CN114170500A - Wheat lodging area extraction system and method - Google Patents
Wheat lodging area extraction system and method Download PDFInfo
- Publication number
- CN114170500A CN114170500A CN202010845296.2A CN202010845296A CN114170500A CN 114170500 A CN114170500 A CN 114170500A CN 202010845296 A CN202010845296 A CN 202010845296A CN 114170500 A CN114170500 A CN 114170500A
- Authority
- CN
- China
- Prior art keywords
- lodging
- wheat
- image
- multispectral
- area
- 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
- 241000209140 Triticum Species 0.000 title claims abstract description 109
- 235000021307 Triticum Nutrition 0.000 title claims abstract description 109
- 238000000034 method Methods 0.000 title claims abstract description 76
- 238000000605 extraction Methods 0.000 title claims abstract description 40
- 238000012360 testing method Methods 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims abstract description 5
- 238000004458 analytical method Methods 0.000 claims description 17
- 238000013145 classification model Methods 0.000 claims description 15
- 238000011156 evaluation Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 8
- 238000011835 investigation Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 238000012795 verification Methods 0.000 description 12
- 238000010224 classification analysis Methods 0.000 description 9
- 238000011160 research Methods 0.000 description 8
- 238000012549 training Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 230000008569 process Effects 0.000 description 3
- 238000007476 Maximum Likelihood Methods 0.000 description 2
- 235000013339 cereals Nutrition 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 101100175003 Oryza sativa subsp. japonica RGB1 gene Proteins 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000012407 engineering method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 238000009331 sowing Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a wheat lodging area extraction system and a method, wherein the extraction system comprises: the acquisition unit is used for acquiring field image data of a test area by carrying a multispectral sensor through an unmanned aerial vehicle; the pretreatment unit is connected with the acquisition unit and is used for processing the acquired field image data of the test area to obtain a wheat canopy multispectral image; the computing unit is connected with the preprocessing unit and is used for computing the texture characteristics of the target crop in the multispectral image of the wheat canopy; and the selecting unit is connected with the calculating unit and is used for respectively combining the calculated textural features with the multispectral image, classifying and analyzing the image by using a Mahalanobis distance classification method and extracting the lodging area of the wheat. The extraction system and the extraction method can accurately extract the lodging area of wheat, reduce the labor cost and are suitable for large-scale popularization and application.
Description
Technical Field
The invention relates to the field of ground observation and crop phenotype extraction based on unmanned aerial vehicle images, in particular to a wheat lodging area extraction system and method.
Background
Wheat is the grain crop with the largest sowing area and the largest yield in the world, and the growth vigor and the yield condition of the wheat influence the grain safety in the world. Lodging often occurs in the growth process of wheat, is a common agricultural disaster, causes a plurality of lodging reasons, and has the biggest influence caused by disastrous weather and strong and weak lodging resistance of wheat varieties. The lodging of wheat is easy to cause diseases and causes quality reduction, and the difficulty of mechanical harvesting is increased. Therefore, when lodging occurs, the yield evaluation of relevant departments can be influenced by untimely acquiring lodging information.
The manual survey method and the remote sensing image method are two common methods for acquiring the lodging area of the wheat at present. The manual investigation method is characterized in that workers observe, measure and count the crop lodging information on the spot, the method is time-consuming, labor-consuming and low in efficiency, and is only suitable for crop lodging investigation in a small range; the remote sensing image method is used for obtaining remote sensing images of the ground, different ground objects can present different texture differences and different spectral reflectivity, and the remote sensing image method is used for distinguishing the ground objects by analyzing the differences. The method can obtain a large-area wheat remote sensing image at one time, the efficiency is much higher than that of a human engineering method, but the remote sensing-based investigation method also has some problems: the acquired image is greatly influenced by weather, and the image formation is seriously influenced in cloudy and rainy and snowy weather; the re-turn period for a fixed location is relatively long and it cannot be ensured that the image of the study area can be completely acquired at one time.
In recent years, unmanned aerial vehicle remote sensing overcomes the defects of manual and satellite remote sensing by virtue of the advantages of easy platform construction, low cost, simple operation, high space-time resolution and the like, the processing method for the data of the remote sensing image of the small unmanned aerial vehicle is more and more mature, mass images can be acquired in time for processing, the unmanned aerial vehicle is suitable for complex farmland environments by acquiring the images, the acquired images with high resolution can be used for dividing the crop plot into cells for research, the research scale is more fine, and the unmanned aerial vehicle remote sensing is very suitable for crop lodging investigation. Unmanned aerial vehicles have become a main tool for rapidly and accurately acquiring crop information in agricultural quantitative remote sensing research, and are a hotspot and a trend of current research.
Disclosure of Invention
The present invention aims to solve at least to some extent at least one of the technical problems of the prior art. Therefore, the invention provides a system and a method for extracting the lodging area of wheat, and the system and the method can be used for accurately extracting the lodging area of wheat, reducing the labor cost and being suitable for large-scale popularization and application.
In one aspect of the invention, the invention provides a wheat lodging area extraction system. According to an embodiment of the invention, the system comprises:
the acquisition unit is used for acquiring field image data of a test area by carrying a multispectral sensor through an unmanned aerial vehicle;
the pretreatment unit is connected with the acquisition unit and is used for processing the acquired field image data of the test area to obtain a wheat canopy multispectral image;
the computing unit is connected with the preprocessing unit and is used for computing the texture characteristics of the target crop in the multispectral image of the wheat canopy;
and the selecting unit is connected with the calculating unit and is used for respectively combining the calculated textural features with the multispectral image, classifying and analyzing the image by using a Mahalanobis distance classification method and extracting the lodging area of the wheat.
According to the wheat lodging area extraction system provided by the embodiment of the invention, the unmanned aerial vehicle is adopted to obtain the wheat image data, the image data is analyzed, the texture characteristics are determined, and the texture characteristics and the image data are subjected to classification analysis so as to extract the wheat lodging area. Further, the inventors discovered that different classification analysis methods affect the accuracy value when comparing the classification analysis results with manual survey data to determine the classification model accuracy. Furthermore, the inventor conducts a large amount of theoretical analysis and experimental verification on various classification methods reported in the current research, and finds that the classification model obtained by adopting the Mahalanobis distance classification method has higher precision value, so that the accuracy of the finally obtained wheat lodging area is ensured.
According to the embodiment of the invention, the wheat lodging area extraction system can also have the following additional technical characteristics:
according to the embodiment of the invention, the field image data of the test area is acquired by carrying the multispectral sensor by the unmanned aerial vehicle at the flying height of 30-50 m and the flying time of 9:00-10:00, 13: 00-14: 00 and 17: 00-18: 00.
According to the embodiment of the invention, the computing unit is used for extracting the texture features of the obtained wheat canopy multispectral image by using the following method:
calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean and variance of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by using the two values.
According to an embodiment of the present invention, the selecting unit is configured to: selecting texture features with small variation coefficient and large relative difference coefficient as variables of the lodging extraction model; selecting an image obtained by combining the average texture characteristics of red, green and blue wave bands with the multispectral images of the flight height and the flight time period; and performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
According to an embodiment of the invention, the extraction system further comprises: and the evaluation unit is connected with the selection unit and used for comparing the manual survey data with the classification model structure obtained by the selection unit and verifying the accuracy of the classification model.
According to an embodiment of the invention, the evaluation unit is configured to: and calculating a difference value by using the lodging area ratio in each cell calculated by the selection unit and the known manually-checked lodging area ratio, wherein the difference value is a classification error.
In another aspect of the invention, the invention provides a method for extracting the lodging area of wheat by using the wheat lodging area extraction system. According to an embodiment of the invention, the method comprises the steps of: s1: collecting field image data of a test area by using an unmanned aerial vehicle carrying a multispectral sensor; s2: preprocessing field image data collected in S1 to obtain a wheat canopy multispectral image; s3: calculating the texture characteristics of the target crop in the multispectral image of the wheat canopy obtained by S2; s4: selecting the average texture characteristics of red, green and blue wave bands, combining the multispectral images of the flight height and the flight time period to obtain an image, classifying and analyzing the image by using a Mahalanobis distance classification method, and extracting the lodging area of the wheat. Therefore, the lodging area of wheat can be accurately extracted by using the method provided by the embodiment of the invention, the labor cost is reduced, and the method is suitable for large-scale popularization and application.
According to the embodiment of the invention, the flying height of the multispectral sensor carried by the unmanned aerial vehicle is 30-50 m, and the flying time is 9:00-10:00, 13: 00-14: 00 and 17: 00-18: 00.
According to the embodiment of the invention, the preprocessing of S2 includes image stitching, geometric correction and radiometric calibration.
According to an embodiment of the present invention, the S3 includes the steps of: and (5) extracting the texture characteristics of the lodging wheat and the normal wheat by using the following method for the multispectral image of the wheat canopy obtained in the step S2: calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean value and the variance of a plurality of texture characteristics of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by utilizing the two values.
According to an embodiment of the present invention, the S4 includes the steps of: s41: selecting texture features with small variation coefficient and large relative difference coefficient as variables of the lodging extraction model; s42: selecting an image obtained by combining the average texture characteristics of red, green and blue wave bands with the multispectral images of the flight height and the flight time period; s43: and performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
According to an embodiment of the invention, the method further comprises: s5: and comparing the wheat lodging area result obtained in the step S4 with manual survey data, and verifying the precision of the classification model.
According to an embodiment of the present invention, the S5 includes the steps of: the ratio of the lodging area in each cell calculated in S46 is compared with the known manually found ratio of the lodging area to calculate a difference, which is a classification error.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 shows a schematic diagram of a wheat lodging area extraction system according to one embodiment of the invention;
fig. 2 shows a flow chart of a wheat lodging area extraction method according to one embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides a wheat lodging area extraction system and a method for extracting a wheat lodging area by using the same, which are respectively described in detail below.
Wheat lodging area extraction system
In one aspect of the invention, the invention provides a wheat lodging area extraction system. According to an embodiment of the invention, referring to fig. 1, the system comprises: the acquiring unit 100, the preprocessing unit 200, the calculating unit 300 and the selecting unit 400, which will be described in detail below.
According to an embodiment of the present invention, the obtaining unit 100 is configured to obtain field image data of a test area by mounting a multispectral sensor on an unmanned aerial vehicle. Adopt unmanned aerial vehicle to obtain image data, it is more convenient, reduced the cost of labor, avoid the human factor to cause image data inaccurate.
According to the embodiment of the invention, the field image data of the test area is acquired by carrying the multispectral sensor by the unmanned aerial vehicle at the flying height of 30-50 m and the flying time of 9:00-10:00, 13: 00-14: 00 and 17: 00-18: 00. The inventor finds that the flying height and flying time of the unmanned aerial vehicle can influence the precision of the classification model, thereby influencing the accuracy of the final result. Furthermore, a large number of experimental researches show that the flying height is 50 meters, the flying time is 9:00-10:00 in the morning, the precision of the classification model is high, and the extracted lodging area of the wheat is accurate.
According to an embodiment of the present invention, the preprocessing unit 200 is connected to the obtaining unit 100, and is configured to process the obtained field image data of the test area to obtain a wheat canopy multispectral image. In each part of wheat, the canopy image information can reflect the lodging state of wheat, and further, the canopy image information is used as a research basis.
According to the embodiment of the present invention, the calculating unit 300 is connected to the preprocessing unit 200, and is configured to calculate the texture features of the target crop in the multispectral image of the wheat canopy.
According to the embodiment of the present invention, the calculating unit 300 is configured to extract the texture features of the obtained wheat canopy multispectral image by using the following method:
calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean and variance of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by using the two values.
The texture features mainly comprise 8 types of mean values, variances, homogeneity, contrast, dissimilarities, information entropies, second moments and correlations, and when the texture features are analyzed each time, the 8 types of texture features are required to be synthesized with the multispectral image respectively, and the synthesized image is supervised, classified and checked for precision by using a Mahalanobis distance method. In the process, the inventor finds that the classification model precision can be ensured by analyzing the mean value and the variance as the texture features, so that the mean value and the variance of normal and lodging wheat can be calculated only by analyzing each time. In addition, the coefficient of variation is related to intra-feature differences, and the relative coefficient of difference is related to inter-feature differences.
According to the embodiment of the invention, the selecting unit 400 is connected with the calculating unit 300 and is used for respectively merging the calculated textural features with the multispectral image, classifying and analyzing the image by using a mahalanobis distance classification method, and extracting the lodging area of the wheat. The inventors have found that different classification analysis methods affect the accuracy value. Furthermore, the inventor conducts a large amount of theoretical analysis and experimental verification on various classification methods reported in the current research, and finds that the classification model obtained by adopting the Mahalanobis distance classification method has higher precision value, so that the accuracy of the finally obtained wheat lodging area is ensured.
According to an embodiment of the invention, the selection unit is configured to:
the smaller the difference in the features is, the larger the difference between the features is, the stronger the classification capability of the features is, so that the texture features with small variation coefficient and large relative difference coefficient are selected as variables of the lodging extraction model;
embedding a new grid tool in ArcGIS, merging the selected texture features in different combination modes to obtain a new image, selecting a training sample and a verification sample from the new image, wherein the verification sample cannot be repeated with the training sample and has a pixel value of 1/2 of the training sample, and performing classification analysis on the merged image by using a Mahalanobis distance classification method;
performing precision verification on each combination mode by using a confusion matrix, performing precision evaluation by using the obtained overall classification precision and a Kappa coefficient thereof as a reference standard, and finally selecting an image obtained by combining mean texture characteristics of red, green and blue wave bands with a multispectral image;
embedding a new grid tool in ArcGIS, merging the mean texture characteristics of red, green and blue wave bands and multispectral images with different flight heights and flight time periods to obtain a new image, selecting a training sample and a verification sample from the new image, and performing classification analysis on the merged image by using a Mahalanobis distance classification method;
performing precision verification on each combination mode by using a confusion matrix, performing precision evaluation by using the obtained overall classification precision and a Kappa coefficient thereof as a reference standard, and finally selecting images obtained by combining mean texture characteristics of red, green and blue wave bands with the flying height and multispectral images of the flying time period;
and performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
According to an embodiment of the invention, the extraction system further comprises: and the evaluation unit is connected with the selection unit and used for comparing the manual survey data with the classification model structure obtained by the selection unit and verifying the accuracy of the classification model. Thereby, whether the final extraction lodging area result is accurate or not is determined.
According to an embodiment of the invention, the evaluation unit is adapted to: and calculating a difference value by using the lodging area ratio in each cell calculated by the selection unit and the known manually-checked lodging area ratio, wherein the difference value is a classification error.
Method for extracting lodging area of wheat
In another aspect of the invention, the invention provides a method for extracting the lodging area of wheat by using the wheat lodging area extraction system. According to an embodiment of the invention, referring to fig. 2, the method comprises:
s1 unmanned plane collecting test area image
In the step, an unmanned aerial vehicle is used for carrying a multispectral sensor to collect field image data of the test area.
According to the embodiment of the invention, the flying height of the multispectral sensor carried by the unmanned aerial vehicle is 30-50 m, and the flying time is 9:00-10:00, 13: 00-14: 00 and 17: 00-18: 00. Therefore, the accuracy of the final result is ensured to be high.
S2 preprocessing to obtain the multispectral image of the wheat canopy
In the step, the field image data collected in the step S1 is preprocessed to obtain a wheat canopy multispectral image.
According to the embodiment of the invention, the preprocessing of S2 includes image stitching, geometric correction and radiometric calibration.
S3 calculating texture features
In this step, the texture features of the target crop in the multispectral image of the wheat canopy obtained in S2 are calculated.
According to an embodiment of the present invention, S3 includes the steps of: and (5) extracting the texture characteristics of the lodging wheat and the normal wheat by using the following method for the multispectral image of the wheat canopy obtained in the step S2: calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean value and the variance of a plurality of texture characteristics of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by utilizing the two values. Thereby, the texture feature can be calculated.
S4 classifying and analyzing the images by using the Mahalanobis distance classification method, and extracting the lodging area of the wheat
In the step, the average texture characteristics of red, green and blue wave bands are selected and combined with the multispectral images of the flight height and the flight time period to obtain an image, the image is classified and analyzed by using a Mahalanobis distance classification method, and the lodging area of the wheat is extracted.
According to an embodiment of the present invention, S4 includes the steps of:
s41, selecting texture features with small variation coefficient and large relative difference coefficient as variables of the lodging extraction model;
s42, selecting average texture characteristics of red, green and blue wave bands, combining the average texture characteristics with the flying height and the multispectral images of the flying time period, and merging the average texture characteristics with the multispectral images to obtain an image;
and S46, performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
According to an embodiment of the invention, the method further comprises: s5: and comparing the wheat lodging area result obtained in the step S4 with manual survey data, and verifying the precision of the classification model. Therefore, whether the obtained wheat lodging area result is accurate or not is judged conveniently.
According to an embodiment of the present invention, S5 includes the steps of: the ratio of the lodging area in each cell calculated in S46 is compared with the known manually found ratio of the lodging area to calculate a difference, which is a classification error.
It will be understood by those skilled in the art that the features and advantages described above for the wheat lodging area extraction system are equally applicable to the method for extracting wheat lodging area, and will not be described in detail herein.
The scheme of the invention will be explained with reference to the examples. It will be appreciated by those skilled in the art that the following examples are illustrative of the invention only and should not be taken as limiting the scope of the invention. The examples, where specific techniques or conditions are not indicated, are to be construed according to the techniques or conditions described in the literature in the art or according to the product specifications. The reagents or instruments used are not indicated by the manufacturer, and are all conventional products commercially available.
Example 1
1. Procedure of experiment
1.1 aerial image
And (3) using an unmanned aerial vehicle to carry a multispectral sensor to obtain a field winter wheat image of the test area.
1.2 image analysis
1.2.1 pretreatment
And preprocessing the collected field image data, including image splicing, geometric correction and radiometric calibration, to obtain the wheat canopy multispectral image.
1.2.2 computing texture features
Extracting the texture characteristics of the obtained multispectral image of the winter wheat canopy by using the following method:
calculating texture information of normal winter wheat and lodging winter wheat by using a filtering tool based on second-order probability statistics, setting a processing window of filtering to be 7X 7, taking a reference window as reference, wherein the transformation values X and Y of the spatial correlation matrix are both 1, and the gray quality level is 64; after texture analysis, the mean value and variance of normal and lodging winter wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by using the two values.
1.2.3 Classification analysis
The smaller the difference in the features is, the larger the difference between the features is, the stronger the classification capability of the features is, and the texture features with small variation coefficient and large relative difference coefficient are selected as variables of the lodging extraction model;
embedding a new grid tool in ArcGIS, merging the selected texture features in different combination modes to obtain a new image, selecting a training sample and a verification sample which are well evaluated, wherein the verification sample cannot be repeated with the training sample and has a pixel value of 1/2 of the training sample, and performing supervision and classification on the merged image by using a Mahalanobis distance classification method;
performing precision verification on each combination mode by using a confusion matrix, performing precision evaluation by using the obtained overall classification precision and a Kappa coefficient thereof as a reference standard, and finally selecting an image obtained by combining mean texture characteristics of red, green and blue wave bands with a multispectral image;
embedding a new grid tool in ArcGIS, merging the mean texture characteristics of red, green and blue wave bands and multispectral images with different flight heights and flight time periods to obtain a new image, selecting a training sample and a verification sample which are well evaluated, and performing supervision and classification on the merged image by using a Mahalanobis distance classification method;
performing precision verification on each combination mode by using a confusion matrix, performing precision evaluation by using the obtained overall classification precision and a Kappa coefficient thereof as a reference standard, and finally selecting images obtained by combining the mean texture characteristics of red, green and blue wave bands with multispectral images with the flying height of 50m and the flying time interval of 9:00-10:00 in the morning;
and performing superposition analysis on the classified lodging plots and the cell boundaries to enable the same image to have lodging information and cell boundary information, counting the size of the lodging area in each cell, and calculating the lodging area ratio in each cell.
1.2.4 evaluation
And calculating a difference value by using the calculated lodging area ratio in each cell and the known manually-found lodging area ratio, wherein the difference value is a classification error.
EXAMPLE 2 Effect of different fly heights
The winter wheat lodging areas were extracted according to the method of example 1, wherein the flying heights were 30 meters, 40 meters, and 50 meters, respectively. The obtained classification error results are shown in the following table, and the results show that the effect is best when the flying height is 50 meters.
| Overall classification accuracy | Kappa coefficient | |
| RGB1_30m | 96.9182％ | 0.9286 |
| RGB1_40m | 98.4609％ | 0.9650 |
| RGB1_50m | 99.4602％ | 0.9875 |
EXAMPLE 3 Effect of different flight times
The lodging area of winter wheat is extracted according to the method in the embodiment 1, wherein the flight time is 9:00-10:00 (RGB1) in the morning, 13: 00-14: 00(RGB2) in the noon and 17: 00-18: 00(RGB3) in the afternoon respectively. The obtained classification error result is shown in the following table, and the result shows that the effect is best when the flight time is 9:00-10:00 in the morning.
| Overall classification accuracy | Kappa coefficient | |
| RGB1_50m | 99.4602％ | 0.9875 |
| RGB2_50m | 95.5221％ | 0.8990 |
| RGB3_50m | 99.2305％ | 0.9822 |
Example 4 different Classification analysis methods
The lodging area of winter wheat is extracted according to the method of the embodiment 1, wherein the classification analysis method is a maximum likelihood method, a minimum distance method, a mahalanobis distance method, a support vector machine method and a parallelepiped method respectively. The obtained classification error results are shown in the following table, and the results show that the result is the best when the Mahalanobis distance method is adopted for analysis.
| Overall classification accuracy | Kappa coefficient | |
| Maximum likelihood method | 99.4602％ | 0.9875 |
| Minimum distance method | 88.7289％ | 0.7374 |
| Mahalanobis distance method | 99.1494％ | 0.9802 |
| Support vector machine method | 99.8201％ | 0.9958 |
| Parallelepiped method | 86.1443％ | 0.6385 |
Note: the used image is an unmanned aerial vehicle multispectral image obtained at a flying height of 50 meters in the morning of 9:00-10: 00.
EXAMPLE 4 different texture features
The lodging area of winter wheat was extracted according to the method of example 1, wherein the textural features were mean, variance, homogeneity, contrast, dissimilarity, entropy, second moment and correlation, respectively. The obtained classification error results are shown in the following table, and the results show that the mean value and the variance are used as texture characteristics, so that the effect is optimal.
| Overall classification accuracy | Kappa coefficient | Overall classification accuracy | Kappa coefficient | ||
| RGB + mean value | 99.9836％ | 0.9996 | RGB + variance | 98.2006％ | 0.9578 |
| RGB + homogeneity | 99.2802％ | 0.9832 | RGB + contrast ratio | 98.2823％ | 0.9596 |
| RGB + dissimilarity | 99.3947％ | 0.9859 | RGB + entropy | 99.4602％ | 0.9874 |
| RGB + second moment | 99.2802％ | 0.9832 | RGB + correlation | 99.2802％ | 0.9832 |
Note: the used image is an unmanned aerial vehicle multispectral image obtained at a flying height of 50 meters in the morning of 9:00-10: 00.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (10)
1. A wheat lodging area extraction system, comprising:
the acquisition unit is used for acquiring field image data of a test area by carrying a multispectral sensor through an unmanned aerial vehicle;
the pretreatment unit is connected with the acquisition unit and is used for processing the acquired field image data of the test area to obtain a wheat canopy multispectral image;
the computing unit is connected with the preprocessing unit and is used for computing the texture characteristics of the target crop in the multispectral image of the wheat canopy;
and the selecting unit is connected with the calculating unit and is used for respectively combining the calculated textural features with the multispectral image, classifying and analyzing the image by using a Mahalanobis distance classification method and extracting the lodging area of the wheat.
2. The extraction system according to claim 1, wherein the field image data of the test area is acquired by carrying a multispectral sensor by an unmanned aerial vehicle at a flying height of 30-50 m and a flying time of 9:00-10:00, 13: 00-14: 00 and 17: 00-18: 00.
3. The extraction system according to claim 1, wherein the computing unit is configured to extract the texture features from the obtained wheat canopy multispectral image by using the following method:
calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean and variance of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by using the two values.
4. The extraction system according to claim 1, wherein the extraction unit is configured to:
selecting texture features with small variation coefficient and large relative difference coefficient as variables of the lodging extraction model;
selecting an image obtained by combining the average texture characteristics of red, green and blue wave bands with the multispectral images of the flight height and the flight time period;
and performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
5. The extraction system according to claim 1, further comprising:
the evaluation unit is connected with the selection unit and used for comparing manual survey data with the classification model structure obtained by the selection unit and verifying the accuracy of the classification model;
optionally, the evaluation unit is configured to:
and calculating a difference value by using the lodging area ratio in each cell calculated by the selection unit and the known manually-checked lodging area ratio, wherein the difference value is a classification error.
6. A method for extracting a wheat lodging area by using the wheat lodging area extraction system as claimed in any one of claims 1 to 5, which is characterized by comprising the following steps:
s1: collecting field image data of a test area by using an unmanned aerial vehicle carrying a multispectral sensor;
s2: preprocessing field image data collected in S1 to obtain a wheat canopy multispectral image;
s3: calculating the texture characteristics of the target crop in the multispectral image of the wheat canopy obtained by S2;
s4: selecting the average texture characteristics of red, green and blue wave bands, combining the multispectral images of the flight height and the flight time period to obtain an image, classifying and analyzing the image by using a Mahalanobis distance classification method, and extracting the lodging area of the wheat.
7. The extraction method according to claim 6, wherein the flying height of the unmanned aerial vehicle carrying the multispectral sensor is 30-50 m, the flying time is 9:00-10:00, 13: 00-14: 00, 17: 00-18: 00;
optionally, the preprocessing of S2 includes image stitching and geometric correction, and radiometric calibration.
8. The extraction method according to claim 6, wherein the S3 includes the steps of:
and (5) extracting the texture characteristics of the lodging wheat and the normal wheat by using the following method for the multispectral image of the wheat canopy obtained in the step S2:
calculating texture information of normal wheat and lodging wheat by using a filtering tool based on second-order probability statistics; after texture analysis, the mean value and the variance of a plurality of texture characteristics of normal and lodging wheat need to be counted, and the variation coefficient and the relative difference coefficient are calculated by utilizing the two values.
9. The extraction method according to claim 8, wherein the S4 includes the steps of:
s41: selecting texture features with small variation coefficient and large relative difference coefficient as variables of the lodging extraction model;
s42: selecting an image obtained by combining the average texture characteristics of red, green and blue wave bands with the multispectral images of the flight height and the flight time period;
s43: and performing superposition analysis on the classified lodging plots and cell boundaries to ensure that the lodging plots also have cell boundary information, counting the size of lodging areas in each cell, and calculating the lodging area ratio in each cell.
10. The extraction method according to claim 9, further comprising:
s5: comparing the wheat lodging area result obtained in the step S4 with manual investigation data, verifying the precision of the classification model,
optionally, the S5 includes the following steps:
the ratio of the lodging area in each cell calculated in S46 is compared with the known manually found ratio of the lodging area to calculate a difference, which is a classification error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010845296.2A CN114170500A (en) | 2020-08-20 | 2020-08-20 | Wheat lodging area extraction system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010845296.2A CN114170500A (en) | 2020-08-20 | 2020-08-20 | Wheat lodging area extraction system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114170500A true CN114170500A (en) | 2022-03-11 |
Family
ID=80475329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010845296.2A Pending CN114170500A (en) | 2020-08-20 | 2020-08-20 | Wheat lodging area extraction system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114170500A (en) |
-
2020
- 2020-08-20 CN CN202010845296.2A patent/CN114170500A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pang et al. | Improved crop row detection with deep neural network for early-season maize stand count in UAV imagery | |
| CN111091052A (en) | Corn lodging area extraction system and method based on maximum likelihood method | |
| CN112101159B (en) | Multi-temporal forest remote sensing image change monitoring method | |
| CN110926430B (en) | Air-ground integrated mangrove forest monitoring system and control method | |
| CN113392775B (en) | Sugarcane seedling automatic identification and counting method based on deep neural network | |
| CN109325431B (en) | Method and device for detecting vegetation coverage in feeding path of grassland grazing sheep | |
| CN110363246B (en) | Fusion method of vegetation index NDVI with high space-time resolution | |
| CN104266982A (en) | Large-area insect pest quantization monitoring system | |
| CN113029971B (en) | Crop canopy nitrogen monitoring method and system | |
| CN102855485B (en) | The automatic testing method of one grow wheat heading | |
| US6990410B2 (en) | Cloud cover assessment: VNIR-SWIR | |
| CN111860150B (en) | Lodging rice identification method and device based on remote sensing image | |
| Xu et al. | Classification method of cultivated land based on UAV visible light remote sensing | |
| Aplin et al. | Predicting missing field boundaries to increase per-field classification accuracy | |
| CN111985445A (en) | Grassland insect pest monitoring system and method based on unmanned aerial vehicle multispectral remote sensing | |
| WO2023197496A1 (en) | Comprehensive evaluation indicator monitoring and evaluation method and system for machine-harvested cotton defoliation effects | |
| Yang et al. | Cotton hail disaster classification based on drone multispectral images at the flowering and boll stage | |
| CN109960972B (en) | Agricultural and forestry crop identification method based on middle-high resolution time sequence remote sensing data | |
| CN112924967A (en) | Remote sensing monitoring method for crop lodging based on radar and optical data combination characteristics and application | |
| CN114170500A (en) | Wheat lodging area extraction system and method | |
| CN116503740A (en) | Unmanned aerial vehicle vision system capable of accurately identifying crop types | |
| CN116416532A (en) | Disease strain diagnosis and strain shortage detection method, system, equipment and storage medium | |
| Wallace et al. | Analysis of remotely sensed data | |
| Akila et al. | Automation in plant growth monitoring using high-precision image classification and virtual height measurement techniques | |
| CN115372281A (en) | Monitoring system and method for physical structure and chemical composition of soil |
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 |