CN115545334A - Land use type prediction method, land use type prediction device, electronic device, and storage medium - Google Patents
Land use type prediction method, land use type prediction device, electronic device, and storage medium Download PDFInfo
- Publication number
- CN115545334A CN115545334A CN202211313851.2A CN202211313851A CN115545334A CN 115545334 A CN115545334 A CN 115545334A CN 202211313851 A CN202211313851 A CN 202211313851A CN 115545334 A CN115545334 A CN 115545334A
- Authority
- CN
- China
- Prior art keywords
- land
- land utilization
- prediction model
- data
- pixel
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000012549 training Methods 0.000 claims abstract description 110
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000012360 testing method Methods 0.000 claims description 37
- 238000013528 artificial neural network Methods 0.000 claims description 29
- 230000000306 recurrent effect Effects 0.000 claims description 27
- 238000004590 computer program Methods 0.000 claims description 15
- 230000015654 memory Effects 0.000 claims description 13
- 238000012795 verification Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 description 8
- 210000002569 neuron Anatomy 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 7
- 210000004027 cell Anatomy 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 239000002689 soil Substances 0.000 description 3
- 230000001413 cellular effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Abstract
The invention provides a method and a device for predicting land utilization types, electronic equipment and a storage medium, which relate to the technical field of remote sensing image information, and the method comprises the following steps: acquiring land utilization remote sensing image data of a target area at a first time; processing the remote sensing image data of land utilization to obtain land utilization data of at least one pixel corresponding to a target area at a first time; inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on the sample land utilization data and the sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time. The method provided by the invention improves the accuracy of land use type prediction.
Description
Technical Field
The invention relates to the technical field of remote sensing image information, in particular to a method and a device for predicting land utilization types, electronic equipment and a storage medium.
Background
Land resources are basic resources on which human beings live, have space-time characteristics, and the reasonable and scientific use of land and the protection of land are important contents for realizing sustainable development. Changes in land use can affect human living environment and economic development to some extent. With the rapid development of Geographic Information Systems (GIS) and Remote Sensing (RS) technologies, the historical data of land utilization has reached twenty-three decades. Therefore, the prediction of the land use spatial distribution has important significance for future land use planning.
In the related technology, the models commonly used for land use prediction are mainly cellular automata Markov (CA-Markov), a FLUS model, a PLUS model, a CLUE-S model and the like, firstly, a land use conversion rule is estimated through an estimation algorithm, and then land use prediction is carried out through the cellular automata models such as the CA-Markov, the FLUS model, the PLUS model, the CLUE-S model and the like according to influence factors, land use data and the conversion rule, wherein the land use prediction models are different only in the selection of the algorithm estimation land use conversion rule and the influence factors, the commonly used estimation algorithm mainly comprises three types, and one type is a traditional statistical method, such as logistic regression; one is an intelligent optimization algorithm, such as a genetic algorithm, a particle swarm algorithm and the like; one is a deep learning algorithm, such as an artificial neural network.
However, the land use prediction model used in the related art only uses the land use data of the period 2 to the period 3 and the related influence factors for prediction, so that the land use prediction model does not fully explore the information of the land use historical data, and the influence factors of land use change are numerous and have different influence degrees, including natural conditions, economic development, human interference and the like, so that some influence factors cannot be quantized and the weight is difficult to judge, and the selection of the influence factors has subjectivity, thereby causing low accuracy of land use type prediction.
Disclosure of Invention
The invention provides a method and a device for predicting land use types, electronic equipment and a storage medium, which are used for solving the defect of low accuracy in prediction of the land use types in the prior art and realizing more accurate prediction of the land use types.
The invention provides a land utilization type prediction method, which comprises the following steps:
acquiring land utilization remote sensing image data of a target area at a first time;
processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area in the first time;
the land use type data of each pixel at the first time is sent to a target land use prediction model, and the land use type of each pixel at the second time output by the target land use prediction model is obtained; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
According to the land utilization type prediction method provided by the invention, the target land utilization prediction model comprises a long-short term memory (LSTM) recurrent neural network module and a multilayer perceptron, the land utilization data of each pixel at a first time is input into the target land utilization prediction model, and the land utilization type of each pixel at a second time output by the target land utilization prediction model is obtained, and the method comprises the following steps:
inputting the land utilization data of each pixel at the first time into the LSTM recurrent neural network module to obtain land utilization characteristic data corresponding to each pixel output by the LSTM recurrent neural network module;
and inputting the land utilization characteristic data corresponding to each pixel into the multilayer perceptron to obtain the land utilization type of each pixel output by the multilayer perceptron at the second time.
According to the land utilization type prediction method provided by the invention, the target land utilization prediction model is obtained based on the training of the following steps:
acquiring a land utilization data set; the land utilization data set comprises historical land utilization data of at least one pixel corresponding to the target area at different times;
dividing the land use data set into a training set and a testing set; the training set comprises at least one group of training data and label data corresponding to each pixel at different time; the test set comprises at least one group of test data and real data corresponding to each pixel at different time;
and training an initial land utilization prediction model based on the training set and the test set to obtain the target land utilization prediction model.
According to the land utilization type prediction method provided by the invention, the training of the initial land utilization prediction model based on the training set and the test set to obtain the target land utilization prediction model comprises the following steps:
training an initial land utilization prediction model by adopting the training set to obtain a first land utilization prediction model;
verifying the first land use prediction model by using the test set;
judging whether a training stopping condition is met; the training stopping condition comprises that the verification precision is not less than a preset threshold value or the iteration times reach preset times;
and determining the target land utilization prediction model based on the judgment result.
According to the soil utilization type prediction method provided by the invention, the target soil utilization prediction model is determined based on the judgment result, and the method comprises the following steps:
under the condition that the training stopping condition is not met, repeatedly executing the step of training the initial land use prediction model by adopting the training set;
and in the case that a training stopping condition is met, taking the first land use prediction model as a target land use prediction model.
According to the land utilization type prediction method provided by the invention, the land utilization type comprises at least one of the following items: cultivated land, woodland, grassland, water area, construction land and unused land.
The present invention also provides a soil utilization type prediction device, including:
the acquisition module is used for acquiring land utilization remote sensing image data of a target area at a first time;
the processing module is used for processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area at the first time;
the prediction module is used for inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
The present invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of land use type prediction as described in any one of the above when the program is executed.
The present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a land use type prediction method as described in any one of the above.
The present invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a land use type prediction method as described in any one of the above.
According to the land utilization type prediction method, the land utilization type prediction device, the electronic equipment and the storage medium, land utilization remote sensing image data of a target area at a first time are obtained, and then the land utilization remote sensing image data are processed to obtain land utilization data of at least one pixel corresponding to the target area at the first time; inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time, accurate prediction of the land utilization type of a plurality of pixels corresponding to the target area at the second time is achieved through the target land utilization prediction model, the accuracy of land utilization type prediction is improved, effective planning of land utilization in the future is facilitated, and the utilization rate of land resources is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a land use type prediction method provided by the present invention;
FIG. 2 is a second schematic flow chart of the land use type prediction method provided by the present invention;
fig. 3 is a schematic structural view of a land use type prediction apparatus provided by the present invention;
fig. 4 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, 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 land use type prediction method of the present invention is described below with reference to fig. 1 to 2.
Fig. 1 is a schematic flow chart of a land use type prediction method provided by the present invention, as shown in fig. 1, the method includes: step 101-step 103; wherein the content of the first and second substances,
It should be noted that the land use type prediction method provided by the invention can be applied to a land resource planning scene. The execution subject of the method may be a land use type prediction apparatus, such as an electronic device, or a control module in the land use type prediction apparatus for executing the land use type prediction method.
Specifically, the target area may be an area of the type of land use to be predicted, and the first time may be any one year prior to the predicted time. The remote sensing image of the area with the land use type to be predicted at the first time is shot through the satellite, so that the land use remote sensing image data of the target area can be obtained.
And 102, processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area in the first time.
Specifically, the target area can be divided into a plurality of pixels, and the land utilization data can be of a land utilization type or other land utilization information.
Optionally, the land use type comprises at least one of: cultivated land, woodland, grassland, water area, construction land and unused land.
In practice, land use data of each pixel at the first time can be obtained by performing One-Hot Encoding (One-Hot Encoding) on the land use remote sensing image data. For example, when the land use type of each pixel at the first time is arable land, then the land use data for that pixel at the first time may be represented as [1,0,0,0,0,0].
103, inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
In particular, the second time may be a predicted time, i.e. any year after the first time. Inputting the land utilization data of each pixel at the first time into the target land utilization prediction model, and obtaining the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
The land use type prediction method provided by the invention comprises the steps of obtaining land use remote sensing image data of a target area at a first time, and processing the land use remote sensing image data to obtain land use data of at least one pixel corresponding to the target area at the first time; inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time, accurate prediction of the land utilization type of a plurality of pixels corresponding to the target area at the second time is achieved through the target land utilization prediction model, the accuracy of land utilization type prediction is improved, effective planning of land utilization in the future is facilitated, and the utilization rate of land resources is improved.
Optionally, the target land use prediction model includes a long-term short-term memory LSTM recurrent neural network module and a multi-layer perceptron, and the specific implementation manner of the step 103 includes:
1) And inputting the land utilization data of each pixel at the first time into the LSTM recurrent neural network module to obtain the land utilization characteristic data corresponding to each pixel output by the LSTM recurrent neural network module.
Specifically, the target land utilization prediction model is an MLP _ LSTM model and comprises a Long Short Term Memory recursion (LSTM) neural network module and a Multi-layer Perceptron (MLP), wherein the LSTM recurrent neural network module is used for extracting land utilization characteristic information of each pixel among land utilization data of a first time, and the land utilization characteristic information represents a weight of which land utilization type the land utilization type of each pixel at a second time in the future is possible. Inputting the land utilization data of each pixel at the first time into an LSTM recurrent neural network module, inputting the output of each LSTM neuron of the hidden layer of the previous layer of the LSTM network into the LSTM neuron corresponding to the hidden layer of the next layer of the LSTM recurrent neural network module for calculation, obtaining the final output sequence of the LSTM network after all calculations, and obtaining the land utilization characteristic data corresponding to each pixel output by the LSTM recurrent neural network module; the LSTM neuron structure comprises a cell state, a forgetting gate, an input gate and an output gate, and the cell state is updated through the forgetting gate, the input gate and the output gate.
2) And inputting the land utilization characteristic data corresponding to each pixel into the multilayer perceptron to obtain the land utilization type of each pixel output by the multilayer perceptron at the second time.
Specifically, the multilayer perceptron is used for determining the land use type of each image element at the second time based on the land use characteristic data corresponding to each image element output by the LSTM recurrent neural network module. And inputting the land utilization characteristic data corresponding to each pixel into the multilayer perceptron, and performing convolution calculation on the land utilization characteristic data by the multilayer perceptron to obtain the land utilization type of each pixel output by the multilayer perceptron at the second time, so that the land utilization type of each pixel corresponding to the target area at the second time is predicted.
The land use type prediction method provided by the invention comprises the steps of inputting land use data of each pixel at the first time into an LSTM recurrent neural network module to obtain land use characteristic data corresponding to each pixel output by the LSTM recurrent neural network module; and inputting the land utilization characteristic data corresponding to each pixel into the multi-layer perceptron to obtain the land utilization type of each pixel output by the multi-layer perceptron at the second time, and realizing that the LSTM recurrent neural network module extracts the land utilization characteristic information of each pixel among the land utilization data of each pixel at the first time through the combination of the LSTM recurrent neural network module and the multi-layer perceptron.
Fig. 2 is a second schematic flowchart of the land use type prediction method provided by the present invention, as shown in fig. 2, the method includes: step 201-step 204; wherein, the first and the second end of the pipe are connected with each other,
And 204, inputting the land use characteristic data corresponding to each pixel into a multilayer sensor, wherein the multilayer sensor comprises an input layer, a convolution layer and an output layer, and the land use type of each pixel output by the multilayer sensor at the second time can be obtained.
Next, a training process of the target land use prediction model will be described in detail.
Optionally, the target land use prediction model is trained based on the following steps:
step a) acquiring a land use data set; the land use data set comprises historical land use data of at least one pixel corresponding to the target area at different times.
Specifically, historical land utilization data of the target area at different times are obtained, and classification and arrangement are performed according to different times, so that a land utilization data set is obtained.
The land use data set is represented by equation (1), wherein,
E＝{(X 1992 ,Y 1997 ),(X 1993 ,Y 1998 ),...,(X year ,Y year+m )} (1)
wherein, X year Historical land use data, Y, from 1992 to 2015, respectively year+m Respectively 1997 to 2020, and m is an integer greater than 1.
It should be noted that the One-Hot coding can be used for coding historical land use data of each pixel in the land use data set at different times.
Step b) dividing the land use data set into a training set and a test set; the training set comprises at least one group of training data and label data corresponding to each pixel at different time; the test set comprises at least one group of test data and real data corresponding to each pixel at different time.
Specifically, the land use data set is divided into a training set and a test set, and the training set and the test set can be obtained according to the land use data set represented by the formula (1), wherein the training set is represented by the formula (2), and X is i As training data, Y i+m As the tag data, among others,
Tr＝{(X 1992 ,Y 1997 ),(X 1993 ,Y 1998 ),...,(X i ,Y i+m )} (2)
wherein X i And Y i+m Expressed by equation (3):
wherein X i Representing the historical land use data for i years,
representing the type of land use represented by the nth pixel in the historical land use data of the ith year. For example, if i ranges from 1992 to 2014, and m is 5, the test set is represented by formula (4):
Te＝{(X 2015 ,Y 2020 )} (4)
it should be noted that, in the training process of the training set, the test set can be used to predict the land use type of the target area in the 2020 land use spatial distribution and compare the land use type with the actual 2020 land use type, so that the accuracy of the MLP _ LSTM model can be verified.
And c) training an initial land use prediction model based on the training set and the testing set to obtain the target land use prediction model.
Specifically, the constructed initial land use prediction model may be trained according to a training set and a test set, so that a target land use prediction model may be obtained.
The land use type prediction method provided by the invention comprises the steps of acquiring a land use data set; the land utilization data set comprises historical land utilization data of at least one pixel corresponding to the target area at different time; dividing a land utilization data set into a training set and a testing set; the training set comprises at least one group of training data and label data corresponding to each pixel at different time; the test set comprises at least one group of test data and real data corresponding to each pixel at different time; the initial land use prediction model is trained according to the training set and the testing set, a target land use prediction model can be obtained, accurate prediction of land use types of a plurality of pixels corresponding to a target area at a second time can be achieved through the target land use prediction model, accuracy of land use type prediction is improved, effective planning of land use in the future is facilitated, and utilization rate of land resources is improved.
Optionally, the training an initial land use prediction model based on the training set and the test set to obtain the target land use prediction model includes:
1) And training the initial land utilization prediction model by adopting the training set to obtain a first land utilization prediction model.
Specifically, at least one group of training data and label data corresponding to each pixel in different time included in a training set are simultaneously input to an LSTM recurrent neural network module in an initial land use prediction model, the LSTM recurrent neural network module is trained layer by layer, the output of each LSTM neuron of a hidden layer of an upper layer of the LSTM network is input to the LSTM neuron corresponding to a hidden layer of a lower layer of the LSTM network for calculation, a final output sequence of the LSTM network is obtained after all calculations, land use characteristic data corresponding to each pixel output by the LSTM recurrent neural network module is obtained, then the land use characteristic data corresponding to each pixel is input to a multi-layer sensor, the multi-layer sensor performs convolution calculation on the land use characteristic data, the land use type of each pixel output by the multi-layer sensor at the second time is obtained, iterative training of the initial land use prediction model is realized, and a first land use prediction model is obtained.
It should be noted that in the training process of the model, at least one set of training data and label data corresponding to each pixel at different time are simultaneously input into the initial land use prediction model, and the spatial position relationship of the historical land use data is reserved.
2) And verifying the first land use prediction model by adopting the test set.
Specifically, after the first land use prediction model is obtained, at least one set of test data and real data corresponding to each pixel included in the test set at different time is input into the first land use prediction model, for example, the test data in 2015 and the real data in 2020, the land use type in 2020 of each pixel output by the first land use prediction model can be obtained, and the predicted land use type in 2020 of each pixel is compared with the real data in 2020, so that the verification accuracy of the first land use prediction model can be obtained.
3) Judging whether a training stopping condition is met; the training stopping condition comprises that the verification precision is not less than a preset threshold value or the iteration times reach preset times.
Specifically, in the process of model training, it may be determined whether a training stop condition is satisfied, where the training stop condition includes that the verification accuracy is not less than a preset threshold or the number of iterations reaches a preset number.
4) And determining the target land utilization prediction model based on the judgment result.
Specifically, according to the result of the judgment, a target land use prediction model may be determined.
Optionally, the determining the target land use prediction model based on the result of the judgment includes:
under the condition that the training stopping condition is not met, repeatedly executing the step of training the initial land use prediction model by adopting the training set; and in the case that the training stopping condition is met, taking the first land use prediction model as a target land use prediction model.
Specifically, in the case where it is determined that the training stop condition is not satisfied, the above-described step of training the initial land use prediction model using the training set is repeatedly performed until the training stop condition is satisfied. And under the condition that the training stopping condition is met, the first land use prediction model can be used as a target land use prediction model, so that the target land use prediction model is obtained.
It should be noted that in the process of model training, information in historical land utilization data of land utilization at different times can be fully excavated, and future land utilization types can be objectively, scientifically and accurately predicted, so that scientific basis is provided for implementing adjustment and optimization of raw land utilization.
The land use type prediction method provided by the invention comprises the steps of training an initial land use prediction model by adopting a training set to obtain a first land use prediction model; in the model training process, verifying the first land utilization prediction model by adopting a test set; judging whether a training stopping condition is met; the training stopping condition comprises that the verification precision is not less than a preset threshold value or the iteration times reach preset times; under the condition that the training stopping condition is not met, repeatedly executing the step of training the initial land utilization prediction model by adopting the training set; under the condition that the training stopping condition is met, the obtained first land utilization prediction model can be used as a target land utilization prediction model, accurate prediction of land utilization types of a plurality of pixels corresponding to a target area at a second time can be achieved through the target land utilization prediction model, accuracy of land utilization type prediction is improved, effective planning of land utilization in the future is facilitated, and utilization rate of land resources is improved.
The land use type prediction apparatus provided by the present invention will be described below, and the land use type prediction apparatus described below and the land use type prediction method described above may be referred to in correspondence with each other.
Fig. 3 is a schematic structural view of a land use type prediction apparatus provided by the present invention, and as shown in fig. 3, the land use type prediction apparatus 300 includes: an acquisition module 301, a processing module 302, and a prediction module 303, wherein,
the acquisition module 301 is configured to acquire land use remote sensing image data of a target area at a first time;
a processing module 302, configured to process the remote sensing image data of land use to obtain land use data of at least one pixel corresponding to the target area at the first time;
the prediction module 303 is configured to input the land utilization data of each pixel at the first time into a target land utilization prediction model, so as to obtain a land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land use prediction model is obtained by training based on sample land use data and sample label data and is used for predicting the land use type of each pixel corresponding to the target area at the second time.
The land use type prediction device obtains land use data of at least one pixel corresponding to a target area at a first time by obtaining land use remote sensing image data of the target area at the first time and then processing the land use remote sensing image data; inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time, accurate prediction of the land utilization type of a plurality of pixels corresponding to the target area at the second time is achieved through the target land utilization prediction model, the accuracy of land utilization type prediction is improved, effective planning of land utilization in the future is facilitated, and the utilization rate of land resources is improved.
Optionally, the target land use prediction model includes a long-short term memory LSTM recurrent neural network module and a multi-layer perceptron, and the prediction module 303 is specifically configured to:
inputting the land utilization data of each pixel at the first time into the LSTM recurrent neural network module to obtain land utilization characteristic data corresponding to each pixel output by the LSTM recurrent neural network module;
and inputting the land utilization characteristic data corresponding to each pixel into the multilayer perceptron to obtain the land utilization type of each pixel output by the multilayer perceptron at the second time.
Optionally, the target land use prediction model is trained based on the following steps:
acquiring a land utilization data set; the land use data set comprises historical land use data of at least one pixel corresponding to the target area at different times;
dividing the land use data set into a training set and a testing set; the training set comprises at least one group of training data and label data corresponding to each pixel at different time; the test set comprises at least one group of test data and real data corresponding to each pixel at different time;
and training an initial land utilization prediction model based on the training set and the test set to obtain the target land utilization prediction model.
Optionally, the training an initial land use prediction model based on the training set and the test set to obtain the target land use prediction model includes:
training an initial land utilization prediction model by adopting the training set to obtain a first land utilization prediction model;
verifying the first land use prediction model by using the test set;
judging whether a training stopping condition is met; the training stopping condition comprises that the verification precision is not less than a preset threshold value or the iteration times reach preset times;
and determining the target land utilization prediction model based on the judgment result.
Optionally, the determining the target land use prediction model based on the result of the judgment includes:
under the condition that the training stopping condition is not met, repeatedly executing the step of training the initial land use prediction model by adopting the training set;
and in the case that the training stopping condition is met, taking the first land use prediction model as a target land use prediction model.
Optionally, the land use type comprises at least one of: cultivated land, woodland, grassland, water area, construction land and unused land.
Fig. 4 is a schematic physical structure diagram of an electronic device provided in the present invention, and as shown in fig. 4, the electronic device may include: a processor (processor) 410, a communication Interface 420, a memory (memory) 430 and a communication bus 440, wherein the processor 410, the communication Interface 420 and the memory 430 are communicated with each other via the communication bus 440. The processor 410 may invoke logic instructions in the memory 430 to perform a land use type prediction method comprising:
acquiring land utilization remote sensing image data of a target area at a first time;
processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area in the first time;
inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
In addition, the logic instructions in the memory 430 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, and various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program, the computer program being storable on a non-transitory computer readable storage medium, the computer program, when executed by a processor, being capable of executing a land use type prediction method provided by the above methods, the method comprising:
acquiring land utilization remote sensing image data of a target area at a first time;
processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area at the first time;
inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform a land use type prediction method provided by the above methods, the method comprising:
acquiring land utilization remote sensing image data of a target area at a first time;
processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area at the first time;
inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for predicting a land use type, comprising:
acquiring land utilization remote sensing image data of a target area at a first time;
processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area in the first time;
inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
2. The land use type prediction method of claim 1, wherein the target land use prediction model comprises a long-short term memory (LSTM) recurrent neural network module and a multi-layer perceptron, and the inputting the land use data of each pixel at a first time into the target land use prediction model to obtain the land use type of each pixel at a second time output by the target land use prediction model comprises:
inputting the land utilization data of each pixel at the first time into the LSTM recurrent neural network module to obtain land utilization characteristic data corresponding to each pixel output by the LSTM recurrent neural network module;
and inputting the land utilization characteristic data corresponding to each pixel into the multilayer perceptron to obtain the land utilization type of each pixel output by the multilayer perceptron at the second time.
3. The land use type prediction method according to claim 1 or 2, characterized in that the target land use prediction model is trained based on the following steps:
acquiring a land utilization data set; the land use data set comprises historical land use data of at least one pixel corresponding to the target area at different times;
dividing the land use data set into a training set and a testing set; the training set comprises at least one group of training data and label data corresponding to each pixel at different time; the test set comprises at least one group of test data and real data corresponding to each pixel at different time;
and training an initial land utilization prediction model based on the training set and the test set to obtain the target land utilization prediction model.
4. The land use type prediction method of claim 3, wherein the training an initial land use prediction model based on the training set and the test set to obtain the target land use prediction model comprises:
training an initial land utilization prediction model by adopting the training set to obtain a first land utilization prediction model;
verifying the first land use prediction model by using the test set;
judging whether a training stopping condition is met; the training stopping condition comprises that the verification precision is not less than a preset threshold or the iteration times reach preset times;
and determining the target land utilization prediction model based on the judgment result.
5. The land use type prediction method according to claim 4, wherein the determining the target land use prediction model based on a result of the judgment includes:
under the condition that the training stopping condition is not met, repeatedly executing the step of training the initial land use prediction model by adopting the training set;
and in the case that the training stopping condition is met, taking the first land use prediction model as a target land use prediction model.
6. The land use type prediction method according to any one of claims 1, wherein the land use type includes at least one of: cultivated land, woodland, grassland, water area, construction land and unused land.
7. An apparatus for predicting a land use type, comprising:
the acquisition module is used for acquiring land utilization remote sensing image data of a target area at a first time;
the processing module is used for processing the land utilization remote sensing image data to obtain land utilization data of at least one pixel corresponding to the target area at the first time;
the prediction module is used for inputting the land utilization data of each pixel at the first time into a target land utilization prediction model to obtain the land utilization type of each pixel at the second time, which is output by the target land utilization prediction model; the target land utilization prediction model is obtained by training based on sample land utilization data and sample label data and is used for predicting the land utilization type of each pixel corresponding to the target area at the second time.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the land use type prediction method according to any one of claims 1 to 6.
9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the land use type prediction method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the land use type prediction method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211313851.2A CN115545334B (en) | 2022-10-25 | 2022-10-25 | Land utilization type prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211313851.2A CN115545334B (en) | 2022-10-25 | 2022-10-25 | Land utilization type prediction method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115545334A true CN115545334A (en) | 2022-12-30 |
| CN115545334B CN115545334B (en) | 2024-04-02 |
Family
ID=84719002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211313851.2A Active CN115545334B (en) | 2022-10-25 | 2022-10-25 | Land utilization type prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115545334B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116246175A (en) * | 2023-05-05 | 2023-06-09 | 西昌学院 | Land utilization information generation method, electronic device, and computer-readable medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108268527A (en) * | 2016-12-30 | 2018-07-10 | 航天星图科技（北京）有限公司 | A kind of method for detecting land use pattern variation |
| US20190303703A1 (en) * | 2018-03-30 | 2019-10-03 | Regents Of The University Of Minnesota | Predicting land covers from satellite images using temporal and spatial contexts |
| CN111738329A (en) * | 2020-06-19 | 2020-10-02 | 中南大学 | Land use classification method for time series remote sensing images |
| CN111931738A (en) * | 2020-09-28 | 2020-11-13 | 南京邮电大学 | Neural network model pre-training method and device for remote sensing image |
| KR20200136535A (en) * | 2019-05-27 | 2020-12-08 | 공주대학교 산학협력단 | Land-cover Change prediction Method Using Regional Land-use Change Expected Values |
| CN113807278A (en) * | 2021-09-23 | 2021-12-17 | 华中师范大学 | Deep learning-based land use classification and change prediction method |
| CN114821340A (en) * | 2022-06-01 | 2022-07-29 | 河南大学 | Land utilization classification method and system |
| CN114998361A (en) * | 2022-06-07 | 2022-09-02 | 山西云时代智慧城市技术发展有限公司 | Agricultural land cover spatio-temporal semantic segmentation method based on transformations-MulMLA |
-
2022
- 2022-10-25 CN CN202211313851.2A patent/CN115545334B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108268527A (en) * | 2016-12-30 | 2018-07-10 | 航天星图科技（北京）有限公司 | A kind of method for detecting land use pattern variation |
| US20190303703A1 (en) * | 2018-03-30 | 2019-10-03 | Regents Of The University Of Minnesota | Predicting land covers from satellite images using temporal and spatial contexts |
| KR20200136535A (en) * | 2019-05-27 | 2020-12-08 | 공주대학교 산학협력단 | Land-cover Change prediction Method Using Regional Land-use Change Expected Values |
| CN111738329A (en) * | 2020-06-19 | 2020-10-02 | 中南大学 | Land use classification method for time series remote sensing images |
| CN111931738A (en) * | 2020-09-28 | 2020-11-13 | 南京邮电大学 | Neural network model pre-training method and device for remote sensing image |
| CN113807278A (en) * | 2021-09-23 | 2021-12-17 | 华中师范大学 | Deep learning-based land use classification and change prediction method |
| CN114821340A (en) * | 2022-06-01 | 2022-07-29 | 河南大学 | Land utilization classification method and system |
| CN114998361A (en) * | 2022-06-07 | 2022-09-02 | 山西云时代智慧城市技术发展有限公司 | Agricultural land cover spatio-temporal semantic segmentation method based on transformations-MulMLA |
Non-Patent Citations (3)
| Title |
|---|
| JUNYE WANG ETAL: "Machine learning in modelling land-use and land cover-change (LULCC): Current status, challenges and prospects", 《SCIENCE OF THE TOTAL ENVIRONMENT》, pages 1 - 17 * |
| MUHAMMAD TALHA ZESHAN ETAL: "Monitoring Land Use Changes and Their Future Prospects Using GIS and ANN-CA for Perak River Basin, Malaysia", 《WATER》, pages 1 - 20 * |
| 陈敏等: "结合 VGGNet 与 Mask R-CNN 的高分辨率 遥感影像建设用地检测", 《遥感技术与应用》, vol. 36, no. 2, pages 256 - 264 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116246175A (en) * | 2023-05-05 | 2023-06-09 | 西昌学院 | Land utilization information generation method, electronic device, and computer-readable medium |
| CN116246175B (en) * | 2023-05-05 | 2023-07-14 | 西昌学院 | Land utilization information generation method, electronic device, and computer-readable medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115545334B (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11714937B2 (en) | Estimating physical parameters of a physical system based on a spatial-temporal emulator | |
| CN111310968A (en) | LSTM neural network circulation hydrological forecasting method based on mutual information | |
| CN111639787A (en) | Spatio-temporal data prediction method based on graph convolution network | |
| Jiang et al. | Prediction of house price based on the back propagation neural network in the keras deep learning framework | |
| Braakmann-Folgmann et al. | Sea level anomaly prediction using recurrent neural networks | |
| US11720727B2 (en) | Method and system for increasing the resolution of physical gridded data | |
| Johnson et al. | agTrend: A Bayesian approach for estimating trends of aggregated abundance | |
| Bai et al. | A forecasting method of forest pests based on the rough set and PSO-BP neural network | |
| CN112560966A (en) | Polarimetric SAR image classification method, medium and equipment based on scattergram convolution network | |
| McCulloch et al. | Calibrating agent-based models using uncertainty quantification methods | |
| CN115545334A (en) | Land use type prediction method, land use type prediction device, electronic device, and storage medium | |
| Maulik et al. | Efficient high-dimensional variational data assimilation with machine-learned reduced-order models | |
| CN112949944B (en) | Intelligent groundwater level prediction method and system based on space-time characteristics | |
| CN116628442B (en) | Groundwater reserve change space downscaling method based on artificial neural network | |
| CN110852415B (en) | Vegetation index prediction method, system and equipment based on neural network algorithm | |
| Tripathy et al. | Deep learning in hydrology and water resources disciplines: Concepts, methods, applications, and research directions | |
| Ju et al. | Hydrologic simulations with artificial neural networks | |
| Shrestha et al. | Encapsulation of parametric uncertainty statistics by various predictive machine learning models: MLUE method | |
| Stepchenko | NDVI index forecasting using a layer recurrent neural network coupled with stepwise regression and the PCA | |
| Vidyarthi et al. | Does ANN really acquire the physics of the system? A study using conceptual components from an established water balance model | |
| Roy et al. | Tropical cyclone intensity forecasting in the Bay of Bengal using a biologically inspired computational model | |
| CN113011657A (en) | Method and device for predicting typhoon water level, electronic equipment and storage medium | |
| Pearson et al. | Predicting ecological outcomes using fuzzy interaction webs | |
| Lippert et al. | Physics-informed inference of aerial animal movements from weather radar data | |
| Lønnum | Deep Learning Metocean Simulation and its Applications in Marine Simulation-based Design |
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 | ||
| GR01 | Patent grant |