obtains, estimates leaves of winter wheat area index.
The present invention has the beneficial effect that：Facilitate implementation accurate estimation, easy to use, easy and simple to handle, time-consuming.
Brief description
Fig. 1 is the dependency relation figure of leaves of winter wheat area index and Canopy cover degrees and R, G, B brightness value in the present invention；
Fig. 2 is the leaves of winter wheat area index fitting effect in the present invention based on BP artificial nerve network model；
Fig. 3 is the pass in the present invention between the leaf area index estimated value based on BP artificial nerve network model and measured value
System's figure.
Specific embodiment
Below in conjunction with the accompanying drawings, the present invention is further illustrated.
In order that the objects, technical solutions and advantages of the present invention become more apparent, below in conjunction with accompanying drawing and be embodied as
Mode, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only in order to explain this
Invention, is not intended to limit the present invention.
As shown in figure 1, this specific embodiment adopts the following technical scheme that：Its evaluation method is：(good with digital camera
Can 60D camera, camera lens 18-135mm IS) shoot wheat canopy image when, camera is fixed at the 1m of wheat canopy top,
Perpendicular to the ground shot；The rectangle grid of homemade 72.5cm × 45cm is placed in field, keeps phase with wheat canopy
With height；During shooting, the viewfinder range be aligned rectangle grid of camera is shot, shooting time is 11:00-13:00；Phase
Machine is set to auto exposure mode, and using multiple spot auto-focusing, ISO is set as 400, and image resolution ratio is set as 4,500,000 pixels
(2592 × 1728), image is stored with RAW form；The brightness value of image slices vegetarian refreshments is to be stored with 14, and its decoding adopts
Dcraw v9.22 software, and set following option：White balance is using the AWB arranging when shooting；Demosaicing adopts
AHD algorithm；Rgb value is converted to sRGB color space and is preserved with 16 tiff formats.
Canopy of winter wheat image segmentation adopts random forests algorithm.Specifically, replicating from canopy image and splicing respectively
Go out wheat and the pedological map picture of 400 × 400 pixels, read ruddiness brightness value (R, the redness of each pixel of stitching image
Intensity), green glow brightness value (G, greenness intensity), blue light brightness value (B, blueness intensity)
Component is simultaneously converted to L*, a*, b* value of CIEL*a*b* color space, as independent variable with R software
One random forest containing 100 classification trees of RandomForest bag training is split to image.From segmentation figure picture, read the winter
R, G, B value of wheat image slices vegetarian refreshments, subsequently calculates Canopy cover degrees (CC, canopy cover).
Formula is：
Leaves of winter wheat area index and the correlation of Canopy cover degrees and R, G, B brightness value：
Leaves of winter wheat area index and three color component ruddiness brightness of Canopy cover degrees (CC) and sRGB color space
The correlation analysis result of value (R), green glow brightness value (G) and blue light brightness value (B) is shown in table 1 and Fig. 1.From table 1 it follows that
In addition to B color component, leaves of winter wheat area index is in significantly correlated with CC, R and G, is wherein proportionate with CC, is in negative with R and G
Related；The coefficient correlation absolute value of leaf area index and CC, R and G, between 0.437～0.910, all reaches the pole level of signifiance
(test in 2014, sample number=71, r0.05=0.234, r0.01=0.304；Test in 2015, sample number=70, r0.05=
0.235, r0.01, and 2 years result of the tests are more consistent (as Fig. 1)=0.306).Correlation analysis result shows, to taking out after turning green
Leaves of winter wheat area index before fringe and CC and R, G be in extremely significantly correlated, wherein related to CC belong to height correlation, it is 2 years
Coefficient correlation is respectively 0.905 and 0.910, and fall within moderate related to R is related, shows to divide based on canopy of winter wheat image
CC and canopy image R of analysis acquisition etc. have higher relevance and between leaf area index, and this also illustrates based on CC and sRGB color
The estimation of the leaves of winter wheat area index of three color components (R, G and B) in space has feasibility, i.e. CC and sRGB color space
Three color components can be used to build the appraising model of leaves of winter wheat area index.
Table 1 leaves of winter wheat area index is as follows with Canopy cover degrees, R, G and B coefficient correlation：
The foundation of leaves of winter wheat area index appraising model：
Using BP Artificial Neural Network, with R, G and B of CC and sRGB color component
As independent variable, leaf area index is fitted.100 data are randomly selected in totally 141 groups of sample datas
(49 groups of data of 2014 and 51 groups of data of 2015) carry out the structure of the appraising model of leaf area index, with the coefficient of determination
(R2), root-mean-square error (RMSE) and relative root-mean-square error (RRMSE) carry out the degree of fitting of evaluation model.
BP artificial neural network is that the one kind being proposed by Rumelhart and McCelland for 1986 is calculated by error Back-Propagation
The Multi-layered Feedforward Networks of method training, are one of artificial nerve network models of being most widely used at present.Its learning rules is logical
Cross weight and the threshold value that backpropagation constantly to adjust each node of network, make the error sum of squares of network minimum.BP artificial neuron
The structure of network generally comprises input layer (input layer), one or more hidden layer (hide layer) and output layer
(output layer), in this research, hidden layer is set to 1 layer, the nodes of hidden layer, through artificial neural network testing setup
For 5 best nodes of estimation effect.Each function of artificial neural network and parameter setting are as follows：Training function adopts
Trainlm function；Hidden layer activation primitive adopts tansig function；Output layer adopts pureline function；Artificial neural network
The weight of activation primitive and threshold value are default value.After artificial neural network training terminates, 41 groups of data (22 of test in 2014
Organize 19 groups of data of data and 2015) as input vector, calculated using Sim function and export leaves of winter wheat area index,
The finally model parameter (being shown in Table 3, table 4) such as the weight of output artificial neural network, deviation.
The leaves of winter wheat area index appraising model matching based on Canopy cover degrees and canopy image sRGB color component for the table 2
Degree：
Table 2 and Fig. 2 are model parameter and the fitting effect of leaf area index appraising model.By table
2 and Fig. 2 understand, model has higher degree of fitting, and the appraising model coefficient of determination for 0.900, RRMSE is
17.394.
Table 3 is come of age based on Canopy cover degrees the artificial neural network leaf area index appraising model of tomographic image color component：
§Input data inputs artificial neural network after being normalized with Mapminmax function first again, and neural
The normalized leaf area index data that network is exported obtains leaf area index data after renormalization.
Table 4 artificial neural network weight and deviation
The estimation precision of leaves of winter wheat area index appraising model：
For verifying the precision of constructed leaf area index appraising model in this research, using 2 years test datas (2014
22 groups of data of year test and 19 groups of data of 2015) estimation result of model to be checked, it the results are shown in Table 5 and figure
3.From table 5 and Fig. 3,
The coefficient of determination of model is that 0.886, RMSE and RRMSE is respectively 0.536 and 18.320%.Model predication value and reality
The 1 of measured value:1 graph of a relation (Fig. 3) understands, most scatterplot is distributed in 1:Near 1 line.Table 5 and Fig. 3 result show that this research is constructed
Leaf area index appraising model there is preferable estimation precision.
The model checking of table 5 leaf area index appraising model and cross validation results：
The above, only in order to technical scheme to be described and unrestricted, those of ordinary skill in the art are to this
Other modifications or equivalent that bright technical scheme is made, without departing from the spirit and scope of technical solution of the present invention,
All should cover in the middle of scope of the presently claimed invention.
Claims (1)
1. a kind of leaves of winter wheat area index evaluation method based on canopy image parameter it is characterised in that：Its evaluation method
For：When shooting wheat canopy image with digital camera, camera is fixed at the 1m of wheat canopy top, perpendicular to the ground carries out
Shoot；The rectangle grid of homemade 72.5cm × 45cm is placed in field, height identical with wheat canopy holding；Will during shooting
The viewfinder range be aligned rectangle grid of camera is shot, and shooting time is 11:00-13:00；Camera is set to automatic exposure
Pattern, using multiple spot auto-focusing, ISO is set as 400, and image resolution ratio is set as 4,500,000 pixels, and image is deposited with RAW form
Storage；The brightness value of image slices vegetarian refreshments is to be stored with 14, and its decoding adopts dcraw v9.22 software, and sets following option：
White balance is using the AWB arranging when shooting；Demosaicing adopts AHD algorithm；Rgb value is converted to sRGB color space
And preserved with 16 tiff formats；Using constructed artificial nerve network model, the Canopy cover degrees being obtained with graphical analysis
Come of age tomographic image R, G, B for input variable, estimates leaves of winter wheat area index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610770439.1A CN106442329A (en) | 2016-08-31 | 2016-08-31 | Winter wheat leaf area index estimation method based on canopy image parameters |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610770439.1A CN106442329A (en) | 2016-08-31 | 2016-08-31 | Winter wheat leaf area index estimation method based on canopy image parameters |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106442329A true CN106442329A (en) | 2017-02-22 |
Family
ID=58091384
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610770439.1A Pending CN106442329A (en) | 2016-08-31 | 2016-08-31 | Winter wheat leaf area index estimation method based on canopy image parameters |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106442329A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108229403A (en) * | 2018-01-08 | 2018-06-29 | 中国科学院遥感与数字地球研究所 | A kind of mixed model construction method for being used to estimate vegetation leaf area index |
| CN111751376A (en) * | 2020-07-25 | 2020-10-09 | 江西省农业科学院农业工程研究所 | Rice nitrogen nutrition estimation method based on canopy image feature derivation |
| CN112330672A (en) * | 2020-11-28 | 2021-02-05 | 华中农业大学 | Crop leaf area index inversion method based on PROSAIL model and under participation of canopy coverage in optimization |
| CN113256567A (en) * | 2021-05-08 | 2021-08-13 | 中国农业大学 | Banana leaf area index detection method and system |
| WO2023197496A1 (en) * | 2022-04-15 | 2023-10-19 | 石河子大学 | Comprehensive evaluation indicator monitoring and evaluation method and system for machine-harvested cotton defoliation effects |
| CN117853936A (en) * | 2024-03-08 | 2024-04-09 | 山东省气候中心 | Method, system, terminal and storage medium for identifying freezing injury of wheat canopy image |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021472A (en) * | 2007-01-17 | 2007-08-22 | 南京农业大学 | Portable multiway crop leaf nitrogen nutrient index nondestructive monitoring device |
| CN102331244A (en) * | 2011-08-10 | 2012-01-25 | 浙江大学 | Method of acquiring leaf area index (LAI) of rice canopy by using common digital camera |
-
2016
- 2016-08-31 CN CN201610770439.1A patent/CN106442329A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021472A (en) * | 2007-01-17 | 2007-08-22 | 南京农业大学 | Portable multiway crop leaf nitrogen nutrient index nondestructive monitoring device |
| CN102331244A (en) * | 2011-08-10 | 2012-01-25 | 浙江大学 | Method of acquiring leaf area index (LAI) of rice canopy by using common digital camera |
Non-Patent Citations (2)
| Title |
|---|
| 崔日鲜等: "基于可见光光谱和ＢＰ人工神经网络的冬小麦生物量估算研究", 《光谱学与光谱分析》 * |
| 崔日鲜等: "基于机器学习和可见光光谱的冬小麦叶片氮积累量估算", 《光谱学与光谱分析》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108229403A (en) * | 2018-01-08 | 2018-06-29 | 中国科学院遥感与数字地球研究所 | A kind of mixed model construction method for being used to estimate vegetation leaf area index |
| CN111751376A (en) * | 2020-07-25 | 2020-10-09 | 江西省农业科学院农业工程研究所 | Rice nitrogen nutrition estimation method based on canopy image feature derivation |
| CN112330672A (en) * | 2020-11-28 | 2021-02-05 | 华中农业大学 | Crop leaf area index inversion method based on PROSAIL model and under participation of canopy coverage in optimization |
| CN113256567A (en) * | 2021-05-08 | 2021-08-13 | 中国农业大学 | Banana leaf area index detection method and system |
| CN113256567B (en) * | 2021-05-08 | 2023-09-22 | 中国农业大学 | Banana leaf area index detection method and system |
| WO2023197496A1 (en) * | 2022-04-15 | 2023-10-19 | 石河子大学 | Comprehensive evaluation indicator monitoring and evaluation method and system for machine-harvested cotton defoliation effects |
| CN117853936A (en) * | 2024-03-08 | 2024-04-09 | 山东省气候中心 | Method, system, terminal and storage medium for identifying freezing injury of wheat canopy image |
| CN117853936B (en) * | 2024-03-08 | 2024-05-31 | 山东省气候中心 | Method, system, terminal and storage medium for identifying freezing injury of wheat canopy image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106442329A (en) | Winter wheat leaf area index estimation method based on canopy image parameters | |
| Liu et al. | Estimating leaf area index using unmanned aerial vehicle data: shallow vs. deep machine learning algorithms | |
| CN106469302B (en) | A kind of face skin quality detection method based on artificial neural network | |
| CN108615071B (en) | Model testing method and device | |
| CN108875821A (en) | The training method and device of disaggregated model, mobile terminal, readable storage medium storing program for executing | |
| CN107067043A (en) | A kind of diseases and pests of agronomic crop detection method | |
| CN107092926A (en) | Service robot object recognition algorithm based on deep learning | |
| CN110479636B (en) | Method and device for automatically sorting tobacco leaves based on neural network | |
| EP4154159B1 (en) | System and method for estimating vegetation coverage in a real-world environment | |
| Concepcion et al. | Estimation of photosynthetic growth signature at the canopy scale using new genetic algorithm-modified visible band triangular greenness index | |
| CN109344738A (en) | The recognition methods of crop diseases and pest crop smothering and device | |
| CN109670509A (en) | Winter wheat seedling stage growing way parameter evaluation method and system based on convolutional neural networks | |
| CN113822198B (en) | Peanut growth monitoring method, system and medium based on UAV-RGB image and deep learning | |
| Song et al. | Assessment of wheat chlorophyll content by the multiple linear regression of leaf image features | |
| CN111476153A (en) | Method for calculating fruit maturity | |
| CN111767943A (en) | Mulching film identification method and device, electronic equipment and storage medium | |
| CN110796148A (en) | Litchi insect pest monitoring and identifying system and litchi insect pest monitoring and identifying method | |
| CN113989181A (en) | Remote sensing monitoring method and electronic equipment for growing period of crops in plastic greenhouse | |
| CN114663761A (en) | Crop growth condition determining method, device, equipment and storage medium | |
| CN113256567B (en) | Banana leaf area index detection method and system | |
| CN113514402A (en) | System and method for predicting chlorophyll content of winter wheat | |
| CN107016401A (en) | Digital camera image-based rice canopy recognition method | |
| Schmidtke | Developing a phone-based imaging tool to inform on fruit volume and potential optimal harvest time | |
| CN111222481B (en) | Method and device for identifying clothes color | |
| Kusnandar et al. | A Novel Method for Optimizing Color Selection Using the Hadamard Product Technique |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170222 |