CN116229261A - Wheat identification method, device and storage medium - Google Patents
Wheat identification method, device and storage medium Download PDFInfo
- Publication number
- CN116229261A CN116229261A CN202310146482.0A CN202310146482A CN116229261A CN 116229261 A CN116229261 A CN 116229261A CN 202310146482 A CN202310146482 A CN 202310146482A CN 116229261 A CN116229261 A CN 116229261A
- Authority
- CN
- China
- Prior art keywords
- wheat
- training
- evi
- ndvi
- identification
- 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 78
- 235000021307 Triticum Nutrition 0.000 title claims abstract description 78
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000003860 storage Methods 0.000 title claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 59
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 44
- 238000001914 filtration Methods 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000004140 cleaning Methods 0.000 claims abstract description 12
- 238000004590 computer program Methods 0.000 claims description 20
- 230000009467 reduction Effects 0.000 claims description 13
- 238000007637 random forest analysis Methods 0.000 claims description 9
- 238000000513 principal component analysis Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 6
- 238000012805 post-processing Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 5
- 235000002566 Capsicum Nutrition 0.000 claims description 4
- 239000006002 Pepper Substances 0.000 claims description 4
- 235000016761 Piper aduncum Nutrition 0.000 claims description 4
- 235000017804 Piper guineense Nutrition 0.000 claims description 4
- 235000008184 Piper nigrum Nutrition 0.000 claims description 4
- 102000006463 Talin Human genes 0.000 claims description 4
- 108010083809 Talin Proteins 0.000 claims description 4
- 238000009499 grossing Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 150000003839 salts Chemical class 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 2
- 244000203593 Piper nigrum Species 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 241000722363 Piper Species 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000007635 classification algorithm Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a wheat identification method, a device and a storage medium, which are used for more effectively utilizing time sequence data so as to realize the identification of winter wheat with high accuracy. According to the wheat identification method provided by the invention, firstly, a preset algorithm is trained by using training data to obtain a wheat identification model, then, a preset area is identified by using the wheat identification model, and then, boundary cleaning and mode filtering processing are carried out on the identified result to obtain a wheat identification result.
Description
Technical Field
The present disclosure relates to the field of computing technologies, and in particular, to a method and apparatus for identifying wheat, and a storage medium.
Background
The main technology of crop identification is to identify crops by using their unique time series characteristics of vegetation indexes. The vegetation index reflects the response characteristics of the plant to different band spectra. Because of the existence of the foreign matter homospectrum condition, accurate identification of crops is difficult to realize by a single-period remote sensing image, and therefore, in the prior art, the crop identification mainly takes a vegetation index time sequence of the whole growth period of the crops as a research object. The most widely used vegetation indexes, which can reflect the vegetation growth condition best, are mainly normalized vegetation index (NDVI) and Enhanced Vegetation Index (EVI).
However, in the prior art, the change trend of the vegetation index time sequence curve is not fully considered, and the error division is easily caused by extracting the wheat planting area only by the threshold value.
Disclosure of Invention
Aiming at the technical problems, the embodiment of the application provides a wheat identification method, a device and a storage medium, which are used for improving the accuracy of wheat identification.
In a first aspect, an embodiment of the present application provides a wheat identification method, including:
training a preset algorithm by using training data to obtain a wheat identification model;
identifying a preset area by using the wheat identification model;
and carrying out boundary cleaning and mode filtering treatment on the identified result to obtain a wheat identification result.
Preferably, the training the preset algorithm using the training data to obtain the wheat recognition model includes:
acquiring satellite images of the whole growth period of wheat, removing cloud images, and calculating normalized vegetation index NDVI and enhanced vegetation index EVI of the rest images;
carrying out mask extraction on the grid data of the NDVI and the EVI by using sample plot data acquired in the field to respectively obtain NDVI and EVI time sequence curves of a wheat sample and other samples;
constructing importance indexes of the NDVI and the EVI, and performing characteristic dimension reduction;
sequencing the importance indexes, and selecting vegetation indexes corresponding to the N most important importance indexes as the characteristics of model training, wherein N is an integer greater than or equal to 1;
training a preset algorithm by taking the time sequence curve as the training data;
wherein the importance index is used for measuring the importance of the NDVI and/or the EVI, and the higher the value of the importance index is, the more important the NDVI and/or the EVI is.
In the method, firstly, the invention calculates the NDVI and EVI time sequence curves of the whole test area, namely, the defect that the NDVI is easy to saturate when the vegetation coverage is high is overcome, the defect that the EVI grows slowly when the vegetation coverage is low and is sensitive to blue buildings is avoided, the vegetation index curve of winter wheat in the whole growing period can be well distinguished from other features, and the identification precision is improved on the basis. And then taking the time sequence curve as the training data to train a preset algorithm. And identifying the preset area by using the wheat identification model. And finally, carrying out boundary cleaning and mode filtering treatment on the identified result to obtain a wheat identification result.
Preferably, the removing the cloud image includes:
and eliminating the influence of the cloud coverage exceeding a preset threshold.
Preferably, the calculating the normalized vegetation index NDVI and the enhanced vegetation index EVI of the remaining images includes:
the NDVI is calculated by the following formula:
the EVI is calculated by the following formula:
wherein Band4 is the reflection value of the red Band, band8 is the reflection value of the near infrared Band, and Band2 is the reflection value of the blue Band.
Other samples include data for one or a combination of:
other crops, vegetation, bare land, buildings.
The feature dimension reduction comprises the following steps:
discarding the features of which the importance index is lower than a preset first threshold value; or alternatively
And carrying out characteristic dimension reduction on the NDVI and the EVI according to Principal Component Analysis (PCA).
Preferably, the preset algorithm includes:
the preset algorithm is a random forest algorithm with grid search.
Preferably, the training data is the time sequence curve, and after training the preset algorithm, the method further includes:
modifying the values of parameters in the grid search until the accuracy meets the preset requirement;
the parameters include one or a combination of the following:
the number of trees;
the maximum depth of the tree.
The training data is the time sequence curve, and after training a preset algorithm, the training data further comprises:
and verifying the trained model by using verification data, and calculating the accuracy.
Preferably, the boundary cleaning of the identified result includes:
and smoothing boundary saw teeth at different category junctions in the identification result.
Preferably, performing mode filtering processing on the identified result includes:
and removing the salt and pepper noise in the identification result through the mode filtering processing.
In a second aspect, embodiments of the present application further provide a wheat identification apparatus, including:
the training module is configured to train a preset algorithm by using training data to obtain a wheat recognition model;
the identification module is configured to identify a preset area by using the wheat identification model;
and the post-processing module is configured to perform boundary cleaning and mode filtering processing on the identified result to obtain a wheat identification result.
In a third aspect, an embodiment of the present application further provides a wheat identification apparatus, including: a memory, a processor, and a user interface;
the memory is used for storing a computer program;
the user interface is used for realizing interaction with a user;
the processor is used for reading the computer program in the memory, and when the processor executes the computer program, the wheat identification method provided by the invention is realized.
In a fourth aspect, an embodiment of the present application further provides a processor readable storage medium, where a computer program is stored in the processor readable storage medium, and when the processor executes the computer program, the wheat identification method provided by the present invention is implemented.
According to the wheat recognition method, the NDVI and EVI are utilized to establish the time sequence characteristic curve of winter wheat, the dimension of the characteristics is reduced according to the importance sequence of the vegetation index characteristics, and partial characteristics which do not help or even affect the recognition accuracy are removed, so that on one hand, the training speed of a model is improved, on the other hand, the accuracy of the model is improved, and the accuracy of winter wheat recognition is improved. Meanwhile, a random forest algorithm with grid search is utilized for model training, so that the accuracy of wheat identification is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of a wheat identification method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a training flow of a wheat recognition model according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a wheat identification apparatus according to an embodiment of the present invention;
fig. 4 is a schematic diagram of another wheat identification apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Some words appearing hereinafter are explained:
1. in the embodiment of the invention, the term "and/or" describes the association relation of the association objects, which means that three relations can exist, for example, a and/or B can be expressed as follows: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.
2. The term "plurality" in the embodiments of the present application means two or more, and other adjectives are similar thereto.
The following description of the technical solutions in the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
It should be noted that, the display sequence of the embodiments of the present application only represents the sequence of the embodiments, and does not represent the advantages or disadvantages of the technical solutions provided by the embodiments.
Referring to fig. 1, a schematic diagram of a wheat identification method according to an embodiment of the present application is shown in fig. 1, and the method includes steps S101 to S103:
s101, training a preset algorithm by using training data to obtain a wheat identification model;
as a preferred example, the preset algorithm may be a classification algorithm, which is a machine learning supervised classification algorithm, and may include, for example, maximum likelihood, neural networks, decision trees, random forests, support vector machines, etc. As a preferred example, the training process may be: firstly, selecting a sample, and matching with a remote sensing image by using measured data or other data sources to obtain sample data; secondly, carrying out algorithm training, and obtaining optimal parameters by using training samples; and verifying by a third algorithm, and verifying the accuracy of the algorithm by using verification data.
As a preferred example, S101 may specifically include the steps as shown in fig. 2:
s201, acquiring satellite images of the whole growth period of wheat, removing cloud images, and calculating normalized vegetation indexes NDVI and enhanced vegetation indexes EVI of the rest images;
it should be noted that in the embodiment of the present invention, satellite images of the whole growth cycle of wheat are acquired instead of satellite images of only part of the growth cycle of wheat. That is, the invention calculates the NDVI and EVI time sequence curves of the whole growth period of the whole test area, i.e. overcomes the defect that the NDVI is easy to saturate when the vegetation coverage is high, and also avoids the defect that the EVI grows slowly when the vegetation coverage is low and is sensitive to blue buildings, so that the vegetation index curve of winter wheat in the whole growth period can be well distinguished from other ground objects, and the recognition precision is improved on the basis.
As a preferred example, in the present invention, eliminating the cloud image may include:
and eliminating the influence of the cloud coverage exceeding a preset threshold.
In the present invention, the cloud coverage means a ratio of an area covered by the cloud to a total area, and the unit is a percentage. The preset threshold is a cloud coverage value which is preset, and is determined according to specific requirements, and the specific value is not limited by the invention. For example, 20% or 15%.
The remaining images are images remaining after the cloud images are removed.
As a preferred example, in the present invention, calculating the normalized vegetation index NDVI and the enhanced vegetation index EVI of the remaining images includes:
the NDVI is calculated by the following formula:
the EVI is calculated by the following formula:
wherein Band4 is the reflection value of the red Band, band8 is the reflection value of the near infrared Band, and Band2 is the reflection value of the blue Band.
S202, carrying out mask extraction on the raster data of the NDVI and the EVI by using sample plot data acquired in the field to respectively obtain NDVI and EVI time sequence curves of a wheat sample and other samples;
in this step, the raster data of the NDVI and the EVI refers to data stored in the format of raster data.
It should be noted that, in the embodiment of the present invention, the other samples refer to samples other than the wheat sample, for example, samples that may include one or a combination of the following: other crops, vegetation, bare land, buildings. Wherein, other crops refer to crops other than wheat.
In the embodiment of the present invention, mask extraction refers to performing cut extraction on the raster data obtained in S202 by using vector data with boundaries.
S203, constructing importance indexes of the NDVI and the EVI, and performing characteristic dimension reduction;
wherein, the importance index is used for measuring the importance of the NDVI and/or the EVI, and the higher the value of the importance index is, the more important the NDVI and/or the EVI is.
As a preferred example, performing feature dimension reduction may include:
discarding the features of which the importance index is lower than a preset first threshold value; or alternatively
And carrying out characteristic dimension reduction on the NDVI and the EVI according to Principal Component Analysis (PCA).
In the invention, the feature is subjected to dimension reduction according to the importance sequence of the vegetation index features (namely the sequence of the importance indexes), and partial features which do not help to the recognition and even influence the recognition accuracy are removed, so that the training speed of the model is improved on one hand, and the accuracy of the model is also improved on the other hand.
S204, sorting the importance indexes, and selecting vegetation indexes corresponding to the N most important importance indexes as the characteristics of model training;
wherein N is an integer of 1 or more, and specific numerical values are determined in advance as needed, and the present invention is not particularly limited.
S205, training a preset algorithm by taking the time sequence curve as the training data;
as a preferred example, the preset algorithm includes a machine learning algorithm that can implement classification, such as a random forest algorithm with grid search, and other machine learning algorithms that can implement classification.
Taking a random forest algorithm with network searching as an example, the grid searching can set a plurality of values for each parameter, and then an optimal solution of each parameter in a given value is selected by using a permutation and combination mode, so that compared with a machine learning algorithm with fixed parameters, the model iterative optimization speed is greatly improved. Meanwhile, compared with a curve integration method and other machine learning algorithms, the random forest algorithm can train a high-accuracy classifier in a short learning process in the classification problem of large data quantity and large feature number.
As a preferred example, after S205, it may further include:
modifying the values of parameters in the grid search until the accuracy meets the preset requirement;
the parameters include one or a combination of the following:
the number of trees;
the maximum depth of the tree.
As a preferred example, after S205, it may further include:
and verifying the trained model by using verification data, and calculating the accuracy. Preferably, both the training data and the validation data are derived from field samples, and may be distributed at a predetermined ratio, for example at a ratio of 8:2, i.e. 80% field samples are used as training data and 20% field samples are used as validation data.
S102, identifying a preset area by using the wheat identification model;
s103, performing boundary cleaning and mode filtering processing on the identified result to obtain a wheat identification result.
As a preferred example, in the present invention, boundary cleaning of the identified result may include:
and smoothing boundary saw teeth at the junctions of different categories in the identification result, wherein the different categories refer to different object categories, such as different crops, crops and non-crops, different non-crops and the like.
As a preferred example, in the present invention, the mode filtering processing of the identified result may include: and removing the salt and pepper noise in the identification result through the mode filtering processing.
According to the wheat recognition method, the NDVI and EVI are utilized to establish the time sequence characteristic curve of winter wheat, the dimension of the characteristics is reduced according to the importance sequence of the vegetation index characteristics, and partial characteristics which do not help or even affect the recognition accuracy are removed, so that on one hand, the training speed of a model is improved, on the other hand, the accuracy of the model is improved, and the accuracy of winter wheat recognition is improved. Meanwhile, a random forest algorithm with grid search is utilized for model training, so that the accuracy of wheat identification is improved.
Based on the same inventive concept, the embodiment of the invention also provides a wheat identification device, as shown in fig. 3, which comprises:
the training module 301 is configured to train a preset algorithm by using training data to obtain a wheat recognition model;
the identifying module 302 is configured to identify a preset area by using the wheat identifying model;
and the post-processing module 303 is configured to perform boundary cleaning and mode filtering processing on the identified result to obtain a wheat identification result.
As a preferred example, training module 301 is further configured to:
acquiring satellite images of the whole growth period of wheat, removing cloud images, and calculating normalized vegetation index NDVI and enhanced vegetation index EVI of the rest images;
carrying out mask extraction on the grid data of the NDVI and the EVI by using sample plot data acquired in the field to respectively obtain NDVI and EVI time sequence curves of a wheat sample and other samples;
constructing importance indexes of the NDVI and the EVI, and performing characteristic dimension reduction;
sequencing the importance indexes, and selecting vegetation indexes corresponding to the N most important importance indexes as the characteristics of model training;
training a preset algorithm by taking the time sequence curve as the training data;
wherein the importance index is used for measuring the importance of the NDVI and/or the EVI, and the higher the value of the importance index is, the more important the NDVI and/or the EVI is;
n is an integer greater than or equal to 1.
Wherein, reject the cloud image and include:
and eliminating the influence of the cloud coverage exceeding a preset threshold.
The calculating the normalized vegetation index NDVI and the enhanced vegetation index EVI of the remaining images includes:
the NDVI is calculated by the following formula:
the EVI is calculated by the following formula:
wherein Band4 is the reflection value of the red Band, band8 is the reflection value of the near infrared Band, and Band2 is the reflection value of the blue Band.
The other samples include data of one or a combination of:
other crops, vegetation, bare land, buildings.
As a preferred example, the training module 301 is further configured to perform feature dimension reduction:
discarding the features of which the importance index is lower than a preset first threshold value; or alternatively
And carrying out characteristic dimension reduction on the NDVI and the EVI according to Principal Component Analysis (PCA).
As a preferred example, the preset algorithm includes:
the preset algorithm is a random forest algorithm with grid search.
As a preferred example, the training module 301 is further configured to, after training the preset algorithm:
modifying the values of parameters in the grid search until the accuracy meets the preset requirement;
the parameters include one or a combination of the following:
the number of trees;
the maximum depth of the tree.
As a preferred example, the training module 301 is further configured to, after training the preset algorithm: and verifying the trained model by using verification data, and calculating the accuracy.
As a preferred example, the post-processing module 303 is further configured to boundary clean-up the identified results: and smoothing boundary saw teeth at different category junctions in the identification result.
As a preferred example, the post-processing module 303 is further configured to perform a mode filtering process on the identified result: and removing the salt and pepper noise in the identification result through the mode filtering processing.
It should be noted that, the training module 301 provided in this embodiment can implement all the functions included in step S101 in the above method, solve the same technical problem, achieve the same technical effect, and are not described herein again;
it should be noted that, the identification module 302 provided in this embodiment can implement all functions included in the step S102 in the above method, solve the same technical problem, achieve the same technical effect, and are not described herein again;
it should be noted that, the post-processing module 303 provided in this embodiment can implement all functions included in step S103 in the above method, solve the same technical problem, achieve the same technical effect, and are not described herein again;
it should be noted that, the device and the method belong to the same inventive concept, solve the same technical problem, achieve the same technical effect, and the device can implement all the above methods, and the same points are not repeated.
Based on the same inventive concept, the embodiment of the invention also provides a wheat identification device, as shown in fig. 4, which comprises:
including a memory 402, a processor 401 and a user interface 403;
the memory 402 is used for storing a computer program;
the user interface 403 is configured to interact with a user;
the processor 401 is configured to read a computer program in the memory 402, where the processor 401 implements:
training a preset algorithm by using training data to obtain a wheat identification model;
identifying a preset area by using the wheat identification model;
and carrying out boundary cleaning and mode filtering treatment on the identified result to obtain a wheat identification result.
Where in FIG. 4, a bus architecture may comprise any number of interconnected buses and bridges, with one or more processors, represented in particular by processor 401, and various circuits of memory, represented by memory 402, linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, power management circuits, etc., which are well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The processor 401 is responsible for managing the bus architecture and general processing, and the memory 402 may store data used by the processor 401 in performing operations.
The processor 401 may be CPU, ASIC, FPGA or CPLD, and the processor 401 may also employ a multi-core architecture.
When the processor 401 executes the computer program stored in the memory 402, any one of the wheat identification methods in the first embodiment is implemented.
It should be noted that the device and the method belong to the same inventive concept, solve the same technical problem, achieve the same technical effect, and the same points are not repeated.
The present application also proposes a processor readable storage medium. The processor-readable storage medium stores a computer program, and the processor implements any one of the wheat identification methods in the first embodiment when executing the computer program.
It should be noted that, in the embodiment of the present application, the division of the units is schematic, which is merely a logic function division, and other division manners may be implemented in actual practice. In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (14)
1. A wheat identification method, comprising:
training a preset algorithm by using training data to obtain a wheat identification model;
identifying a preset area by using the wheat identification model;
and carrying out boundary cleaning and mode filtering treatment on the identified result to obtain a wheat identification result.
2. The method of claim 1, wherein training a predetermined algorithm using training data to obtain a wheat recognition model comprises:
acquiring satellite images of the whole growth period of wheat, removing cloud images, and calculating normalized vegetation index NDVI and enhanced vegetation index EVI of the rest images;
carrying out mask extraction on the grid data of the NDVI and the EVI by using sample plot data acquired in the field to respectively obtain NDVI and EVI time sequence curves of a wheat sample and other samples;
constructing importance indexes of the NDVI and the EVI, and performing characteristic dimension reduction;
sequencing the importance indexes, and selecting vegetation indexes corresponding to the N most important importance indexes as the characteristics of model training;
training a preset algorithm by taking the time sequence curve as the training data;
wherein the importance index is used for measuring the importance of the NDVI and/or the EVI, and the higher the value of the importance index is, the more important the NDVI and/or the EVI is;
n is an integer greater than or equal to 1.
3. The method of claim 2, wherein the culling the cloud image comprises:
and eliminating the influence of the cloud coverage exceeding a preset threshold.
4. The method of claim 2, wherein calculating the normalized vegetation index NDVI and enhanced vegetation index EVI for the remaining images comprises:
the NDVI is calculated by the following formula:
the EVI is calculated by the following formula:
wherein Band4 is the reflection value of the red Band, band8 is the reflection value of the near infrared Band, and Band2 is the reflection value of the blue Band.
5. The method of claim 2, wherein the other samples comprise one or a combination of data:
other crops, vegetation, bare land, buildings.
6. The method of claim 2, wherein performing feature dimension reduction comprises:
discarding the features of which the importance index is lower than a preset first threshold value; or alternatively
And carrying out characteristic dimension reduction on the NDVI and the EVI according to Principal Component Analysis (PCA).
7. The method of claim 2, wherein the predetermined algorithm comprises:
the preset algorithm is a random forest algorithm with grid search.
8. A method according to claim 3, wherein the time sequence curve is used as the training data, and further comprising, after training a predetermined algorithm:
modifying the values of parameters in the grid search until the accuracy meets the preset requirement;
the parameters include one or a combination of the following:
the number of trees;
the maximum depth of the tree.
9. The method according to claim 2, wherein training the preset algorithm using the timing curve as the training data further comprises:
and verifying the trained model by using verification data, and calculating the accuracy.
10. The method of claim 1, wherein boundary cleaning the identified results comprises:
and smoothing boundary saw teeth at different category junctions in the identification result.
11. The method of claim 1, wherein performing a mode filter process on the identified result comprises:
and removing the salt and pepper noise in the identification result through the mode filtering processing.
12. A wheat identification apparatus, comprising:
the training module is configured to train a preset algorithm by using training data to obtain a wheat recognition model;
the identification module is configured to identify a preset area by using the wheat identification model;
and the post-processing module is configured to perform boundary cleaning and mode filtering processing on the identified result to obtain a wheat identification result.
13. A wheat identification apparatus comprising a memory, a processor and a user interface;
the memory is used for storing a computer program;
the user interface is used for realizing interaction with a user;
the processor being adapted to read a computer program in the memory, the processor implementing the wheat identification method according to one of claims 1 to 11 when the computer program is executed.
14. A processor-readable storage medium, characterized in that the processor-readable storage medium stores a computer program, which when executed by the processor implements the wheat identification method according to one of claims 1 to 11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310146482.0A CN116229261A (en) | 2023-02-22 | 2023-02-22 | Wheat identification method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310146482.0A CN116229261A (en) | 2023-02-22 | 2023-02-22 | Wheat identification method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116229261A true CN116229261A (en) | 2023-06-06 |
Family
ID=86576401
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310146482.0A Pending CN116229261A (en) | 2023-02-22 | 2023-02-22 | Wheat identification method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116229261A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850694A (en) * | 2015-05-13 | 2015-08-19 | 福州大学 | Winter wheat remote sensing monitoring method based on vegetation index increment in growing period |
| CN109583345A (en) * | 2018-11-21 | 2019-04-05 | 平安科技（深圳）有限公司 | Roads recognition method, device, computer installation and computer readable storage medium |
| CN109800815A (en) * | 2019-01-24 | 2019-05-24 | 北华航天工业学院 | Training method, wheat recognition methods and training system based on Random Forest model |
| CN113963260A (en) * | 2021-10-20 | 2022-01-21 | 中科三清科技有限公司 | Extraction method and device for winter wheat planting area and computer equipment |
| CN114529097A (en) * | 2022-02-26 | 2022-05-24 | 黑龙江八一农垦大学 | Multi-scale crop phenological period remote sensing dimensionality reduction prediction method |
| CN114694043A (en) * | 2022-03-30 | 2022-07-01 | 中国人民解放军空军军医大学 | Ground wounded person identification method, device and medium of airborne multispectral multi-domain preferred features under complex scene |
-
2023
- 2023-02-22 CN CN202310146482.0A patent/CN116229261A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850694A (en) * | 2015-05-13 | 2015-08-19 | 福州大学 | Winter wheat remote sensing monitoring method based on vegetation index increment in growing period |
| CN109583345A (en) * | 2018-11-21 | 2019-04-05 | 平安科技（深圳）有限公司 | Roads recognition method, device, computer installation and computer readable storage medium |
| CN109800815A (en) * | 2019-01-24 | 2019-05-24 | 北华航天工业学院 | Training method, wheat recognition methods and training system based on Random Forest model |
| CN113963260A (en) * | 2021-10-20 | 2022-01-21 | 中科三清科技有限公司 | Extraction method and device for winter wheat planting area and computer equipment |
| CN114529097A (en) * | 2022-02-26 | 2022-05-24 | 黑龙江八一农垦大学 | Multi-scale crop phenological period remote sensing dimensionality reduction prediction method |
| CN114694043A (en) * | 2022-03-30 | 2022-07-01 | 中国人民解放军空军军医大学 | Ground wounded person identification method, device and medium of airborne multispectral multi-domain preferred features under complex scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | A vision-based robust grape berry counting algorithm for fast calibration-free bunch weight estimation in the field | |
| CN111243736B (en) | Survival risk assessment method and system | |
| Revathy et al. | Comparative analysis of C4. 5 and C5. 0 algorithms on crop pest data | |
| CN112232476A (en) | Method and device for updating test sample set | |
| CN109816509A (en) | Generation method, terminal device and the medium of scorecard model | |
| CN113537063B (en) | Forest environment remote sensing monitoring method and terminal based on comprehensive remote sensing technology | |
| CN111401339A (en) | Method and device for identifying age of person in face image and electronic equipment | |
| CN107368526A (en) | A kind of data processing method and device | |
| CN111242191A (en) | Credit rating method and device based on multi-classifier integration | |
| Herdiyeni et al. | Chilli quality classification using deep learning | |
| Zermas et al. | Extracting phenotypic characteristics of corn crops through the use of reconstructed 3D models | |
| CN111709468A (en) | Training method and device for directional artificial intelligence and storage medium | |
| CN115438040A (en) | Pathological archive information management method and system | |
| Nirale et al. | Analytical Study on IoT and Machine Learning based Grading and Sorting System for Fruits | |
| Lallich et al. | Improving classification by removing or relabeling mislabeled instances | |
| CN116229261A (en) | Wheat identification method, device and storage medium | |
| CN111696122A (en) | Crop phenotype parameter extraction method and device | |
| CN112308289A (en) | Rice yield prediction method and device | |
| CN111598580A (en) | XGboost algorithm-based block chain product detection method, system and device | |
| CN115034693B (en) | Biological information data security management method, system and storage medium based on Internet of things | |
| Karyemsetty et al. | A Machine Learning Approach to Classification of Okra | |
| Alizadeh Khameneh | Tree detection and species identification using LiDAR data | |
| CN114689526A (en) | Banana wilt nondestructive detection method, device and detection equipment | |
| Zhang et al. | Adapting the Segment Anything Model for Plant Recognition and Automated Phenotypic Parameter Measurement | |
| Fernandes et al. | Towards Precise Pruning Points Detection using Semantic-Instance-Aware Plant Models for Grapevine Winter Pruning Automation |
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 |