CN117114197B - Symptom migration range identification method based on deep learning - Google Patents
Symptom migration range identification method based on deep learning Download PDFInfo
- Publication number
- CN117114197B CN117114197B CN202311156732.5A CN202311156732A CN117114197B CN 117114197 B CN117114197 B CN 117114197B CN 202311156732 A CN202311156732 A CN 202311156732A CN 117114197 B CN117114197 B CN 117114197B
- Authority
- CN
- China
- Prior art keywords
- area
- road
- land
- grid
- planning
- 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.)
- Active
Links
- 208000024891 symptom Diseases 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000013508 migration Methods 0.000 title claims abstract description 31
- 230000005012 migration Effects 0.000 title claims abstract description 31
- 238000013135 deep learning Methods 0.000 title claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 14
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 11
- 230000007704 transition Effects 0.000 claims abstract description 6
- 238000004891 communication Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 14
- 208000011580 syndromic disease Diseases 0.000 claims description 11
- 230000000877 morphologic effect Effects 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000005260 corrosion Methods 0.000 claims description 4
- 230000007797 corrosion Effects 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 abstract description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000013136 deep learning model Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000010339 dilation Effects 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/16—Real estate
- G06Q50/165—Land development
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
- G06V30/414—Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/42—Document-oriented image-based pattern recognition based on the type of document
- G06V30/422—Technical drawings; Geographical maps
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Geometry (AREA)
- Evolutionary Biology (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Educational Administration (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mathematical Physics (AREA)
- Computer Graphics (AREA)
- Computing Systems (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a symptom transition range identification method based on deep learning, which comprises the following steps: s1: acquiring an area map, and dividing the area map into a plurality of grid planning blocks based on the central line of a road in the map; s2: acquiring the sign migration historical data of the sign migration in each grid planning block; s3: training a convolutional neural network model aiming at the residential land, the industrial land, the commercial land and the greening land planning types according to the grid planning block subjected to the symptom migration and the symptom migration history data; s5: sequentially inputting all grid planning blocks into convolutional neural network models aiming at different land planning types for prediction; the method is beneficial to preserving the integrity of the area in each grid, scientifically dividing the area map, and preserving the road parameters in each grid planning block, so that the characteristics of the map form in the grid planning block can be preserved, and the neural network model can be trained in the later period.
Description
Technical Field
The invention relates to the technical field of land planning, in particular to a symptom migration range identification method based on deep learning.
Background
The characterization migration range identification method is a method for identifying and determining the land planning range of an area by analyzing and processing related data such as the existing land utilization, land type, topography and the like. The method mainly carries out comprehensive analysis and feature extraction on regional map attributes through a Geographic Information System (GIS) technology so as to obtain information such as the spatial range and the type of a land planning range. By identifying the land planning range, scientific basis can be provided for decisions such as urban planning, land utilization optimization, resource protection and the like. The method generally comprises the steps of data preprocessing, feature extraction, identification and the like, and can be implemented by using various algorithms and models. The land planning range recognition method is widely applied to the fields of land planning, urban construction and management, environmental protection and the like.
In the existing syndrome range identification technology, a map is generally segmented by adopting a fixed grid, so that the features reserved in segmented blocks are less or incomplete; while at the same time. The existing land planning range identification method has poor pertinence, and is difficult to give out the planning of land types according to different land planning types.
Disclosure of Invention
The invention aims to provide a symptom transition range identification method based on deep learning, which solves the following technical problems:
in the existing syndrome range identification technology, when a map is segmented, fewer or incomplete features are reserved in segmented blocks; meanwhile, the pertinence is poor, and the planning of the land type is difficult to be given for different land planning types.
The aim of the invention can be achieved by the following technical scheme:
The symptom migration range identification method based on deep learning comprises the following steps:
s1: acquiring an area map, and dividing the area map into a plurality of grid planning blocks based on the central line of a road in the map;
s2: acquiring the sign migration historical data of the sign migration in each grid planning block;
S3: training a convolutional neural network model aiming at the residential land, the industrial land, the commercial land and the greening land planning types according to the grid planning block subjected to the symptom migration and the symptom migration history data;
s5: all grid planning blocks are sequentially input into convolutional neural network models aiming at different land planning types for prediction, and the symptom range of a prediction result is identified into a regional map, wherein the regional map displays the symptom range planning suitable for living land, industrial land, commercial land and greening land.
As a further scheme of the invention: the symptom history data comprises the symptom range, the original land type, the land planning type after symptom and the symptom area of the symptom area.
As a further scheme of the invention: the step S3 comprises the following steps:
s301: screening out grid planning blocks subjected to symptom migration, wherein the land planning types after symptom migration in the screened out grid planning blocks are divided into: living, industrial, commercial and greening land;
S302: distributing the screened grid planning blocks to a living land set, an industrial land set, a commercial land set and a greening land set according to the land planning types after the migration;
S303: convolutional neural network models for residential, industrial, commercial and greening land planning types are trained by grid planning blocks and their corresponding symptom transition history data in the residential, industrial, commercial and greening land sets, respectively.
As a further scheme of the invention: the step S1 comprises the following steps:
S101: acquiring regional map data, road network map data in a region and regional boundary data;
s102: determining the whole grid range to be divided according to the regional boundary data;
S103: extracting a road center line according to road network map data in the region;
S104: generating grids for primarily dividing grid planning blocks according to the central line of the road;
S105: merging grids with the area smaller than the minimum value of the area range according to the minimum value of the area of the preset grid area until the area of all the merged grids is larger than the minimum value of the area of the preset grid area, so as to obtain grids for finally dividing the area map;
s106: dividing the regional map into a plurality of grid planning blocks according to grids which divide the regional map finally.
As a further scheme of the invention: in step S103, a road center line is extracted from the road network map data by a skeleton extraction method or a connected region method.
As a further scheme of the invention: in the step S103, the extracting the road center line from the road network map data by the skeleton extraction method includes the following steps:
Extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
Performing morphological operations on the binarized road network image, wherein the morphological operations include corrosion and expansion;
converting the road network image into a topological structure of a road center line through a Zhang-Suen skeleton extraction algorithm;
And optimizing and smoothing the road center line according to the topological structure of the road center line to obtain the extracted road center line.
As a further scheme of the invention: in S103, the method for extracting the road center line from the road network map data by the connected region method includes the following steps:
A1: extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
A2: dividing a white road part in an image into different communication areas through a communication area algorithm;
a3: combining the connected areas with the phase distance smaller than the set threshold value into the same connected area according to the connection relation of the road pixels of each connected area;
a4: optimizing and smoothing the edges of the combined communication areas;
a5: deleting the communication area with the area smaller than the set threshold value;
a6: the obtained communication region of the white road portion has the center line of the communication region of the white road portion as the extracted road center line.
As a further scheme of the invention: in the step S104, generating a preliminary grid for dividing the grid planning block according to the road center line, including the following steps:
and (5) carrying out line segmentation on the center line of the road. Segmenting a road center line according to an intersection or a curve shape to obtain a plurality of segments;
If the distance between two nearest points between different line segments is smaller than a set threshold value, combining the line segments through a Bezier curve fitting algorithm, and fitting the line segments into a smooth curve;
and taking the curve after optimization processing as a grid for dividing the grid planning block.
As a further scheme of the invention: in S105, according to a preset area range threshold of the grid, the combination of the grids with the area smaller than the minimum value of the area range includes the following steps:
S1051: calculating the area of each cell; estimating the area of the cell by calculating the number of points inside the cell or calculating the cell boundary area;
S1052: traversing all the cells in sequence, and taking the cells with the cell areas smaller than the minimum value of the area range as cells to be processed;
s1053: combining the cells to be processed with adjacent cells;
S1054: step S1052 is repeated until there are no cells whose cell areas are smaller than the minimum value of the area range among all the cells.
As a further scheme of the invention: the step S1053 includes the steps of:
Screening all cells adjacent to the cell to be processed;
calculating the areas of the adjacent cells, and finding the cell with the smallest area;
And merging the cells to be processed into the cells with the minimum area by fusing the boundaries of the cells to be processed and the boundaries of the cells with the minimum area.
The invention has the beneficial effects that:
The regional map is divided into the grid planning blocks based on the central line of the road in the map, so that the regional map is reserved for the integrity of the region in each grid, the regional map is scientifically segmented, meanwhile, the road parameters are reserved in each grid planning block, the characteristics of the map form in the grid planning block are reserved, and the training of the neural network model in the later stage is facilitated.
The present invention can transform a complex overall classification problem into a plurality of relatively simple sub-classification problems by partitioning the regions. This may improve the accuracy of classification because the deep learning model may learn each sub-classification problem more intensively. And after the area is divided, the scale of each sub-classification problem is relatively small. The computational resources and time required for deep learning training for each sub-classification problem are less than for classifying the entire region. By using the living, industrial, commercial and greening land as sub-classifications, each sub-classification may have different features and characteristics, and by performing deep learning training for each sub-classification problem, the structure and parameters of the model may be individually adjusted according to the characteristics of each sub-classification problem, so as to better adapt to the features of the living, industrial, commercial and greening land classification problems.
Drawings
The invention is further described below with reference to the accompanying drawings.
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the present invention is a method for identifying a symptom transition range based on deep learning, comprising:
s1: acquiring an area map, and dividing the area map into a plurality of grid planning blocks based on the central line of a road in the map;
s2: acquiring the sign migration historical data of the sign migration in each grid planning block;
S3: training a convolutional neural network model aiming at the residential land, the industrial land, the commercial land and the greening land planning types according to the grid planning block subjected to the symptom migration and the symptom migration history data;
s5: all grid planning blocks are sequentially input into convolutional neural network models aiming at different land planning types for prediction, and the symptom range of a prediction result is identified into a regional map, wherein the regional map displays the symptom range planning suitable for living land, industrial land, commercial land and greening land.
Specifically, the regional map is divided into a plurality of grid planning blocks based on the central line of the road in the map, so that the regional map is reserved for the integrity of the region in each grid, the regional map is scientifically segmented, meanwhile, road parameters are reserved in each grid planning block, the characteristics of the map form in the grid planning block are reserved conveniently, and the training of the neural network model in the later stage is facilitated.
By partitioning the region, a complex overall classification problem can be converted into a plurality of relatively simple sub-classification problems. This may improve the accuracy of classification because the deep learning model may learn each sub-classification problem more intensively. And after the area is divided, the scale of each sub-classification problem is relatively small. The computational resources and time required for deep learning training for each sub-classification problem are less than for classifying the entire region. By using the living land, the industrial land, the commercial land and the greening land as the sub-classifications, each sub-classification may have different characteristics and features, and by performing deep learning training on each sub-classification problem, the structure and parameters of the model may be individually adjusted according to the features of each sub-classification problem, so as to better adapt to the characteristics of each classification problem.
In one embodiment of the present invention, the symptom history data includes a symptom range, an original land type, a post-symptom land planning type, and a symptom area of the symptom area.
In one embodiment of the present invention, the step S3 includes the following steps:
s301: screening out grid planning blocks subjected to symptom migration, wherein the land planning types after symptom migration in the screened out grid planning blocks are divided into: living, industrial, commercial and greening land;
S302: distributing the screened grid planning blocks to a living land set, an industrial land set, a commercial land set and a greening land set according to the land planning types after the migration;
S303: convolutional neural network models for residential, industrial, commercial and greening land planning types are trained by grid planning blocks and their corresponding symptom transition history data in the residential, industrial, commercial and greening land sets, respectively.
Specifically, by using the living, industrial, commercial and greening land as sub-classifications, each sub-classification may have different features and characteristics, and by performing deep learning training on each sub-classification problem, the structure and parameters of the model may be individually adjusted according to the features of each sub-classification problem, so as to better adapt to the features of each classification problem. Meanwhile, through deep learning training aiming at each sub-classification problem, more training samples and labels can be obtained, and the generalization capability of the model is improved. The deep learning model of each sub-classification problem can better capture the characteristics of the respective classification problem, thereby improving the prediction capability of the model for different land planning types.
In one embodiment of the present invention, the step S1 includes the following steps:
S101: acquiring regional map data, road network map data in a region and regional boundary data;
s102: determining the whole grid range to be divided according to the regional boundary data;
S103: extracting a road center line according to road network map data in the region;
S104: generating grids for primarily dividing grid planning blocks according to the central line of the road;
S105: merging grids with the area smaller than the minimum value of the area range according to the minimum value of the area of the preset grid area until the area of all the merged grids is larger than the minimum value of the area of the preset grid area, so as to obtain grids for finally dividing the area map;
s106: dividing the regional map into a plurality of grid planning blocks according to grids which divide the regional map finally.
Specifically, the grids for primarily dividing the grid planning blocks are generated through the road center line, the regional map is divided into a plurality of grid planning blocks, the integrity of the region in each grid is reserved, the regional map is scientifically divided, and the grid planning efficiency is improved;
the grids with the area smaller than the minimum value of the area range are combined according to the minimum value of the area of the preset grid area, so that the area of the grids is not too small, the characteristics in the grids are not too small, and the training effect of the model is influenced.
In one embodiment of the present invention, in the step S103, the road center line is extracted from the road network map data by a skeleton extraction method or a connected region method.
In one embodiment of the present invention, in the step S103, the extracting the road center line from the road network map data by the skeleton extraction method includes the following steps:
Extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
Performing morphological operations on the binarized road network image, wherein the morphological operations include corrosion and expansion;
converting the road network image into a topological structure of a road center line through a Zhang-Suen skeleton extraction algorithm;
And optimizing and smoothing the road center line according to the topological structure of the road center line to obtain the extracted road center line.
Specifically, the binarized road network image is eroded by changing the shape of the image by performing a local area reduction operation on the image. Each pixel point in the road network image is replaced by the minimum value in the neighborhood of the pixel point through corrosion operation, so that the small area with thinner edge and isolated area is disappeared or reduced. The erosion operation may make the tiny objects in the road network image disappear, the connection area separate, and the size is reduced. The binarized road network image is inflated by changing the shape of the image by performing an expanding operation of a local area on the road network image. The dilation operation replaces each pixel in the road network image with a maximum in its neighborhood, connecting the boundary dilation with the surrounding area. The binarized road network image is subjected to morphological operation, noise can be removed, and broken road parts are connected, so that the road network image becomes more coherent.
Meanwhile, converting the road network image into a topological structure of a road center line through a Zhang-Suen skeleton extraction algorithm; the road network image can be converted into a skeleton map, i.e. a topology of the road centre line. Compared with the original image, the skeleton diagram is simpler and more compact, and redundant information in the image is reduced. Meanwhile, the topological structure of the skeleton diagram keeps the connection relation between roads, so that the road center line can be extracted more accurately. The calculation amount of the skeleton diagram is smaller, and the skeleton diagram is more efficient for application scenes needing to process large-scale road network data. Through the Zhang-Suen skeleton extraction algorithm, the data storage and calculation amount can be reduced, and the processing efficiency is improved. The skeleton extraction algorithm has certain robustness to road change, and can remove small-section breaks or intricate road branches in the image, so that the topological structure of the road center line is clearer and simpler.
In one embodiment of the present invention, in S103, the extracting the road center line from the road network map data by the connected area method includes the following steps:
A1: extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
A2: dividing a white road part in an image into different communication areas through a communication area algorithm;
a3: combining the connected areas with the phase distance smaller than the set threshold value into the same connected area according to the connection relation of the road pixels of each connected area;
a4: optimizing and smoothing the edges of the combined communication areas;
a5: deleting the communication area with the area smaller than the set threshold value;
a6: the obtained communication region of the white road portion has the center line of the communication region of the white road portion as the extracted road center line.
In one embodiment of the present invention, in the step S104, generating a preliminary grid for dividing the grid planning block according to the road center line includes the following steps:
and (5) carrying out line segmentation on the center line of the road. Segmenting a road center line according to an intersection or a curve shape to obtain a plurality of segments;
If the distance between two nearest points between different line segments is smaller than a set threshold value, combining the line segments through a Bezier curve fitting algorithm, and fitting the line segments into a smooth curve;
and taking the curve after optimization processing as a grid for dividing the grid planning block.
Specifically, the distance between two nearest points between different line segments is smaller than the set threshold, the line segments are combined through a Bezier curve fitting algorithm, the separated areas among dense small roads appearing on a map are smaller, discontinuous conditions are easy to appear among the small roads, and the line segments are fitted into a smooth curve, so that more characteristics are reserved in the grid, and the situation that too many area grid surfaces are smaller than the set threshold is avoided.
In one embodiment of the present invention, in S105, according to a preset area range threshold of the grid, the combining the grids with the area smaller than the area range minimum value includes the following steps:
S1051: calculating the area of each cell; estimating the area of the cell by calculating the number of points inside the cell or calculating the cell boundary area;
S1052: traversing all the cells in sequence, and taking the cells with the cell areas smaller than the minimum value of the area range as cells to be processed;
s1053: combining the cells to be processed with adjacent cells;
S1054: step S1052 is repeated until there are no cells whose cell areas are smaller than the minimum value of the area range among all the cells.
In one embodiment of the present invention, the step S1053 includes the steps of:
Screening all cells adjacent to the cell to be processed;
calculating the areas of the adjacent cells, and finding the cell with the smallest area;
And merging the cells to be processed into the cells with the minimum area by fusing the boundaries of the cells to be processed and the boundaries of the cells with the minimum area.
In the description of the present invention, it should be understood that the terms "upper," "lower," "left," "right," and the like indicate an orientation or a positional relationship based on that shown in the drawings, and are merely for convenience of description and for simplifying the description, and do not indicate or imply that the apparatus or element in question must have a specific orientation, as well as a specific orientation configuration and operation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present invention, unless otherwise indicated, the meaning of "a plurality" is two or more.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "connected," and the like are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
The foregoing describes one embodiment of the present invention in detail, but the description is only a preferred embodiment of the present invention and should not be construed as limiting the scope of the invention. All equivalent changes and modifications within the scope of the present invention are intended to be covered by the present invention.
Claims (5)
1. The symptom migration range identification method based on deep learning is characterized by comprising the following steps of:
s1: acquiring an area map, and dividing the area map into a plurality of grid planning blocks based on the central line of a road in the map;
s2: acquiring the sign migration historical data of the sign migration in each grid planning block;
S3: training a convolutional neural network model aiming at the residential land, the industrial land, the commercial land and the greening land planning types according to the grid planning block subjected to the symptom migration and the symptom migration history data;
S5: sequentially inputting all grid planning blocks into convolutional neural network models aiming at different land planning types for prediction, and identifying the symptom range of a prediction result into a regional map, wherein the regional map displays the symptom range planning suitable for living land, industrial land, commercial land and greening land;
The step S1 comprises the following steps:
S101: acquiring regional map data, road network map data in a region and regional boundary data;
s102: determining the whole grid range to be divided according to the regional boundary data;
S103: extracting a road center line from road network map data by a skeleton extraction method or a connected region method;
S104: generating grids for primarily dividing grid planning blocks according to the central line of the road;
S105: merging grids with the area smaller than the minimum value of the area range according to the minimum value of the area of the preset grid area until the area of all the merged grids is larger than the minimum value of the area of the preset grid area, so as to obtain grids for finally dividing the area map;
s106: dividing the regional map into a plurality of grid planning blocks according to grids which divide the regional map finally;
In the step S103, the extracting the road center line from the road network map data by the skeleton extraction method includes the following steps:
Extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
Performing morphological operations on the binarized road network image, wherein the morphological operations include corrosion and expansion;
converting the road network image into a topological structure of a road center line through a Zhang-Suen skeleton extraction algorithm;
Optimizing and smoothing the road center line according to the topological structure of the road center line to obtain an extracted road center line;
In S103, the method for extracting the road center line from the road network map data by the connected region method includes the following steps:
A1: extracting a road network image from the road network map, binarizing the road network image, setting a road part to be white, and setting a non-road part to be black;
A2: dividing a white road part in an image into different communication areas through a communication area algorithm;
a3: combining the connected areas with the phase distance smaller than the set threshold value into the same connected area according to the connection relation of the road pixels of each connected area;
a4: optimizing and smoothing the edges of the combined communication areas;
a5: deleting the communication area with the area smaller than the set threshold value;
a6: the obtained communication area of the white road part takes the central line of the communication area of the white road part as the extracted central line of the road;
in the step S104, generating a preliminary grid for dividing the grid planning block according to the road center line, including the following steps:
The method comprises the steps of carrying out line segmentation on a road center line, and segmenting the road center line according to an intersection or a curve shape to obtain a plurality of line segments;
If the distance between two nearest points between different line segments is smaller than a set threshold value, combining the line segments through a Bezier curve fitting algorithm, and fitting the line segments into a smooth curve;
and taking the curve after optimization processing as a grid for dividing the grid planning block.
2. The depth learning-based syndrome identification method according to claim 1, wherein the syndrome history data includes syndrome range, original land type, post-syndrome land planning type, and syndrome area of a syndrome area.
3. The depth learning-based syndrome identification method according to claim 2, wherein the step S3 comprises the steps of:
s301: screening out grid planning blocks subjected to symptom migration, wherein the land planning types after symptom migration in the screened out grid planning blocks are divided into: living, industrial, commercial and greening land;
S302: distributing the screened grid planning blocks to a living land set, an industrial land set, a commercial land set and a greening land set according to the land planning types after the migration;
S303: convolutional neural network models for residential, industrial, commercial and greening land planning types are trained by grid planning blocks and their corresponding symptom transition history data in the residential, industrial, commercial and greening land sets, respectively.
4. The depth learning-based syndrome identification method according to claim 1, wherein in S105, according to a preset area range threshold of the grid, the grids with areas smaller than the minimum value of the area range are combined, and the method comprises the following steps:
S1051: calculating the area of each cell; estimating the area of the cell by calculating the number of points inside the cell or calculating the cell boundary area;
S1052: traversing all the cells in sequence, and taking the cells with the cell areas smaller than the minimum value of the area range as cells to be processed;
s1053: combining the cells to be processed with adjacent cells;
S1054: step S1052 is repeated until there are no cells whose cell areas are smaller than the minimum value of the area range among all the cells.
5. The deep learning-based syndrome identification method according to claim 4, wherein the step S1053 includes the steps of:
Screening all cells adjacent to the cell to be processed;
calculating the areas of the adjacent cells, and finding the cell with the smallest area;
And merging the cells to be processed into the cells with the minimum area by fusing the boundaries of the cells to be processed and the boundaries of the cells with the minimum area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311156732.5A CN117114197B (en) | 2023-09-08 | 2023-09-08 | Symptom migration range identification method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311156732.5A CN117114197B (en) | 2023-09-08 | 2023-09-08 | Symptom migration range identification method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117114197A CN117114197A (en) | 2023-11-24 |
| CN117114197B true CN117114197B (en) | 2024-09-10 |
Family
ID=88794672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311156732.5A Active CN117114197B (en) | 2023-09-08 | 2023-09-08 | Symptom migration range identification method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117114197B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105761192A (en) * | 2016-02-19 | 2016-07-13 | 北京舜土国源信息技术有限公司 | Intelligent method and intelligent integrated system for village-town area land use planning |
| CN113297174A (en) * | 2021-05-24 | 2021-08-24 | 中南大学 | Land use change simulation method based on deep learning |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309248B (en) * | 2019-06-26 | 2022-04-29 | 东南大学 | Method for automatically dividing traffic cells of traffic road network based on Voronoi diagram |
| CN116484266B (en) * | 2023-05-18 | 2023-11-24 | 广东国地规划科技股份有限公司 | Fine urban land type recognition model training method |
-
2023
- 2023-09-08 CN CN202311156732.5A patent/CN117114197B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105761192A (en) * | 2016-02-19 | 2016-07-13 | 北京舜土国源信息技术有限公司 | Intelligent method and intelligent integrated system for village-town area land use planning |
| CN113297174A (en) * | 2021-05-24 | 2021-08-24 | 中南大学 | Land use change simulation method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117114197A (en) | 2023-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108986119B (en) | Image segmentation method and device, computer equipment and readable storage medium | |
| CN110135351B (en) | Built-up area boundary identification method and equipment based on urban building space data | |
| CN108415975B (en) | BDCH-DBSCAN-based taxi passenger carrying hot spot identification method | |
| WO2018107939A1 (en) | Edge completeness-based optimal identification method for image segmentation | |
| CN105975913B (en) | Road network extraction method based on adaptive cluster learning | |
| CN110533064B (en) | Partial discharge map mode identification method and system of GIS (geographic information System) equipment | |
| CN110956207B (en) | Method for detecting full-element change of optical remote sensing image | |
| CN111291811B (en) | Region dividing and classifying method, device and computer readable storage medium | |
| CN113807301B (en) | Automatic extraction method and automatic extraction system for newly-added construction land | |
| CN116911055B (en) | Digital twinning-based intelligent city planning management system | |
| CN116778146A (en) | Road information extraction method and device based on multi-mode data | |
| CN103198479A (en) | SAR image segmentation method based on semantic information classification | |
| CN113724279B (en) | System, method, equipment and storage medium for automatically dividing traffic cells into road networks | |
| CN115239644B (en) | Concrete defect identification method, device, computer equipment and storage medium | |
| CN105893703A (en) | Polygon-based main road selection method of urban road network | |
| CN113240623A (en) | Pavement disease detection method and device | |
| CN116543310B (en) | Road line extraction method based on Voronoi diagram and kernel density | |
| CN116935065A (en) | Lane line instance detection method and system based on fusing and fusion | |
| CN114299394A (en) | Intelligent interpretation method for remote sensing image | |
| CN113902830B (en) | Method for generating track road network | |
| US20240200973A1 (en) | Method, data processing apparatus and computer program product for generating map data | |
| CN117114197B (en) | Symptom migration range identification method based on deep learning | |
| CN117132508A (en) | Digital twin data driving method and system based on GIS+BIM technology | |
| CN107832732A (en) | Method for detecting lane lines based on ternary tree traversal | |
| CN115620169B (en) | Building main angle correction method based on regional consistency |
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 | ||
| GR01 | Patent grant |