CN115601651A - Intelligent identification method and device for rural land remediation potential - Google Patents
Intelligent identification method and device for rural land remediation potential Download PDFInfo
- Publication number
- CN115601651A CN115601651A CN202211074997.6A CN202211074997A CN115601651A CN 115601651 A CN115601651 A CN 115601651A CN 202211074997 A CN202211074997 A CN 202211074997A CN 115601651 A CN115601651 A CN 115601651A
- Authority
- CN
- China
- Prior art keywords
- land
- rural
- potential
- rural land
- remote sensing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble 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
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- 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
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
Abstract
The application discloses an intelligent identification method and device for rural land improvement potential, a rural land object to be improved is obtained through vector data of a spatial range of the rural land to be improved and high-resolution remote sensing image data, image features of the rural land object to be improved are extracted, a rural land improvement potential identification model is established, automatic identification of the rural land improvement potential is achieved, and identified potential plots are corrected through participation type investigation. The method and the device make full use of the advantage of rich and diversified ground feature information of the high-resolution remote sensing image, and realize intelligent identification of the rural land remediation potential land parcel based on a mode of combining random forest classification identification and participatory investigation, so that the identified remediation potential land parcel is more accurate in scale, more accurate in spatial position and higher in utilizability.
Description
Technical Field
The application relates to the technical field of rural land reclamation, in particular to an intelligent identification method and device for rural land reclamation potential.
Background
At present, the dual contradiction of the imbalance of urban and rural land configuration and the low efficiency of rural land utilization exists in China, and the transformation of construction land supply from incremental land to stock land is a key exit of the benign development of urban and rural areas in the key period of urban and rural development transformation. As a powerful means for overall urban and rural development and promotion of social new rural construction, rural land improvement can optimize urban and rural construction land structure through spatial replacement and relieve urban and rural human-ground relationship, and becomes important content of national soil comprehensive improvement and village planning in a new period. The rural land reclamation potential mainly refers to the potential of increasing the available land area through utilizing the land reclamation plate in the rural land. The method is a precondition and basis for developing rural land improvement work, and is of great importance for guiding rural areas to scientifically develop land improvement, inventory land utilization and promotion of land intensive saving and utilization.
The rural land reclamation potential identification technology mainly relates to a standard land method for per capita construction, a standard land method for per household construction, an idle land sampling investigation method, a building volume ratio method, a town system planning method and the like. The standard method for per capita construction land, the standard method for per household construction land and the sampling survey method for idle land are several basic common technical methods for identifying rural land remediation potential.
And identifying the rural construction land reclamation potential according to the current situation scale of the rural construction land and the difference value of the product of the determined rural construction land standard and the rural population at the end of the planning period by the everyone construction land standard method. Wherein, the population prediction determines the natural growth rate and the mechanical variation of population on the basis of analyzing the change of the population of the rural areas in the historical years so as to calculate the population of the rural areas at the end of a planning period; the standard of the per capita construction land is determined according to the indexes of the per capita construction land specified in the planning standard of villages and small towns (GB 50188-2007) and the related planning standard of the local area.
The measuring and calculating idea of the standard method for the residential quarter construction land is similar to that of the standard method for the civil quarter construction land, and the method measures and calculates the remediation potential according to the standard of the residential quarter rural construction land specified by the country or the local area.
The method for sampling and surveying the idle land comprises the steps of selecting a typical village which can represent idle conditions of rural construction land used in the whole evaluation area as a sampling point, surveying the idle land area and the land idle rate in the sampling point, and taking the idle land area and the land idle rate as the idle rate of the rural construction land used in the evaluation area so as to calculate the remediation potential of the rural construction land used in the whole area.
The core of rural land improvement potential identification is to determine the land storage patches which can be reused in an improvement mode in the object to be improved. The land for stock is in an idle and unused state in the current situation, comprises various types such as barren grassland, bare land, open land and the like, has complex and fine plaque shapes and land feature characteristics, has strong subjective dependence on potential identification and realization, and needs to be comprehensively judged by means of high-precision images rich in multi-dimensional information, on-site investigation and the like.
However, the existing rural land reclamation potential identification technology based on the per-capita or per-household construction land standard is mainly suitable for large-scale areas, the diversified utilization state of the rural land under the small scale cannot be considered, the potential identification result is large due to the fact that the current per-capita or per-household construction land area is too large, the accuracy is poor, and the practical availability of the identified remediation potential is not high. Although more idle land areas identified by idle land sampling survey can be converted into practical potentials, the accuracy of the identification result depends heavily on the number of regional samples, the rationality of sampling point selection directly relates to whether the idle rate can truly reflect the idle conditions of rural areas in the region and the size of the remediation potential, and the difficulty level of the remediation potential implementation is difficult to distinguish.
Generally speaking, the comprehensive improvement of the state soil under the background of increment conversion stock needs to improve the land utilization of rural land stock and the accuracy of reutilization, the complexity of rural land utilization characteristics is neglected by the commonly used rural land improvement potential identification technology in the existing land improvement, and the identified problems of inaccurate land block scale of the improvement potential, uncertain spatial position and the like are difficult to adapt to the requirement of fine management and utilization of rural land under the new situation.
Disclosure of Invention
Therefore, the method and the device for intelligently identifying the rural land improvement potential are provided to solve the problems that the size of the identified land improvement potential land is not accurate and the spatial position is uncertain in the prior art.
In order to achieve the above purpose, the present application provides the following technical solutions:
in a first aspect, an intelligent identification method for rural land reclamation potential comprises the following steps:
acquiring space range vector data and high-resolution remote sensing image data of rural land to be renovated;
preprocessing the high-resolution remote sensing image data;
carrying out constraint segmentation on the preprocessed high-resolution remote sensing image data by utilizing the rural land space range vector data to be rectified to obtain an optimal segmentation scale, and generating a rural land object to be rectified;
extracting the land cover characteristics of the rural land object to be remediated;
inputting the land cover characteristics into a trained random forest classification model to perform classification and identification on the rural land to be renovated;
the identified idle type is used as a rural land reclamation potential land parcel;
and correcting the identification result of the remediation potential land parcel which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land parcel.
Further, when the random forest classification model is trained to establish a sample data set of a land object of a rural area to be remediated, the method specifically comprises the following steps:
selecting a sample based on the high-resolution remote sensing image data by taking the rural land object to be renovated as a father class;
marking the type of the sample according to the land cover characteristics of the rural land object to be renovated;
classifying and assigning different types of samples to be used as sample label files;
superposing the sample label file and the high-resolution remote sensing image data;
and reading the sample class attribute value and the sample space coordinate information in the sample label file, correspondingly capturing sample spectral characteristics under the same space coordinate position pixel with the high-resolution remote sensing image data as sample characteristic attributes, and establishing a rural land object sample data set to be renovated.
Furthermore, the resolution of the high-resolution remote sensing image data is within 1 meter.
Further, preprocessing the rural land space range vector data to be treated and the high-resolution remote sensing image data, specifically comprising: and carrying out geometric correction on the high-resolution remote sensing image data, and carrying out projection conversion by taking the vector data projection space of the rural land space range to be renovated as a reference.
Furthermore, the geometric correction precision is not lower than the precision of the vector data of the spatial range of the rural land to be treated and is not more than 2 pixels.
Further, the ground cover characteristics comprise spectral bands and spectral indexes.
Further, the spectrum bands are red, green, blue and near infrared, and the spectrum indexes are normalized vegetation indexes, normalized water body indexes and specific value residential area indexes;
wherein, the first and the second end of the pipe are connected with each other,
NDVI = ((NIR-R)/(NIR + R)), NDVI is normalized vegetation index, NIR is pixel value of near infrared band, and R is pixel value of red light band;
NDWI = (G-NIR)/(G + NIR), NDWI is normalized water body index, NIR is pixel value of near infrared band, G is pixel value of green light band;
RRI = B/NIR, RRI being the fractional population index, NIR being the pixel value of the near infrared band; and B is the pixel value of the blue light band.
Further, the idle types include grass, bare land, hardened open land, and miscellaneous objects stockpiled land.
Further, the modifying the land parcel identification result which does not conform to the utilization status comprises: spot shape modification, spot attribute modification and topology inspection.
In a second aspect, an intelligent identification device for rural land reclamation potential comprises:
the data acquisition module is used for acquiring the vector data of the rural land space range to be renovated and the high-resolution remote sensing image data;
the preprocessing module is used for preprocessing the high-resolution remote sensing image data;
the to-be-remediated rural land object generating module is used for carrying out constraint segmentation on the preprocessed high-resolution remote sensing image data by utilizing the to-be-remediated rural land space range vector data to obtain an optimal segmentation scale and generating a to-be-remediated rural land object;
the to-be-remediated rural land object feature extraction module is used for extracting the land cover features of the to-be-remediated rural land object;
the classification recognition module is used for inputting the land cover characteristics into a trained random forest classification model to perform classification recognition on the rural land to be renovated;
the rural land reclamation potential land parcel identification module is used for taking the identified idle type as a rural land reclamation potential land parcel;
and the potential correction module is used for correcting the identification result of the remediation potential land block which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land block.
Compared with the prior art, the method has the following beneficial effects:
the application provides an intelligent identification method and device for rural land improvement potential, a rural land object to be improved is obtained through vector data of a spatial range of the rural land to be improved and high-resolution remote sensing image data, image features of the rural land object to be improved are extracted, a rural land improvement potential identification model is established, automatic identification of the rural land improvement potential is achieved, and identified potential plots are corrected through participation type investigation. The method and the device make full use of the advantages of rich and diversified ground feature information of the high-resolution remote sensing image, and realize intelligent identification of the rural land remediation potential land parcel based on a mode of combining random forest classification identification and participatory investigation, so that the identified remediation potential land parcel is more accurate in scale and more definite in spatial position.
Drawings
To more intuitively illustrate the prior art and the present application, several exemplary drawings are given below. It should be understood that the specific shapes, configurations, shown in the drawings, are not generally considered limitations on the practice of the present application; for example, it is within the ability of those skilled in the art to make routine adjustments or further optimizations based on the technical concepts disclosed in the present application and the exemplary drawings, for the increase/decrease/attribution of certain units (components), specific shapes, positional relationships, connection manners, dimensional ratios, and the like.
Fig. 1 is a flowchart of an intelligent identification method for rural land reclamation potential provided in an embodiment of the present application;
fig. 2 is a functional flow chart of an intelligent rural land reclamation potential identification method provided in an embodiment of the present application;
fig. 3 is a flowchart of random forest classification according to an embodiment of the present application.
Detailed Description
The present application will be described in further detail below with reference to specific embodiments in conjunction with the accompanying drawings.
In the description of the present application: "plurality" means two or more unless otherwise specified. The terms "first", "second", "third", and the like in this application are intended to distinguish one referenced item from another without having a special meaning in technical connotation (e.g., should not be construed as emphasizing a degree or order of importance, etc.). The terms "comprising," "including," "having," and the like, are intended to be inclusive and mean "not limited to" (some elements, components, materials, steps, etc.).
In the present application, terms such as "upper", "lower", "left", "right", "middle", and the like are generally used for easy visual understanding with reference to the drawings, and are not intended to absolutely limit the positional relationship in an actual product. Changes in these relative positional relationships are also considered to be within the scope of the present disclosure without departing from the technical concepts disclosed in the present disclosure.
Example one
Referring to fig. 1 and fig. 2, the present embodiment provides an intelligent identification method for rural land reclamation potential, including:
s1: acquiring rural land space range vector data and high-resolution remote sensing image data to be renovated;
specifically, the initial relevant data comprises rural land space range vector data to be remediated and high-resolution remote sensing image data. In order to accurately identify the land reclamation potential land blocks by using the target image features of the rural land to be reclaimed, the resolution of the used high-resolution remote sensing image is within 1 meter, and the vector data precision of the spatial range of the rural land to be reclaimed is also within the range.
S2: preprocessing the high-resolution remote sensing image data;
specifically, the data preprocessing is mainly used for carrying out geometric correction on the high-resolution remote sensing image and carrying out projection conversion by taking the projection space of the rural land vector data to be rectified as a reference. Because various geometric deformations exist in the satellite remote sensing image, the aerial remote sensing image or the unmanned aerial vehicle image in the acquisition process. When the vector data of the spatial range of the rural land to be remediated is applied, geometric correction and projection conversion are needed to be carried out on the high-resolution image, and in order to ensure the spatial consistency with the vector data of the spatial range of the rural land to be remediated, the geometric correction precision is not lower than the precision of the vector data of the spatial range of the rural land to be remediated and is not more than 2 pixels. The specific correction and projection conversion method can refer to the remote sensing image processing related documents or software application specifications, and the description of the invention is not repeated.
S3: carrying out constraint segmentation on the preprocessed high-resolution remote sensing image data by utilizing the rural land space range vector data to be rectified to obtain an optimal segmentation scale, and generating a rural land object to be rectified;
specifically, the rural land object to be remediated is obtained by adopting a remote sensing image segmentation mode based on the rural land space range vector data to be remediated and the high-resolution remote sensing image data. The vector data of the rural land to be rectified is used for eliminating the road range in the village, the high-resolution remote sensing image is further constrained and segmented, a reasonable range is provided for subsequent rural land object feature extraction to be rectified, and the finally segmented and extracted image data is consistent with the vector range of the existing rural land to be rectified.
More specifically, under the support of eCoginization 9.0, the high-resolution remote sensing image is subjected to multi-scale segmentation by combining the characteristics of the image, corresponding parameters are continuously debugged to obtain the optimal segmentation scale, and a rural land image object to be rectified is generated. Further utilizing vector data of the rural land to be regulated to judge whether an object in the constraint segmentation result is a rural land object to be regulated or not, and forming a rural land object set to be regulated by all the rural land objects to be regulated into an O = { r = 1 ，r 2 ，…，r n }，r i Is one of the rural land objects to be remediated.
S4: extracting the land cover characteristics of the rural land object to be renovated;
specifically, the land cover characteristics of the rural land object to be remedied are used as the basis for classification and identification of the rural land to be remedied, and the classification and identification characteristics of the land cover comprise two types of spectral bands and spectral indexes. In order to improve the precision of classification and identification of the rural land to be remediated, the spectrum band and the spectrum index are used as input variables of the random forest classification and identification of the rural land to be remediated.
The spectral bands of the input variables are mainly red (R), green (G), blue (B) and Near Infrared (NIR), and the spectral indexes of the input variables are mainly normalized vegetation index (NDVI), normalized water body index (NDWI) and specific value residential area index (RRI);
wherein NDVI = ((NIR-R)/(NIR + R))
NIR = pixel value of near infrared band; r = pixel value of red light band
NDWI＝(G-NIR)/(G+NIR)
NIR = pixel value of near infrared band; g = pixel value of green band
RRI＝B/NIR
NIR = pixel value of near infrared band; b = pixel value of blue band
The object features of the two types are combined to construct a multi-feature set for classifying and identifying the random forest of the rural land to be remediated, and the accuracy of classification and identification is improved.
S5: inputting the land cover characteristics into a trained random forest classification model to perform classification and identification on the rural land to be renovated;
specifically, the method comprises the steps of training a random forest classification model and carrying out verification testing by using a to-be-remediated rural land object marked as a sample to obtain the random forest classification model meeting the precision requirement, carrying out classification identification on the to-be-remediated rural land, and providing a basis for identification of rural land remediation potential. The method comprises the following three substeps:
s51: establishing a rural land object sample data set to be renovated;
specifically, a rural land object to be renovated is used as a father class, and a sample is selected based on a high-resolution remote sensing image. According to the characteristics of the land cover of the rural land to be renovated, the marked samples are divided into grasslands, buildings, water bodies, trees, bare land, hardened open land, sundry stacking land and the like. The method for visual interpretation of the ArcGIS platform is adopted to vectorize typical samples of grassland, buildings, water bodies, trees, bare land, hardened open land and other land types, and a marked sample set is divided into two parts, namely a training sample and a verification sample. The number of training samples accounts for 70% of the total sample set, the number of testing samples accounts for 30%, the training samples are used for feature selection and random forest classification model modeling, and the verification samples are used for precision evaluation.
Classifying and assigning different types of samples, exporting a label file used as the characteristics of the obtained samples, further superposing the processed high-resolution remote sensing image and the sample label file, reading the sample class attribute value and the sample space coordinate information in the sample label file, correspondingly capturing the sample spectrum characteristics under the pixels of the same space coordinate position of the remote sensing image as the sample characteristic attributes, and establishing a sample data set combining the sample class, the sample coordinates and the sample characteristics.
S52: identifying internal land class of rural land object to be renovated
The random forest is used as an important integrated machine learning algorithm, has the characteristics of high efficiency, flexibility, accuracy, strong selection capability and the like, and is widely applied to medium and high resolution image classification. The random forest comprises a plurality of decision trees, and for each decision tree, a classification result is obtained when one sample is input. The random forest classification algorithm can process high-dimensional data, and due to the randomness of feature subset selection, the random forest classification does not need feature selection and fully guarantees the independence between trees.
Based on the high-resolution remote sensing image, the random forest classification method is combined to identify the ground object types with fine and complex characteristics in a large range, a random forest classifier is selected to perform learning classification, the rural land object to be remediated is taken as a father class, a random forest classification model is constructed based on the sample data set and the characteristic set of the rural land object to be remediated, and the internal ground class identification of the rural land object to be remediated is performed.
Referring to fig. 3, the learning process of the random forest classification model includes the following steps:
s521: giving a training set with the size of N and the maximum tree number M in a random forest, and randomly and reversely selecting N samples from training samples as a sample training set of each tree;
s522: setting the dimension of a sample attribute as Q, specifying a constant Q, wherein Q is less than Q, randomly selecting Q from the attributes Q as an attribute subset of the tree, and selecting the optimal Q from the attributes Q when the tree grows;
s523: calculating the number M of the decision trees generated in the process, comparing M with M, repeating the steps S521-S522 if M is smaller than M, and skipping the cycle if M is not smaller than M to generate M random forests of decision trees;
s524: all trees grow as much as possible without a pruning process;
s525: for the data set to be classified, after the decision of each tree, the final classification result is determined based on the classification that results in the highest number of votes in the decision.
Ideally, the result of feature selection is that each branch of the decision tree is the cleanest class, i.e., for each branch, the samples output by the same leaf node will be assigned to the same class.
S53: identification and verification of land classes inside to-be-remediated rural land object
And evaluating the precision of the classifier identification result after random forest learning. If the verification result shows that the recognition precision is low, returning to adjust the training sample or adjust the parameters of the random forest classification model; and if the accuracy requirement is met, storing the classification model and applying the classification model to the identification of the non-sample object of the rural land to be remediated.
S6: the identified idle type is used as a rural land reclamation potential land parcel;
specifically, a random forest classification model meeting the precision requirement is adopted to perform land classification recognition of non-sample objects of rural land to be remediated, and land cover types such as buildings, trees, grasslands, water bodies, bare land, hardened open land, sundry stacking land and the like in the rural land are recognized. And judging whether the rural land is idle or not based on the land cover characteristics, selecting idle placement types such as grasslands, bare lands, hardened open lands, sundries stacking lands and the like as rural land reclamation potential land parcels, and outputting the primary identification result by using a vector or grid map.
S7: and correcting the identification result of the remediation potential land parcel which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land parcel.
Specifically, according to the preliminary identification result of the rural land reclamation potential land parcel, the utilization status of the rural land reclamation potential land parcel is subjected to participatory investigation, and the identification result of the reclamation potential land parcel which does not conform to the utilization status is corrected. And carrying out pattern spot shape correction, pattern spot attribute correction and topology inspection on the rural land reclamation potential pattern spots to obtain block vector or grid data of the rural land reclamation potential, and summarizing and establishing a spatial data layer of the rural land reclamation potential.
The method and the device make full use of the advantage of rich and diversified ground feature information of the high-resolution remote sensing image, realize intelligent identification of the rural land improvement potential by adopting a mode of combining random forest classification identification and participatory investigation, overcome the defects of the traditional rural land improvement potential identification technology in the aspects of identification accuracy, space uncertainty, potential availability and the like, reduce the workload of rural land improvement potential identification, and provide reference basis for accelerating the rural stock quantity and land use utilization work.
Example two
This embodiment provides a rural area soil renovation potentiality intelligent recognition device, includes:
the data acquisition module is used for acquiring the vector data of the rural land space range to be renovated and the high-resolution remote sensing image data;
the preprocessing module is used for preprocessing the high-resolution remote sensing image data;
the to-be-remediated rural land object generating module is used for carrying out constraint segmentation on the preprocessed high-resolution remote sensing image data by utilizing the to-be-remediated rural land space range vector data to obtain an optimal segmentation scale and generating a to-be-remediated rural land object;
the to-be-remediated rural land object feature extraction module is used for extracting the land cover features of the to-be-remediated rural land object;
the classification recognition module is used for inputting the land cover characteristics into a trained random forest classification model to perform classification recognition on the rural land to be renovated;
the rural land reclamation potential plot identification module is used for taking the identified idle type as a rural land reclamation potential plot;
and the potential correction module is used for correcting the identification result of the remediation potential land block which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land block.
The specific definition of the intelligent identification device for the rural land reclamation potential can be referred to the definition of the intelligent identification method for the rural land reclamation potential in the above, and is not described in detail here.
All the technical features of the above embodiments can be arbitrarily combined (as long as there is no contradiction between the combinations of the technical features), and for brevity of description, all the possible combinations of the technical features in the above embodiments are not described; these examples, which are not explicitly described, should be considered to be within the scope of the present description.
The present application has been described in considerable detail with reference to certain embodiments and examples thereof. It should be understood that several general adaptations or further innovations of these specific embodiments can also be made based on the technical idea of the present application; however, such conventional modifications and further innovations can also fall into the scope of the claims of the present application as long as they do not depart from the technical idea of the present application.
Claims (10)
1. An intelligent identification method for rural land remediation potential is characterized by comprising the following steps:
acquiring space range vector data and high-resolution remote sensing image data of rural land to be renovated;
preprocessing the high-resolution remote sensing image data;
carrying out constraint segmentation on the preprocessed high-resolution remote sensing image data by utilizing the rural land space range vector data to be rectified to obtain an optimal segmentation scale, and generating a rural land object to be rectified;
extracting the land cover characteristics of the rural land object to be renovated;
inputting the land cover characteristics into a trained random forest classification model to perform classification and identification on the rural land to be renovated;
the identified idle type is used as a rural land reclamation potential plot;
and correcting the identification result of the remediation potential land parcel which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land parcel.
2. The intelligent rural land reclamation potential identification method according to claim 1, wherein the training of the random forest classification model to establish the target sample data set of the rural land to be remediated specifically comprises:
selecting a sample based on the high-resolution remote sensing image data by taking the rural land object to be renovated as a father class;
marking the type of the sample according to the land cover characteristics of the rural land object to be renovated;
classifying and assigning different types of samples to be used as sample label files;
superposing the sample label file and the high-resolution remote sensing image data;
and reading the sample class attribute value and the sample space coordinate information in the sample label file, correspondingly capturing sample spectral characteristics under the same space coordinate position pixel with the high-resolution remote sensing image data as sample characteristic attributes, and establishing a rural land object sample data set to be renovated.
3. The intelligent rural land reclamation potential identification method of claim 1, wherein the resolution of the high-resolution remote sensing image data is within 1 meter.
4. The intelligent rural land reclamation potential identification method according to claim 1, wherein the preprocessing of the spatial range vector data of the rural land to be remediated and the high-resolution remote sensing image data specifically comprises: and carrying out geometric correction on the high-resolution remote sensing image data, and carrying out projection conversion by taking the projection space of the vector data in the rural land space range to be rectified as a reference.
5. The intelligent rural land remediation potential identification method of claim 4, wherein the geometric correction precision is not lower than the precision of the rural land space range vector data to be remediated and is not more than 2 pixels.
6. The intelligent rural land reclamation potential identification method of claim 1, wherein the land cover characteristics comprise spectral bands and spectral indices.
7. The intelligent rural land remediation potential identification method of claim 6, wherein the spectral bands are red, green, blue and near infrared, and the spectral indices are a normalized vegetation index, a normalized water body index and a ratio residential land index;
wherein, the first and the second end of the pipe are connected with each other,
NDVI = ((NIR-R)/(NIR + R)), NDVI is normalized vegetation index, NIR is pixel value of near infrared band, and R is pixel value of red light band;
NDWI = (G-NIR)/(G + NIR), NDWI is normalized water body index, NIR is pixel value of near infrared band, G is pixel value of green light band;
RRI = B/NIR, RRI being a specific residential area index, NIR being a pixel value of the near infrared band; and B is the pixel value of the blue light band.
8. The intelligent rural land reclamation potential identification method as recited in claim 1, wherein the idle types comprise grassland, bare land, hardened open land and sundry piled land.
9. The method for intelligently identifying rural land reclamation potential as claimed in claim 1, wherein the step of correcting the identification result of the remediation potential land blocks which do not conform to the current utilization situation comprises the following steps: spot shape correction, spot attribute modification and topology inspection.
10. The utility model provides a rural area soil renovation potentiality intelligent recognition device which characterized in that includes:
the data acquisition module is used for acquiring the vector data of the rural land space range to be renovated and the high-resolution remote sensing image data;
the preprocessing module is used for preprocessing the high-resolution remote sensing image data;
the rural land object to be remediated generating module is used for utilizing the vector data of the spatial range of the rural land to be remediated to carry out constraint segmentation on the preprocessed high-resolution remote sensing image data to obtain an optimal segmentation scale and generate a rural land object to be remediated;
the to-be-remediated rural land object feature extraction module is used for extracting the land cover features of the to-be-remediated rural land object;
the classification recognition module is used for inputting the land cover characteristics into a trained random forest classification model to perform classification recognition on the rural land to be remedied;
the rural land reclamation potential land parcel identification module is used for taking the identified idle type as a rural land reclamation potential land parcel;
and the potential correction module is used for correcting the identification result of the remediation potential land block which does not conform to the current utilization situation according to the participation type investigation result of the rural land remediation potential land block.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211074997.6A CN115601651A (en) | 2022-09-02 | 2022-09-02 | Intelligent identification method and device for rural land remediation potential |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211074997.6A CN115601651A (en) | 2022-09-02 | 2022-09-02 | Intelligent identification method and device for rural land remediation potential |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115601651A true CN115601651A (en) | 2023-01-13 |
Family
ID=84842435
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211074997.6A Pending CN115601651A (en) | 2022-09-02 | 2022-09-02 | Intelligent identification method and device for rural land remediation potential |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115601651A (en) |
-
2022
- 2022-09-02 CN CN202211074997.6A patent/CN115601651A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9064151B2 (en) | Device and method for detecting plantation rows | |
| CN102521624B (en) | Classification method for land use types and system | |
| CN113360587B (en) | Land surveying and mapping equipment and method based on GIS technology | |
| CN107239795A (en) | SAR image change detecting system and method based on sparse self-encoding encoder and convolutional neural networks | |
| CN110263735A (en) | A method of tree species classification being carried out to artificial forest high-spectral data using Three dimensional convolution neural network | |
| CN113449594A (en) | Multilayer network combined remote sensing image ground semantic segmentation and area calculation method | |
| CN114241326B (en) | Progressive intelligent production method and system for ground feature elements of remote sensing images | |
| CN104809724A (en) | Automatic precise registration method for multiband remote sensing images | |
| CN111222536A (en) | City green space information extraction method based on decision tree classification | |
| CN111914767A (en) | Scattered-pollution enterprise detection method and system based on multi-source remote sensing data | |
| CN106228136A (en) | Panorama streetscape method for secret protection based on converging channels feature | |
| CN110596008A (en) | Plot-based soil nutrient digital mapping method for agricultural region of Chinese Hongsheng plain | |
| CN109447111A (en) | A kind of remote sensing supervised classification method based on subclass training sample | |
| CN113705990A (en) | Natural resource information processing method and device, electronic equipment and storage medium | |
| CN116740474A (en) | Remote sensing image classification method based on anchoring stripe attention mechanism | |
| CN116486289A (en) | Gas pipeline high-consequence area identification method driven by multi-source data and knowledge | |
| CN107480634A (en) | A kind of geographical national conditions ground mulching monitoring method based on multistage target classification | |
| Ming et al. | Cropland extraction based on OBIA and adaptive scale pre-estimation | |
| CN113139511B (en) | Water and soil conservation production and construction project disturbance pattern spot extraction method, medium and equipment | |
| CN116822980A (en) | Base vector guided typical earth surface element intelligent monitoring method and system | |
| CN110929739B (en) | Automatic impervious surface range remote sensing iterative extraction method | |
| CN115601651A (en) | Intelligent identification method and device for rural land remediation potential | |
| CN116310802A (en) | Method and device for monitoring change of residence based on multi-scale fusion model | |
| CN105760421A (en) | Land-use type classifying method | |
| CN111598036B (en) | Urban group geographic environment knowledge base construction method and system of distributed architecture |
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 |