CN108052876B - Regional development assessment method and device based on image recognition - Google Patents
Regional development assessment method and device based on image recognition Download PDFInfo
- Publication number
- CN108052876B CN108052876B CN201711216014.7A CN201711216014A CN108052876B CN 108052876 B CN108052876 B CN 108052876B CN 201711216014 A CN201711216014 A CN 201711216014A CN 108052876 B CN108052876 B CN 108052876B
- Authority
- CN
- China
- Prior art keywords
- satellite
- picture
- type
- main body
- area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
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
- G06V20/176—Urban or other man-made structures
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Operations Research (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a regional development condition evaluation method based on image recognition, which comprises the following steps: acquiring a satellite picture of a region to be identified; dividing the satellite picture into a plurality of corresponding block pictures according to color clustering; extracting features of each block picture to obtain a corresponding feature vector, and obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and the established main body identification model; merging block pictures of adjacent similar subjects so as to divide the satellite picture into a plurality of type areas; and obtaining the geographic position of each type of area of the satellite picture according to the longitude and latitude data of the map, and calculating the area size of each type of area according to the area of each type of area. The invention also discloses an electronic device, a storage medium and a device. The invention realizes the evaluation of the development condition of the region by identifying the satellite map of the region to obtain the corresponding main body and the area change of the main body.
Description
Technical Field
The invention relates to the field of city evaluation, in particular to a city co-fusion evaluation method and device based on image recognition.
Background
At present, modern cities are the centers of human living, cultural education, comprehensive business, knowledge and technical innovation and the center of gravity of environmental governance. With the increasing proportion of urban population to the general population, the great strategy of our country is to accelerate the urban expansion, drive the surrounding area construction and the development of the radiation satellite city. How to objectively evaluate the development and the co-fusion condition of the city, and the evaluation result is utilized to promote mature planning calibration and optimization, promote the coordinated development of the urban area and realize the strategic decision of the urban development, which becomes one of the problems that people pay more attention to and are worth discussing.
However, the conventional urban development assessment method usually utilizes factors such as urban population indexes or economic indexes to assess, but for a city, the method is a complex human engineering ecosystem which comprises various dynamic factors such as nature, humanity, ecology, economy, society and the like. Therefore, the traditional evaluation method can not reflect the overall action of the city and the development state of comprehensive functions, has data unicity and evaluation one-sidedness, and can not evaluate the co-fusion condition of city construction and surrounding urban area construction; when multiple factors are considered, the data acquisition is incomplete due to the complex requirements on data acquisition, processing and the like and the complex calculation process; in addition, when the standards according to which data are collected and calculated are different, the obtained data indexes have great difference, and the urban co-fusion development state cannot be effectively and quickly evaluated. In addition, evaluation is required not only for changes in a city, such as changes in areas such as harbors, forests, oceans, etc., but the same problem exists, such that the requirements for data acquisition, processing, etc., and the calculation process are complicated, and an effective and rapid evaluation of an area cannot be performed.
Disclosure of Invention
In order to overcome the defects of the prior art, an object of the present invention is to provide a method for evaluating a development status of an area based on image recognition, which can solve the problem that the development status of a certain area cannot be evaluated effectively and quickly in the prior art.
Another object of the present invention is to provide an electronic device, which can solve the problem that the development status of a certain area cannot be effectively and quickly evaluated in the prior art.
It is another object of the present invention to provide a computer-readable storage medium, which can solve the problem in the prior art that the development status of a certain area cannot be evaluated effectively and quickly.
The fourth objective of the present invention is to provide an apparatus for evaluating the development status of a region based on image recognition, which can solve the problem that the development status of a certain region cannot be evaluated effectively and quickly in the prior art.
One of the purposes of the invention is realized by adopting the following technical scheme:
the regional development condition evaluation method based on image recognition comprises the following steps of:
a model establishing step: establishing a subject identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
an acquisition step: acquiring a satellite picture of a region to be identified;
a segmentation step: dividing the satellite picture into a plurality of corresponding block pictures according to color clustering;
an identification step: extracting features of each block picture to obtain a corresponding feature vector, and obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and a main body identification model;
and (3) merging steps: merging block pictures of adjacent similar subjects so as to divide the satellite picture into a plurality of type areas;
a calculation step: and obtaining the geographic position of each type of area of the satellite picture according to the longitude and latitude data of the map, and calculating the area size of each type of area according to the area of each type of area.
Further, the method also comprises the following processing steps: firstly, respectively obtaining a plurality of satellite pictures of a region to be identified in a statistical period, and sequentially performing a segmentation step, an identification step, a combination step and a calculation step on each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
Further, after the plurality of satellite pictures of the area to be identified in the statistical period are obtained, the plurality of satellite pictures of the area to be identified are sequenced according to the sequence of dates.
Further, the method also comprises the following evaluation steps: and scoring the region to be identified according to the area change data of the type region on the same geographical position of each satellite picture of the region to be identified in the statistical period and a preset scoring model.
Further, the model building step further comprises: firstly, acquiring a plurality of satellite pictures of each type of main body; then, extracting the features of each satellite picture of each type of main body to obtain a corresponding feature vector; and finally, carrying out recognition training on the feature vectors corresponding to the plurality of satellite pictures of each type of main body to obtain a feature vector set corresponding to the satellite pictures of each type of main body, namely a main body recognition model.
Further, the step of obtaining further comprises a preprocessing step after the step of obtaining: carrying out a preprocessing process on the satellite picture; wherein the pretreatment process comprises one or more of the following methods in combination: image binarization, interference point removal, centroid alignment and linear interpolation amplification.
The second purpose of the invention is realized by adopting the following technical scheme:
an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the image recognition based region development status assessment method as described above when executing the program.
The third purpose of the invention is realized by adopting the following technical scheme:
a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the image recognition-based region development status assessment method as described above.
The fourth purpose of the invention is realized by adopting the following technical scheme:
the regional development condition evaluation device based on image recognition comprises:
the model establishing module is used for establishing a main body identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
the acquisition module is used for acquiring a satellite picture of an area to be identified; the segmentation module is used for dividing the satellite picture into a plurality of corresponding block pictures according to the color clustering;
the characteristic extraction module is used for extracting the characteristics of each block picture to obtain a corresponding characteristic vector;
the identification module is used for obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and the main body identification model;
the merging module is used for merging the block pictures of the adjacent similar main bodies so as to divide the satellite picture into a plurality of type areas;
and the calculation module is used for obtaining the geographic position of each type region of the satellite picture according to the longitude and latitude data of the map and calculating the area size of each region type according to the area of each region type.
The system further comprises a processing module, a storage module and a processing module, wherein the processing module is used for respectively obtaining a plurality of satellite pictures of the region to be identified in the statistical period, and sequentially executing the steps of segmenting, identifying, combining and calculating each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
Compared with the prior art, the invention has the beneficial effects that: according to the invention, the identification models of various different types of main bodies are established in advance, then the satellite picture of a certain region is processed and the main body contained in the satellite picture is identified, and further the area size of the corresponding main body is obtained, so that the development condition data of the region can be counted according to the area size data of the main body contained in the satellite picture, and the development condition of the region can be effectively and quickly evaluated.
Drawings
FIG. 1 is a flowchart of a method for evaluating a regional development status based on image recognition according to the present invention;
FIG. 2 is a second flowchart of the method for evaluating the development status of a region based on image recognition according to the present invention;
fig. 3 is a block diagram of a regional development status evaluation apparatus based on image recognition according to the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and the detailed description, and it should be noted that any combination of the embodiments or technical features described below can be used to form a new embodiment without conflict.
Examples
The invention analyzes the satellite pictures of a certain area, such as cities, harbors, forests, oceans and the like, by the image recognition technology to recognize the development changes of the area. For example, a satellite picture of a certain region is identified periodically, and then the periodic front-back changes of various factors in the region are compared, so as to evaluate the development condition of the corresponding region according to the changes of various factors.
For example, by analyzing satellite pictures within 10 km of a boundary between two cities, building density changes of businesses, schools, standard residential areas and the like in the area are identified, and further, the urban co-fusion development condition can be evaluated according to the data. For example, the area of a standard residential area is larger and larger, which indicates that the city is more suitable for people to live in; the larger the area of the business center, the more the economic development and the like are focused on the area.
For example, the satellite picture at the junction of the sea and the land is identified, the sea area and the periodic change, the proportion and the like of the land area are obtained through identification, and the change data of the sea and land area are obtained, so that the environment influence condition caused by global warming is evaluated.
For example, the change of regional greening, forest vegetation, farmland area and the like is obtained by identifying the satellite picture of the city, and then the conditions of desertification control and vegetation-terminated environment construction are evaluated. For example, the area of greening, the area of farmland and the area of forest vegetation are gradually increased, which indicates that the environmental improvement is more and more emphasized, the environment is better and better, etc.
For example, data such as whether a port, a harbor and goods stacking change situation are established in the coastal city is obtained by identifying satellite pictures of the coastal city, and then the external trade economy development condition of the coastal city is judged.
As shown in fig. 1, the present invention provides an embodiment of a method for evaluating a regional development status based on image recognition, which specifically includes the following steps:
and S1, acquiring a plurality of satellite pictures of the to-be-identified area in the statistical period, wherein the satellite pictures are satellite pictures of different dates in the statistical period. The area to be identified may be a junction of two cities, a certain city, a coastal city, a sea-land junction, a port, etc. For example, within a statistical period of 10 years, the satellite pictures of the region to be identified are acquired every other year, so that 10 satellite pictures can be obtained. When an area is developed, the infrastructure of the corresponding map changes, for example, the park is enlarged, the residential area is increased, and the like, on the corresponding satellite map, the proportion of each infrastructure in the satellite picture is increased.
And S2, sequencing the plurality of satellite pictures in the area to be identified according to the sequence of the dates. Meanwhile, when the satellite picture is acquired, the satellite picture is preprocessed, and the processing performance of the system on the picture can be improved after the satellite picture is preprocessed. The pretreatment process may include one or more of the following combinations of methods: image binarization, interference point removal, centroid alignment and linear interpolation amplification. Such as converting pictures uniformly to the same size, format, etc. Sequencing the plurality of preprocessed satellite pictures according to the sequence of dates, so that the development condition of the region can be evaluated according to the front-back change of a main body in the plurality of satellite pictures along with the time in a comparative statistical period.
And S3, sequentially selecting one satellite picture from the plurality of satellite pictures of the region to be identified, and identifying to obtain the area size and the geographic position of the type region contained in the corresponding satellite picture.
And S4, comparing the area size change data of the type area on the same geographical position of each satellite picture of the area to be identified in the statistical period according to the time sequence, and evaluating the area to be identified according to the area size change of each type area to obtain an evaluation result.
In addition, when the regions to be identified are evaluated according to the change of the area size of the type regions corresponding to the same geographic position, the evaluation is carried out through a pre-established scoring mechanism, for example, the regions are scored as 1-5 points when the area is increased, and the regions are scored as-1-5 points when the area is decreased; when the area is unchanged, the value is recorded as 0, etc. The corresponding score may also be set according to the specific range of area change. The scoring mechanism is, of course, specifically configured according to specific requirements.
For example, according to the longitude and latitude data, geographic positions corresponding to types of areas such as businesses, schools, standard residential areas and the like within 10 kilometers of a boundary of two cities in a satellite map are analyzed.
Comparing the changes of the front and back areas of the urban buildings and infrastructure construction (such as commercial comprehensive areas, schools, parks, residential districts and the like) in the front and back maps in the statistical period, and further scoring the areas within 10 kilometers of the urban boundary according to a scoring mechanism to judge the construction and migration conditions of urban living areas, central business circles and cultural education.
For example: evaluating the changes of main bodies such as standard houses, schools, businesses, traffic, roads and the like within a range of 10 kilometers of the Guangzhou and Foshan junction, for example, sequencing the areas of each main body of the region according to the time sequence to obtain the data such as the area change of each main body and the area percentage of each main body type occupying the region so as to evaluate the change conditions of various buildings at the Guangfu junction in a certain period, for example, if the area of the standard house is gradually increased along with the change of the time, more and more residents are located at the Guangfu junction; the roads are more and more, which shows that the city co-fusion at the boundary of Guangzhou and Buddha mountain is well developed.
If the transformation condition of the shed area in a certain city is counted, all satellite maps in a counting period of the city are extracted according to the longitude and latitude data of the city, all shed areas in the maps are judged, the area change data of various types of buildings before and after the shed area in all the maps in the period are sequentially calculated, and the transformation progress condition of the shed area in the city can be obtained according to the area statistical data and the time period.
If the area of the boundary house and the commercial type building of a certain city is rapidly expanded, the boundary house and the commercial type building of the certain city reach the boundary of an adjacent city and still have an expansion sign on a construction site, two cities are closely connected through tracks and traffic, and a division area without obvious difference exists, it can be judged that the city is radiated to the adjacent city and county in the periphery, and the sign of integration of the two cities and the city is obvious.
In addition, as shown in fig. 2, the specific implementation method for processing a satellite picture to obtain the area size and the geographic position of the type region included in the satellite picture includes the following steps:
and S31, dividing the satellite picture into a plurality of corresponding block pictures according to the color clustering. Due to the characteristics of the map, the colors displayed in the map are different for different subjects, for example, the colors displayed for rivers, lands, buildings, oceans, roads and the like are all different, so that the satellite picture can be divided into a plurality of different blocks according to the color clustering technology.
And S32, extracting the features of each block picture to obtain a corresponding feature vector.
S33, obtaining the main body corresponding to each block picture according to the feature vector corresponding to each block picture and the main body identification model.
The main body identification model stores a set of feature vectors of satellite pictures of each type of main body, so that when the main body in the region is identified, the satellite pictures in the region are firstly divided into a plurality of blocks, then feature extraction is carried out on each block picture to obtain the feature vectors, the feature vectors of each block picture are matched with the feature vectors corresponding to the satellite pictures of any type of main body in the main body identification model, and then the main body corresponding to each block picture, namely the main body contained in each satellite picture, can be obtained.
For a city that includes various types of standard residential housing, businesses, schools, farmlands, ports, docks, municipal parks, sports centers, etc., each type of infrastructure for the city corresponds to a region of the satellite map display. In order to identify the main bodies contained in the satellite pictures, the invention establishes a set of feature vectors of the satellite pictures of various types of main bodies, namely a main body identification model, by collecting the satellite pictures of various types of main bodies in advance and extracting, identifying and training the feature vectors, and the establishing process is as follows:
A. the method comprises the steps of obtaining satellite pictures of the same main body type and preprocessing each satellite picture. For example, for a city area or a city boundary area, the main body type can be various types of buildings; for a land-sea interface area, the body type may be marine, land, etc.; for urban areas, forest vegetation, farmlands, deserts and the like can be used; for coastal urban areas, the subject type may be port buildings, port cargo, etc.
B. And performing feature extraction on each preprocessed satellite picture to obtain a corresponding feature vector. When extracting the feature vector, for example, according to the combined features of the appearance shape, color, area size, and spatial distribution of the subject type, the satellite picture of the subject is divided into 25 grid regions of 5 × 5, and the ratio of the number of points in each grid region to the total number of points of the article is calculated to obtain a 25-dimensional feature vector.
In addition, there are individual appearance and color combinations, area sizes and spatial arrangements for the bodies of various types of buildings or infrastructures, such as schools: a block of road network partition area comprising a playing field and a roof level building; sports center: buildings and sports grounds with curved surface areas; business body: the area of the continuous curved surface is included after the road network is cut; farmland: the green area is divided into blocks and a pond, and the pond has obvious light reflecting characteristics; a park: large-area lakes with green areas and blue patches; a standard house comprises: the flat roof color blocks are regularly arranged and have similar shapes and areas; a shed area: the color blocks are arranged in a disordered and dense manner and are similar to the plane roof in area.
Therefore, when feature extraction is performed on a satellite picture of a subject, the feature extraction can be performed according to the appearance, color, area-space distribution, and the like of the subject.
C. And performing identification training on the plurality of satellite pictures of the main body of the same type, extracting a plurality of standard feature vectors of the main body of the same type, and establishing a standard feature vector library of the main body of the type. In recognition training, a correct value indicating the type of subject is required.
By the method, a set of a plurality of feature vectors of various different types of subjects can be established, namely the subject recognition model.
During identification, a plurality of block pictures are obtained by dividing the satellite pictures of the area to be identified, then, feature extraction is carried out on each block picture to obtain a corresponding feature vector, and finally, the feature vector of each block picture is compared with a plurality of feature vectors of the satellite pictures of the main body of any type in the identification model one by one. When the similarity reaches more than 80%, the main body corresponding to the feature vector in the identification model is the main body corresponding to the block, and all the main bodies contained in the satellite picture are obtained.
And S34, merging the adjacent block pictures of the same type of main bodies, and further dividing the satellite picture into a plurality of type areas. Each type region includes the same type of body therein. And one defines the merged type area according to the main body type when dividing the type area. For example, for the neighboring block picture a and block picture B, the main subjects of the block picture a and the block picture B are residential cells, the block picture a and the block picture B may be merged to obtain a type region, and the type region may be named as a residential cell. Of course, it is possible that the same type of area is more in the satellite picture, which may be numbered when named, such as residential cell 1, residential cell 2, etc. For example, the merged area type may be a business center area, school area, problem center area, shed house area, sea, land, port cargo, port building, etc.
S35, calculating the area of each type region according to the area of each type region;
and S36, obtaining the geographic position of each type area of the satellite picture according to the longitude and latitude data of the satellite picture.
In addition, there is no sequence for the actual execution of S35 and S36.
For example, in this embodiment, the method is exemplified by a development evaluation of a blend of a city boundary area:
a: and acquiring a plurality of satellite maps at the urban junction in the statistical period, and preprocessing each satellite picture.
For example for an area within 10 km of a two city boundary: firstly, determining a statistical period, and acquiring satellite pictures of a plurality of dates of a city boundary area in the statistical period. For example, the statistical period is five years, then satellite pictures of the city boundary area of each half year are obtained, and 10 satellite pictures are acquired: a1, A2, A3, A4, … and A10, then sequencing the acquired satellite pictures according to the sequence of the dates, and preprocessing each satellite picture.
B: and dividing each satellite map into a plurality of block pictures according to the color clustering. For example, the boundary between the sea and the land in the satellite map is found out by using an edge detection technology according to the color difference of the main bodies such as the land and the sea, and the satellite map of the sea part and the land part is obtained respectively. And judging the road network in the map by color clustering on the satellite map of the land part, and carrying out region segmentation on the satellite map by using the color band of the road network to obtain a plurality of blocks. The road network is changed into a continuous strip-shaped color area after the color clustering processing. Because the corresponding colors of different areas are different for the satellite map, the satellite map can be divided into a plurality of block pictures by carrying out area division through color clustering.
C: and obtaining a main body corresponding to each block picture in each satellite map according to each block picture of each satellite map and the main body identification model.
D: and merging the adjacent block pictures of the same type of main bodies into corresponding type areas. Such as merging adjacent business hubs into business hub areas, merging adjacent residential quarters into residential quarter areas, merging adjacent school buildings, etc. into school areas, merging adjacent cultural education buildings into cultural hub areas, merging adjacent shed areas into shed areas.
E: and calculating the area of the corresponding type region according to the area of each type region.
F: and judging the geographic position of each type of area in the corresponding satellite map according to the longitude and latitude data of each satellite map.
G: and comparing the area size change data of the type area on the same geographical position of each satellite map of the city boundary area in the statistical period according to the time sequence.
H: and scoring the city boundary area according to the area change of the type area on the same geographical position of each satellite map of the city boundary area in the statistical period and a preset scoring model. Therefore, people can obtain the development condition of the city boundary area through the scoring result, for example, the higher the scoring result is, the better the development of the city boundary area is, and the better the city compatibility is.
The invention can be used for identifying the main bodies such as building facilities and the like at the junction area of cities to obtain the evaluation of the city development co-fusion condition, can also be used for identifying the main bodies such as oceans, lands and the like in a global satellite picture to evaluate the environmental influence caused by global climate warming, realizes the urban desertification control and the evaluation of the environment construction condition of vegetation planting by identifying the main bodies such as greenbelts, forest vegetation, farmlands and the like in a certain area, and judges the outward trade economic development condition of the cities by identifying the main bodies arranged at the foundations such as ports and the like of coastal cities, the main bodies of stacked goods and the like.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the program:
a model establishing step: establishing a subject identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
an acquisition step: acquiring a satellite picture of a region to be identified;
a segmentation step: dividing the satellite picture into a plurality of corresponding block pictures according to color clustering;
an identification step: extracting features of each block picture to obtain a corresponding feature vector, and obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and a main body identification model;
and (3) merging steps: merging block pictures of adjacent similar subjects so as to divide the satellite picture into a plurality of type areas;
a calculation step: and obtaining the geographic position of each type of area of the satellite picture according to the longitude and latitude data of the map, and calculating the area size of each type of area according to the area of each type of area.
Further, the method also comprises the following processing steps: firstly, respectively obtaining a plurality of satellite pictures of a region to be identified in a statistical period, and sequentially performing a segmentation step, an identification step, a combination step and a calculation step on each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
Further, after the plurality of satellite pictures of the area to be identified in the statistical period are obtained, the plurality of satellite pictures of the area to be identified are sequenced according to the sequence of dates.
Further, the method also comprises the following evaluation steps: and scoring the region to be identified according to the area change data of the type region on the same geographical position of each satellite picture of the region to be identified in the statistical period and a preset scoring model.
Further, the model building step further comprises: firstly, acquiring a plurality of satellite pictures of each type of main body; then, extracting the features of each satellite picture of each type of main body to obtain a corresponding feature vector; and finally, carrying out recognition training on the feature vectors corresponding to the plurality of satellite pictures of each type of main body to obtain a feature vector set corresponding to the satellite pictures of each type of main body, namely a main body recognition model.
Further, the step of obtaining further comprises a preprocessing step after the step of obtaining: carrying out a preprocessing process on the satellite picture; wherein the pretreatment process comprises one or more of the following methods in combination: image binarization, interference point removal, centroid alignment and linear interpolation amplification.
The invention also provides a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of:
a model establishing step: establishing a subject identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
an acquisition step: acquiring a satellite picture of a region to be identified;
a segmentation step: dividing the satellite picture into a plurality of corresponding block pictures according to color clustering;
an identification step: extracting features of each block picture to obtain a corresponding feature vector, and obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and a main body identification model;
and (3) merging steps: merging block pictures of adjacent similar subjects so as to divide the satellite picture into a plurality of type areas;
a calculation step: and obtaining the geographic position of each type of area of the satellite picture according to the longitude and latitude data of the map, and calculating the area size of each type of area according to the area of each type of area.
Further, the method also comprises the following processing steps: firstly, respectively obtaining a plurality of satellite pictures of a region to be identified in a statistical period, and sequentially performing a segmentation step, an identification step, a combination step and a calculation step on each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
Further, after the plurality of satellite pictures of the area to be identified in the statistical period are obtained, the plurality of satellite pictures of the area to be identified are sequenced according to the sequence of dates.
Further, the method also comprises the following evaluation steps: and scoring the region to be identified according to the area change data of the type region on the same geographical position of each satellite picture of the region to be identified in the statistical period and a preset scoring model.
Further, the model building step further comprises: firstly, acquiring a plurality of satellite pictures of each type of main body; then, extracting the features of each satellite picture of each type of main body to obtain a corresponding feature vector; and finally, carrying out recognition training on the feature vectors corresponding to the plurality of satellite pictures of each type of main body to obtain a feature vector set corresponding to the satellite pictures of each type of main body, namely a main body recognition model.
Further, the step of obtaining further comprises a preprocessing step after the step of obtaining: carrying out a preprocessing process on the satellite picture; wherein the pretreatment process comprises one or more of the following methods in combination: image binarization, interference point removal, centroid alignment and linear interpolation amplification.
As shown in fig. 3, the apparatus for evaluating the development condition of a region based on image recognition includes:
the model establishing module is used for establishing a main body identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
the acquisition module is used for acquiring a satellite picture of an area to be identified; the segmentation module is used for dividing the satellite picture into a plurality of corresponding block pictures according to the color clustering;
the characteristic extraction module is used for extracting the characteristics of each block picture to obtain a corresponding characteristic vector;
the identification module is used for obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and the main body identification model;
the merging module is used for merging the block pictures of the adjacent similar main bodies so as to divide the satellite picture into a plurality of type areas;
and the calculation module is used for obtaining the geographic position of each type region of the satellite picture according to the longitude and latitude data of the map and calculating the area size of each region type according to the area of each region type.
The system further comprises a processing module, a storage module and a processing module, wherein the processing module is used for respectively obtaining a plurality of satellite pictures of the region to be identified in the statistical period, and sequentially executing the steps of segmenting, identifying, combining and calculating each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
The evaluation module is used for scoring the region to be identified according to the area change data of the type region on the same geographical position of each satellite picture of the region to be identified in the statistical period and a preset scoring model.
Further, the acquisition module is followed by a preprocessing module for performing a preprocessing process on the satellite picture, wherein the preprocessing process includes one or more of the following methods: image binarization, interference point removal, centroid alignment and linear interpolation amplification.
The above embodiments are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby, and any insubstantial changes and substitutions made by those skilled in the art based on the present invention are within the protection scope of the present invention.
Claims (9)
1. The regional development condition evaluation method based on image recognition is characterized by comprising the following steps of:
a model establishing step: establishing a subject identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
an acquisition step: acquiring a satellite picture of a region to be identified;
a segmentation step: dividing the satellite picture into a plurality of corresponding block pictures according to color clustering;
an identification step: extracting features of each block picture to obtain a corresponding feature vector, and obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and a main body identification model;
and (3) merging steps: merging block pictures of adjacent similar subjects so as to divide the satellite picture into a plurality of type areas;
a calculation step: obtaining the geographic position of each type region of the satellite picture according to the longitude and latitude data of the map, and calculating the area size of each region type according to the area of each region type; the model building step further comprises: firstly, acquiring a plurality of satellite pictures of each type of main body; then, extracting the features of each satellite picture of each type of main body to obtain a corresponding feature vector; finally, feature vectors corresponding to a plurality of satellite pictures of each type of main body are identified and trained to obtain a set of feature vectors corresponding to the satellite pictures of each type of main body, namely a main body identification model; wherein, extracting the features of each satellite picture of each type of main body to obtain the corresponding feature vector specifically comprises: the satellite picture of the main body is divided into a plurality of grid areas in an average mode, and the ratio of the number of points in each grid area to the total number of points of the articles is calculated to obtain the multi-dimensional feature vector.
2. The method of claim 1, wherein: further comprising the processing steps of: firstly, respectively obtaining a plurality of satellite pictures of a region to be identified in a statistical period, and sequentially performing a segmentation step, an identification step, a combination step and a calculation step on each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
3. The method of claim 2, wherein: and after the plurality of satellite pictures of the area to be identified in the statistical period are obtained, the plurality of satellite pictures of the area to be identified are sequenced according to the sequence of dates.
4. The method of claim 2, wherein: further comprising the evaluation step of: and scoring the region to be identified according to the area change data of the type region on the same geographical position of each satellite picture of the region to be identified in the statistical period and a preset scoring model.
5. The method of claim 1, wherein: the acquisition step further comprises a preprocessing step: carrying out a preprocessing process on the satellite picture; wherein the pretreatment process comprises one or more of the following methods in combination: image binarization, interference point removal, centroid alignment and linear interpolation amplification.
6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein: the processor, when executing the program, performs the steps of the image recognition-based regional development status assessment method according to any of claims 1-5.
7. A computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program realizing the steps of the image recognition based regional development status assessment method according to any of claims 1-5 when being executed by a processor.
8. The regional development condition evaluation device based on image recognition is characterized by comprising:
the model establishing module is used for establishing a main body identification model; the subject recognition model stores a set of feature vectors for satellite pictures of each type of subject;
the acquisition module is used for acquiring a satellite picture of an area to be identified; the segmentation module is used for dividing the satellite picture into a plurality of corresponding block pictures according to the color clustering;
the characteristic extraction module is used for extracting the characteristics of each block picture to obtain a corresponding characteristic vector;
the identification module is used for obtaining a main body corresponding to each block picture according to the feature vector corresponding to each block picture and the main body identification model;
the merging module is used for merging the block pictures of the adjacent similar main bodies so as to divide the satellite picture into a plurality of type areas;
the calculation module is used for obtaining the geographic position of each type region of the satellite picture according to the longitude and latitude data of the map and calculating the area size of each region type according to the area of each region type; the model building step further comprises: firstly, acquiring a plurality of satellite pictures of each type of main body; then, extracting the features of each satellite picture of each type of main body to obtain a corresponding feature vector; finally, feature vectors corresponding to a plurality of satellite pictures of each type of main body are identified and trained to obtain a set of feature vectors corresponding to the satellite pictures of each type of main body, namely a main body identification model; wherein, extracting the features of each satellite picture of each type of main body to obtain the corresponding feature vector specifically comprises: the satellite picture of the main body is divided into a plurality of grid areas in an average mode, and the ratio of the number of points in each grid area to the total number of points of the articles is calculated to obtain the multi-dimensional feature vector.
9. The apparatus of claim 8, wherein: the system comprises a processing module and a processing module, wherein the processing module is used for respectively obtaining a plurality of satellite pictures of a region to be identified in a statistical period, and sequentially executing a segmentation step, an identification step, a combination step and a calculation step on each satellite picture to obtain the geographic position of each type region and the area size of each region type of each satellite picture; and then calculating the area size change data of different types of regions on the same geographical position of each satellite picture of the region to be identified in the statistical period according to the sequence of time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711216014.7A CN108052876B (en) | 2017-11-28 | 2017-11-28 | Regional development assessment method and device based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711216014.7A CN108052876B (en) | 2017-11-28 | 2017-11-28 | Regional development assessment method and device based on image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108052876A CN108052876A (en) | 2018-05-18 |
| CN108052876B true CN108052876B (en) | 2022-02-11 |
Family
ID=62120840
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711216014.7A Active CN108052876B (en) | 2017-11-28 | 2017-11-28 | Regional development assessment method and device based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108052876B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110827048B (en) * | 2018-08-08 | 2023-05-23 | 菜鸟智能物流控股有限公司 | Geographic region segmentation method and device based on road network and electronic equipment |
| CN111241881B (en) * | 2018-11-29 | 2022-12-27 | 中国移动通信集团四川有限公司 | Method, apparatus, device and medium for region identification |
| CN110398902B (en) * | 2019-06-19 | 2021-09-24 | 上海机电工程研究所 | Photoelectric signal simulation error analysis method |
| CN110290426B (en) * | 2019-06-24 | 2022-04-19 | 腾讯科技（深圳）有限公司 | Method, device and equipment for displaying resources and storage medium |
| CN111861834A (en) * | 2020-07-14 | 2020-10-30 | 自然资源部第一海洋研究所 | Coastal zone land area space usage monitoring and surveying method based on water resource dependence degree |
| CN112200012A (en) * | 2020-09-15 | 2021-01-08 | 汉海信息技术(上海)有限公司 | Map data processing method and device and electronic equipment |
| CN112613426B (en) * | 2020-12-26 | 2024-08-13 | 北京国腾联信科技有限公司 | Resource region determining method, device, equipment and storage medium |
| CN114581786B (en) * | 2021-12-28 | 2022-11-25 | 深圳市城市产业发展集团有限公司 | Method and device for estimating building area according to ground image |
| CN115619712A (en) * | 2022-09-06 | 2023-01-17 | 江苏省水利科学研究院 | Unmanned aerial vehicle aerial photography technology-based water area monitoring and evaluating method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102928A (en) * | 2014-06-26 | 2014-10-15 | 华中科技大学 | Remote sensing image classification method based on texton |
| CN104881681A (en) * | 2015-05-22 | 2015-09-02 | 浙江大学 | Image sequence category labeling method based on mixed graph model |
| CN104036507B (en) * | 2014-06-13 | 2017-02-15 | 中国地质大学（武汉） | Light satellite based urban change area and evolution type rapid extraction method |
| CN107247938A (en) * | 2017-06-08 | 2017-10-13 | 中国科学院遥感与数字地球研究所 | A kind of method of high-resolution remote sensing image City Building function classification |
| CN107392925A (en) * | 2017-08-01 | 2017-11-24 | 西安电子科技大学 | Remote sensing image terrain classification method based on super-pixel coding and convolutional neural networks |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2012318250A1 (en) * | 2011-10-19 | 2013-05-09 | The University Of Sydney | Image processing and object classification |
-
2017
- 2017-11-28 CN CN201711216014.7A patent/CN108052876B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104036507B (en) * | 2014-06-13 | 2017-02-15 | 中国地质大学（武汉） | Light satellite based urban change area and evolution type rapid extraction method |
| CN104102928A (en) * | 2014-06-26 | 2014-10-15 | 华中科技大学 | Remote sensing image classification method based on texton |
| CN104881681A (en) * | 2015-05-22 | 2015-09-02 | 浙江大学 | Image sequence category labeling method based on mixed graph model |
| CN107247938A (en) * | 2017-06-08 | 2017-10-13 | 中国科学院遥感与数字地球研究所 | A kind of method of high-resolution remote sensing image City Building function classification |
| CN107392925A (en) * | 2017-08-01 | 2017-11-24 | 西安电子科技大学 | Remote sensing image terrain classification method based on super-pixel coding and convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 土地利用视角下的南京城市空间多维扩展研究;乔伟峰;《中国博士学位论文全文数据库 经济与管理科学辑》;20170715(第7期);正文第24-33、48-65页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108052876A (en) | 2018-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052876B (en) | Regional development assessment method and device based on image recognition | |
| Voltersen et al. | Object-based land cover mapping and comprehensive feature calculation for an automated derivation of urban structure types at block level | |
| CN108764193B (en) | Urban functional area dividing method fusing POI and remote sensing image | |
| Kuang et al. | A comparative analysis of megacity expansions in China and the US: Patterns, rates and driving forces | |
| Wei et al. | Normalization of time series DMSP-OLS nighttime light images for urban growth analysis with pseudo invariant features | |
| Ren et al. | Developing a rapid method for 3-dimensional urban morphology extraction using open-source data | |
| CN110348506B (en) | Land use classification method based on remote sensing image, storage medium and computing device | |
| Weeks | Using remote sensing and geographic information systems to identify the underlying properties of urban environments | |
| Olayiwola et al. | Land use dynamics and expansion of the built-up area in Benin City, Nigeria | |
| Diep et al. | Assessment on controlling factors of urbanization possibility in a newly developing city of the Vietnamese Mekong delta using logistic regression analysis | |
| Mantas et al. | A geospatial approach to monitoring impervious surfaces in watersheds using Landsat data (the Mondego Basin, Portugal as a case study) | |
| CN113468982A (en) | Method, device and storage medium for classifying urban functional areas | |
| Zhao et al. | Human settlement evaluation in mountain areas based on remote sensing, GIS and ecological niche modeling | |
| CN115062682A (en) | Target function area identification method, device, storage medium and equipment | |
| Tian et al. | Suburban sprawl measurement and landscape analysis of cropland and ecological land: A case study of Jiangsu Province, China | |
| CN113191553A (en) | Population space distribution estimation method and system based on building scale | |
| Roychowdhury et al. | Delineating urban, suburban and rural areas using Landsat and DMSP-OLS night-time images | |
| Liu et al. | Automated extraction of urban roadside trees from mobile laser scanning point clouds based on a voxel growing method | |
| Nitze et al. | A labeling intercomparison of retrogressive thaw slumps by a diverse group of domain experts | |
| Niu | Estimating housing vacancy rate in Qingdao city with NPP-VIIRS nighttime light and geographical national conditions monitoring data | |
| Fafchamps et al. | The evolution of built-up areas in Ghana since 1975 | |
| Eniolorunda et al. | Assessment ofurban growth pattern in sokoto metroplis, nigeria using multitemporal sateelite data | |
| Uchida et al. | Socio-economic impact of Trans-Siberian railway after the collapse of Soviet Union by integrated spatial data analysis | |
| DEMİR et al. | Analysis Temporal Land Use/Land Cover Change Based on Landscape Pattern and Dynamic Metrics in Protected Mary Valley, Trabzon from 1987 to 2015 | |
| CN117893923B (en) | Method for constructing space scene based on blue-green gray scale |
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 |