CN117351368B - Natural village boundary acquisition method and device, electronic equipment and storage medium - Google Patents
Natural village boundary acquisition method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN117351368B CN117351368B CN202311657980.8A CN202311657980A CN117351368B CN 117351368 B CN117351368 B CN 117351368B CN 202311657980 A CN202311657980 A CN 202311657980A CN 117351368 B CN117351368 B CN 117351368B
- Authority
- CN
- China
- Prior art keywords
- concave
- target
- house
- natural village
- village
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000003860 storage Methods 0.000 title claims abstract description 10
- 238000013136 deep learning model Methods 0.000 claims abstract description 36
- 238000012549 training Methods 0.000 claims description 59
- 238000012545 processing Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 10
- 238000005520 cutting process Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000010865 sewage Substances 0.000 abstract description 13
- 238000009826 distribution Methods 0.000 abstract description 8
- 238000013461 design Methods 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000013440 design planning Methods 0.000 abstract description 3
- 239000013256 coordination polymer Substances 0.000 description 15
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 5
- 238000013439 planning Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 4
- 238000002372 labelling Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 230000001746 atrial effect Effects 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012732 spatial analysis Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
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
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
Abstract
The invention is suitable for a blow-down pipe network design technology, and provides a natural village boundary acquisition method, a natural village boundary acquisition device, electronic equipment and a storage medium. The method comprises the steps of obtaining a target remote sensing image containing a target administrative village region; extracting each house surface position in the target remote sensing image by adopting a pre-trained deep learning model; generating Thiessen polygons according to pre-marked natural village POIs; clustering according to the positions of all house surfaces at preset distances to generate concave bags; traversing each concave bag one by one, judging the position relation between the target concave bag and the Thiessen polygon and the distribution condition of POIs according to each traversed target concave bag, setting the natural village attribution attribute of the house face position, and solving the concave bags of the roof positions belonging to the same natural village by utilizing the natural village attribute of the classified house face position to obtain the range of each natural village so as to obtain more accurate natural village boundary suitable for sewage pipe network design planning.
Description
Technical Field
The invention belongs to the technical field of design of sewage pipes and networks, and particularly relates to a method and a device for acquiring natural village boundaries, electronic equipment and a storage medium.
Background
When the sewage pipe network is designed for the administrative village, the design is required to be based on a single natural village in order to improve the design accuracy. Thus, it is necessary to extract individual natural village boundaries.
However, the government in the middle gate place does not have natural village boundary data, or the boundary contour obtained by directly dividing the remote sensing image in the prior art is too simplified, so that the requirement of fine design cannot be met. The natural village deviation is large, so that the laying cost of the drainage pipe network is increased sharply. In addition, the processing cost of manually dividing the remote sensing image is high, and large-scale automatic processing cannot be performed.
Disclosure of Invention
The embodiment of the invention provides a natural village boundary acquisition method, which aims at solving the problem of how to acquire a natural village boundary which is accurate and suitable for sewage pipe network planning.
The embodiment of the invention is realized in such a way that the acquisition method of the natural village boundary comprises the following steps:
acquiring a target remote sensing image containing a target administrative village region;
extracting each house surface position in the target remote sensing image by adopting a pre-trained deep learning model;
generating Thiessen polygons according to pre-marked natural village POIs;
Clustering according to the positions of the house surfaces and preset distances to generate concave bags;
traversing each concave bag one by one, judging whether the target concave bag is intersected with more than one Thiessen polygon according to each traversed target concave bag, if not, setting the natural village attribution attribute of the house surface position in the target concave bag as a natural village corresponding to the POI, and marking the target concave bag as processed;
if the target concave bag is intersected with more than one Thiessen polygon, judging whether the target concave bag contains more than 1 POI, if not, setting the natural village attribution attribute of the house surface position in the target concave bag as the natural village corresponding to the POI of the Thiessen polygon with the largest area occupied by the target concave bag;
if the target concave packet contains more than 1 POI, cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas, wherein N is an integer greater than 1;
judging whether each target ladle sub-region in the N ladle sub-regions contains POIs one by one, if so, setting the natural village attribution attribute of the house roof position in the target ladle sub-region as the natural village corresponding to the POIs, and if not, merging the house surface to the processed natural village nearest to the target ladle sub-region;
After traversing each concave bag one by one, the concave bags are obtained by utilizing the natural village attribute of the classified house surface positions and the roofing positions belonging to the same natural village, so as to obtain the range of each natural village.
The embodiment of the invention also provides a device for acquiring the natural village boundary, which comprises the following steps:
the target remote sensing image acquisition module is used for acquiring a target remote sensing image containing a target administrative village region;
the house face position acquisition module is used for extracting each house face position in the target remote sensing image by adopting a pre-trained deep learning model;
the Thiessen polygon generation module is used for generating Thiessen polygons according to the pre-marked natural village POIs;
the first concave packet generation module is used for generating concave packets in a clustering mode according to the positions of the house surfaces and preset distances;
the first concave packet traversing module is used for traversing each concave packet one by one, judging whether the target concave packet is intersected with more than one Thiessen polygon according to each traversed target concave packet, if not, setting the natural village attribution attribute of the house surface position in the target concave packet as a natural village corresponding to the POI, and marking the target concave packet as processed;
the first concave packet traversing module is further configured to determine whether the target concave packet contains more than 1 POI if the target concave packet intersects more than one taylon polygon, and if not, set a natural village attribution attribute of a house surface position in the target concave packet as a natural village corresponding to the POI to which the taylon polygon with the largest area occupied by the target concave packet belongs;
The first concave packet traversing module is further used for cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas if the target concave packet contains more than 1 POI, wherein N is an integer greater than 1;
the first concave packet traversing module is further used for judging whether each target concave packet sub-area in the N concave packet sub-areas contains POIs one by one, if so, setting the natural village attribution attribute of the house roof position in the target concave packet sub-area as the natural village corresponding to the POIs, and if not, merging the house surface to the processed natural village nearest to the target concave packet sub-area;
and the second concave packet generation module is used for solving concave packets of roof positions belonging to the same natural village by utilizing the natural village attribute of the classified house surface positions after traversing each concave packet one by one, so as to obtain the range of each natural village.
The embodiment of the invention also provides electronic equipment, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the method for acquiring the natural village boundary when executing the computer program.
Embodiments of the present invention also provide a computer-readable storage medium storing a computer program which, when executed by a processor, implements a method of acquiring natural village boundaries as described above.
According to the method, on one hand, the home relation between the house roof and the natural village in each concave pocket is adjusted through the concave pocket formed by clustering the house surface positions, the position relation of Thiessen polygons generated according to the pre-marked natural village POIs and the distribution condition of the POIs. Specifically, by traversing each concave packet one by one, judging whether the target concave packet is intersected with more than one Thiessen polygon for each traversed target concave packet, if not, classifying the house surface position in the target concave packet into a natural village corresponding to the POI, namely setting the natural village attribution attribute of the house surface position in the target concave packet as the natural village corresponding to the POI, and marking the target concave packet as processed; if the target concave bag is intersected with more than one Thiessen polygon, judging whether the target concave bag contains more than 1 POI, if not, setting the home attribute of the house surface position natural village in the target concave bag as the natural village corresponding to the POI of which the target concave bag area occupies a relatively large Thiessen polygon; if the target concave packet contains more than 1 POI, cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas, wherein N is an integer greater than 1; judging whether each target concave sub-area in the N concave sub-areas contains POIs one by one, if so, classifying the house roof position in the target concave sub-areas into a natural village corresponding to the POIs, namely, setting the natural village attribution attribute of the house face position as the natural village corresponding to the POIs, if not, merging the house face into the nearest processed natural village of the target concave sub-areas, adjusting the concave-bag boundary generated by clustering, and repartitioning the house face possibly not belonging to the natural village or far from the natural village to obtain the natural village boundary more suitable for the design planning of the sewage pipe network, so that the construction accuracy of facilities such as the sewage pipe network based on the natural village boundary planning is greatly improved, and the construction cost of facilities such as the sewage pipe network based on the natural village boundary planning is greatly reduced. On the other hand, the method and the device extract each house surface position in the target remote sensing image by adopting the pre-trained deep learning model, further form concave bags according to house surface position clustering, automatically extract natural village boundaries according to the position relation of Thiessen polygons generated by pre-marked natural village POIs and the distribution condition of the POIs, solve the problem of high processing cost of manually segmenting the remote sensing image, and improve the efficiency of natural village boundary extraction.
Drawings
Fig. 1 is a flow chart of a method for acquiring a natural village boundary according to an embodiment of the present application;
FIG. 2 is a schematic diagram of generating Thiessen polygons from pre-labeled natural village POIs provided in an embodiment of the present application;
FIG. 3 is a schematic diagram of concave bags generated by clustering at preset distances according to each house roof provided by the embodiment of the application;
fig. 4 is a schematic diagram of a process of traversing each concave packet one by one according to the method for acquiring a natural village boundary provided in the embodiment of the present application;
FIG. 5 is a schematic diagram of the intersection of a Thiessen polygon and a concave bag provided by an embodiment of the present application;
FIG. 6 is a schematic diagram of the intersection of a concave packet with a Thiessen polygon provided by an embodiment of the present application;
FIG. 7 is a schematic diagram of the intersection of a concave packet with two Thiessen polygons provided by an embodiment of the present application;
FIG. 8 is a schematic illustration of a concave bag provided by an embodiment of the present application intersecting three Thiessen polygons and containing three POIs;
FIG. 9 is a schematic view of the extent of each final natural village provided by embodiments of the present application;
FIG. 10 is a flowchart of a method for acquiring natural village boundaries according to another embodiment of the present invention;
FIG. 11 is a schematic diagram of a training remote sensing image of a training area provided in an embodiment of the present application;
FIG. 12 is a schematic diagram of a training remote sensing image labeled house surface provided in an embodiment of the present application;
Fig. 13 is a schematic view of slicing a training remote sensing image according to a preset size according to an embodiment of the present application;
FIG. 14 is a schematic view of a roofing annotation mask slice provided by embodiments of the present application;
FIG. 15 is a flowchart of a method for acquiring natural village boundaries according to another embodiment of the present invention;
fig. 16 is a schematic diagram of an acquiring device for a natural village boundary according to an embodiment of the present application;
fig. 17 is a schematic diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
To facilitate an understanding of the present application, terms referred to herein are explained below.
One of the administrative villages, the administrative division of the chinese-style base, typically includes a plurality of natural villages. As a broad area to be analyzed in this application.
Natural villages, which refers to a range of naturally occurring villages, typically contain more concentrated points of residence, are intended to extract boundaries of natural villages of unknown distribution.
GIS (Geographic Information System ), computer system for geographic data management, analysis, visualization. The application manages and analyzes natural village data in combination with a GIS database.
GDAL (Geospatial Data Abstraction Library, meaning a geospatial data abstraction library) is an open source geospatial data processing tool set.
Shapefile, a commonly used GIS vector data format, is used to store geometric and attribute information of geographic elements.
Post GIS is the expansion of the spatial database of Post greSQL, and realizes the functions of spatial data management and spatial analysis in the database.
POIs (Point of Interest, points of interest), typically represent locations of geographical elements of a certain significance, in this application place name location points of natural villages.
VP (Voronoi Polygons, thiessen Polygons), a spatially segmented structure, generates Polygons representing a region's extent from a set of points. The method is used for representing the spatial influence range of the natural village POI. In the application, a GIS built-in algorithm is adopted to solve the Thiessen polygon.
It will be appreciated that the straight lines of the dashed lines (- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -is) is a feature of a Thiessen polygon, and that the lines of the repeated dashed lines are shown as being nearly continuous lines, but are recognized by those skilled in the art as being part of a Thiessen polygon contour.
CP (Concave Hull Polygon, concave-convex polygon), also known as concave-convex, is a concave boundary that better represents the shape of a set of points. The boundaries of the pockets may be non-convex, including concave portions, and are typically established by the distribution density of dots, forming the pockets in the high density areas. In the application, a GIS built-in algorithm is adopted to solve the concave bag.
Example 1
Referring to fig. 1, an embodiment of the present application provides a method for acquiring a natural village boundary, which may be applied to an electronic device provided by the implementation of the present application, where the method for acquiring a natural village boundary includes steps S110 to S160, and specifically includes the following steps:
s110, acquiring a target remote sensing image containing a target administrative village region;
s120, extracting the position of each house surface in the target remote sensing image by adopting a pre-trained deep learning model;
s130, generating Thiessen polygons according to pre-marked natural village POIs;
s140, clustering according to the positions of the house surfaces to generate concave bags at preset distances;
s150, traversing each concave packet one by one,
s151, judging whether each traversed target concave packet is intersected with more than one Thiessen polygon, if not, setting the natural village attribution attribute of the house surface position in the target concave packet as a natural village corresponding to the POI, and marking the target concave packet as processed;
S152, if the target concave bag is intersected with more than one Thiessen polygon, judging whether the target concave bag contains more than 1 POI, if not, setting the natural village attribution attribute of the house surface position in the target concave bag as the natural village corresponding to the POI of the Thiessen polygon with the largest area occupied by the target concave bag;
s153, if the target concave packet contains more than 1 POI, cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas, wherein N is an integer greater than 1;
s154, judging whether each target concave-bag sub-area in N concave-bag sub-areas contains POI one by one, if so, setting the natural village attribution attribute of the house roof position in the target concave-bag sub-area as the natural village corresponding to the POI, and if not, merging the house surface to the processed natural village nearest to the target concave-bag sub-area;
s160, after traversing each concave bag one by one, the concave bags are obtained by utilizing the natural village attribute of the classified house surface positions and the roof positions belonging to the same natural village, so that the range of each natural village is obtained.
In some embodiments, acquiring the target remote sensing image including the target administrative village region may be based on administrative village information input by the user, such as administrative village name, and the like, and acquiring the remote sensing image including the target administrative village from the map provider server.
In some embodiments, in order to solve the problem of high processing cost of manually performing segmentation processing on the remote sensing image and improve the efficiency of natural village boundary extraction, a pre-trained deep learning model may be used to extract each house surface position in the target remote sensing image. For example, multiple administrative village remote sensing images marking the roof of the house may be used to train the deep learning model. In some specific examples, the deep learning model can be a U-Net neural network model.
In some embodiments, each house face position in the target remote sensing image is extracted by using a deep learning model, a binary image is generated after the house face position of the remote sensing image is identified, and the house face position is determined according to the binary image. Here, the house surface position refers to position information, and the position information may include not only latitude and longitude information of a house surface, but also relative position information of the house surface in a map, and may be selected according to requirements when implementing the present application.
In some embodiments, referring to fig. 2, a Thiessen polygon is generated from pre-labeled natural village POIs. In particular, the input data may be a set of natural village POI points within the administrative village, the POI points geographically representing the locations of the respective natural villages. Using st_voronoi polygons () function of PostGIS, computation of a tesen polygon is performed with these POI points as generation points. The function first calculates the distance of each generated point from its surrounding points. Then, a polygon is constructed with the perpendicular line between the two points as sides, and each polygon contains one generation point. And finally outputting Thiessen polygons corresponding to all the natural village POI generating points in the administrative village range. Each Thiessen polygon represents the area and extent of influence of one natural village POI. The POI data can be obtained from an electronic map or can be plotted manually. The text position in fig. 2 illustrates that the POI is marked here, and the content of the specific marked place name is irrelevant to the application. The straight line bar constituted by the dash-dot line (- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -, is a Thiessen polygonal contour). The curve formed by the dashed line (- - - -) is the administrative outline.
In some embodiments, referring to fig. 3, the pockets are clustered at a preset distance from each of the house face locations. Specifically, the generating a concave packet by clustering according to the house surface positions and a preset distance includes: calculating the distance between every two house surface positions; classifying a house face position to be classified, which is less than a preset distance from a reference house face position, and the reference house face position into the same category, and setting the natural village attribution attribute of the house face position to be classified into the same natural village as the reference house face position, wherein the reference house face position is the house face position to which the natural village attribution attribute is assigned; and solving the concave bag for the house surface position of which the home attribute of the natural village is the same natural village.
The distance calculation and the concave packet analysis are realized by calling a space analysis function of the geographic information system database, which belongs to the prior art and is not repeated. In some specific examples, the preset distance used for clustering may specifically be selected to be 30 meters.
The curve formed by the dashed lines (- - - - -) in FIG. 3 is an administrative village outline within which the highlighted polygonal area is a recess created by clustering the house roof.
In some embodiments, referring to fig. 1, step S150 generally embodies the inventive concept of traversing each pocket one by one, clustering the pockets according to house surface positions, generating the positional relationship of the Thiessen polygons according to pre-labeled natural village POIs, and adjusting the home relationship of the house roof and the natural village in each pocket according to the distribution of POIs.
Referring to fig. 4, in a specific example, in step S150, it may be determined whether all CPs are traversed using a loop variable i, whether CP sub-areas are traversed using a loop variable j, and the selection of target CPs one by one and the selection of target CP sub-areas one by one may be implemented by gradually increasing i and j. It can be understood that all CPs are traversed, and the currently traversed CPs are target CPs; traversing the CP subregions of all CPs, wherein the CP subregions currently traversed are target CP subregions. How to implement the traversal and how to implement the traversal scheme equivalent to the embodiment of the present application, which belongs to the specific implementation manner that can be understood by those skilled in the art, and will not be described in detail.
In fig. 4, after generating the concave packets according to the positions of the house surfaces by clustering with preset distances, traversing each concave packet, selecting one concave packet as the current processed target concave packet for judgment, and selecting the next concave packet as the current processed target concave packet for judgment after the current target concave packet is processed until all concave packets are processed.
As shown in FIG. 5, the Thiessen polygon intersects the concave bag, which is cut. As in fig. 5, one concave packet, which is shown as a highlighted polygonal area, is cut into 5 areas CP1 to CP5 by a plurality of taisen polygons. At this time, it is necessary to determine whether the target pocket intersects with a plurality of Thiessen polygons.
As shown in fig. 6, if a recess intersects only one tesson polygon, the house surface position contained in the recess is determined as the natural village corresponding to the POI belonging to the tesson polygon.
As shown in fig. 7, if the dip intersects a plurality of taisen polygons, it is determined whether the dip contains a plurality of natural village POIs. If the concave bag only contains one POI, judging that the house surface in the concave bag belongs to a natural village corresponding to the POI.
As shown in fig. 8, if the concave bag contains a plurality of POI points, the concave bag is cut by using a tessellation polygon, so as to obtain a plurality of concave steamed stuffed bun areas. Specifically, the st_intersection function in PostGIS may be used to segment CPs that span multiple VPs and contain multiple POIs. In PostGIS, the st_interaction function is used to calculate the Intersection of two geometric objects. Here, intersection refers to a portion in which two geometric objects coexist.
And traversing each concave bag sub-area of each target concave bag, and judging the house surface contained in the concave bag sub-area to belong to the natural village corresponding to the POI if the POI point is contained in the concave bag sub-area. For a concave steamed stuffed bun area not containing POIs, the house surface is attributed to the nearest processed natural village. Specifically, the Distance solution is performed through an st_distance function in the PostGIS, ascending order is performed according to the Distance, and the first natural village is taken as the nearest natural village.
The steps of selecting the target CP in fig. 4 until determining whether all CPs are traversed are completed are repeated until all the pockets and their pocket sub-areas are processed. And finally, judging that all house surfaces belong to the unique natural villages according to the rules, and finishing the natural village division of the whole area.
As shown in fig. 9, the natural villages on the roof of the reassigned house are notched, and finally, the straight line bar formed by the range of each natural village (the dash-dot line in the figure (-s) -is a Thiessen polygonal outline, the curve formed by the broken line (-s) -is an administrative village outline, and the area of the highlighted polygon in the administrative village outline is the range of the reassigned house face natural village). The process shown in fig. 4 fully considers the requirement that the boundary of the natural village is defined as compactly as possible when designing the sewage pipe network, and ensures that the treatment result of the home position belonging to the natural village meets the design requirement of the sewage pipe network, thereby being the core step for realizing the technical scheme of the application. It will be appreciated that the straight lines of the dashed lines (- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -is) is a feature of a Thiessen polygon, and that the lines of the repeated dashed lines are shown as being nearly continuous lines, but are recognized by those skilled in the art as being part of a Thiessen polygon contour.
In the application, on one hand, the home relation between the house roof and the natural village in each concave pocket is adjusted through the concave pocket formed by clustering the house surface positions, the position relation of Thiessen polygons generated according to the pre-marked natural village POIs and the distribution condition of the POIs. Specifically, by traversing each concave packet one by one, judging whether the target concave packet is intersected with more than one Thiessen polygon for each traversed target concave packet, if not, classifying the house surface position in the target concave packet into a natural village corresponding to the POI, namely setting the natural village attribution attribute of the house surface position in the target concave packet as the natural village corresponding to the POI, and marking the target concave packet as processed; if the target concave bag is intersected with more than one Thiessen polygon, judging whether the target concave bag contains more than 1 POI, if not, setting the home attribute of the house surface position natural village in the target concave bag as the natural village corresponding to the POI of which the target concave bag area occupies a relatively large Thiessen polygon; if the target concave packet contains more than 1 POI, cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas, wherein N is an integer greater than 1; judging whether each target ladle sub-area in the N ladle sub-areas contains POIs one by one, if so, classifying the house roof positions in the target ladle sub-areas into natural villages corresponding to the POIs, namely, setting the natural village attribution attribute of the house face positions as the natural villages corresponding to the POIs, if not, merging the house faces into the nearest processed natural villages of the target ladle sub-areas, adjusting the clustered house face positions, and re-dividing the house faces possibly not belonging to the natural villages or far from the natural villages to obtain natural village boundaries more suitable for sewage pipe network design planning, so that the construction cost of facilities such as sewage pipe networks and the like planned based on the natural village boundaries is greatly reduced. On the other hand, the method and the device extract each house surface position in the target remote sensing image by adopting the pre-trained deep learning model, further the concave packet formed by clustering the house surface positions and the position relation of generating the Thiessen polygons according to the pre-marked natural village POIs automatically extract the natural village boundary, solve the problem of high processing cost of manually segmenting the remote sensing image, and improve the efficiency of extracting the natural village boundary.
Example two
Referring to fig. 10, in some embodiments of the present application, before the step of acquiring the target remote sensing image including the target administrative village region in step S110, steps S001 to S003 are further included, specifically as follows:
s001, acquiring a training remote sensing image of a training area, wherein the training remote sensing image is marked with a house roof;
s002, slicing the training remote sensing image and the house surface label according to preset dimensions to obtain a slice training data set;
and S003, training the deep learning model to be trained by using the slice training data set to obtain a pre-trained deep learning model.
In some embodiments, a U-Net neural network model can be employed as the deep learning model.
In some embodiments, as shown in FIG. 11, a training remote sensing image of a training area is acquired, the training remote sensing image having been labeled with a house roof. Specifically, a remote sensing image for training the deep learning model is obtained through a remote sensing image data service provider, and the training area can be a selected administrative village area for training. As shown in FIG. 12, these training remote sensing images for training the deep learning model may be used to label the house surface by way of manual labeling (highlighting the labeling location in FIG. 12).
In some embodiments, as shown in fig. 13, the training remote sensing image and the house surface label are sliced according to a preset size to obtain a slice training data set. Specifically, a suitable slice size may be preset according to the input size requirement of the deep learning model, such as 256×256 (units are pixels). Each training image is rasterized at the size and a certain overlap ratio, for example, 50%, and the training image includes the remote sensing image after slicing and the labeling mask after slicing. After cutting, each slice image contains a local area of the original large image, and the corresponding relation between the labeling mask and the remote sensing image is reserved. And repeating the cutting on all the training images to finally obtain a large number of slice images containing the marked house roof.
In some embodiments, the deep learning model to be trained is trained using the slice training data set shown in fig. 13 and 14, resulting in a pre-trained deep learning model.
It can be appreciated that the adoption of slice images to form a new training set can efficiently train the deep learning model, and avoid the problem of insufficient memory of the processor. The slice training data is also more beneficial to the model to learn house surface characteristics at different positions and different proportions. The model has stronger and comprehensive house surface recognition capability under various conditions through the enhancement of slice training data. The training samples are obtained by adopting a slicing technology, so that support is provided for improving the generalization performance of the model and processing large scene data.
Example III
Referring to fig. 15, in some embodiments of the present application, step S120 extracts each house surface position in the target remote sensing image by using a pre-trained deep learning model, including steps S121 and S122, specifically as follows:
s121, carrying out sliding window processing in the target remote sensing image by adopting a preset path, and extracting the position of a house roof in each sliding window;
and S122, integrating the house surface positions in all sliding windows to obtain each house surface position in the target remote sensing image.
In some embodiments, the house face extraction is performed on the large-size target remote sensing image required for the house face extraction by using the deep learning model with the recognition capability on the house face, which is obtained through the previous training. A suitable sliding window size, such as 256×256 (in pixels), is preset according to the image size and the model input size.
For example, the sliding window process may be performed in a zigzag path from the top left side of the image to the bottom right side of the image at a predetermined overlapping rate, for example, 50%. Each sliding window image is input into a pre-trained deep learning model that predicts the room roof within the window. The deep learning model predicts all sliding windows in sequence to obtain house surface prediction results of each window. And according to the overlapping area of the sliding windows, splicing and fusing the results of different sliding windows to form the prediction of the whole graph. And finally, summarizing and integrating all sliding windows to obtain all house surface prediction results in the whole target remote sensing image.
It can be appreciated that, because the slice training data and the sliding window prediction technology are adopted in the method, on one hand, the sliding window only loads the local area of the image, so that the occupation of a processor and a memory is greatly reduced, and the method is more economical and efficient than loading the whole image. This allows the limited computational resources to process large-scale remote sensing images as well. Obviously, the characteristic of small resource occupation is very suitable for processing large-area administrative villages. On the other hand, the sliding window with fixed size can uniformly convert the target remote sensing image prediction problems with different sizes into the image prediction problems with the same size. This simplifies the model design without having to adjust the model separately for different sizes. The sliding window technique can be well adapted to this complex situation, considering the large differences in area between different administrative villages. Therefore, the slice training data and the sliding window prediction technology are adopted simultaneously, so that the occupation of computing resources is reduced, the adaptability of the model is improved, and the efficient processing of administrative villages with different scales is supported.
Example IV
In some embodiments of the present application, step S140, generating a concave packet by clustering with a preset distance according to each house surface position, includes the following steps:
clustering to generate concave bags according to the positions of the house surfaces at preset distances, wherein the concave bags comprise: calculating the distance between every two house surface positions; classifying a house face position to be classified, which is less than a preset distance from a reference house face position, and the reference house face position into the same category, and setting the natural village attribution attribute of the house face position to be classified into the same natural village as the reference house face position, wherein the reference house face position is the house face position to which the natural village attribution attribute is assigned; and solving the concave bag for the house surface position of which the home attribute of the natural village is the same natural village.
The distance calculation and the concave packet analysis are realized by calling a space analysis function of a geographic information system database.
In a specific embodiment, the binary map generated by the pre-trained deep learning model sliding window prediction is converted into a house surface vector shape using the gdal. The GDAL, polygonize () method is a function in the GDAL library for vectorizing raster data. The main function of the method is to convert the raster image into vector Polygon (Polygon) data. And (3) warehousing house surface data in the previous step into a PostGIS to obtain a data table resa.
The step of clustering the generated concave packets at preset distances according to the house surface positions can be realized by adopting the processing mode described by the following pseudo codes.
Input: administrative village boundary, house face data, cluster distance
resa contains the gid and geom fields.
resa table structure:
a temporary table tmp is created from the resa house facing table, and also includes gid and geom fields, and dmn and chk fields are added, wherein dmn default is NULL (NULL value) and chk default is FALSE (FALSE value).
tmp table structure:
and performing data operation based on the PostGIS.
dmn_number is initialized to 1.
And (3) carrying out updating operation on the tmp table, setting the dmn field of the record with the smallest gid as 1, and setting the chk field as FALSE.
The outer loop is started and the record outer of dmn=dmn_number and chk is FALSE is queried.
And starting the inner layer circulation, and inquiring all record innr with dmn being empty in tmp.
Judging the distance between the outer and the inr, if the distance is smaller than the preset clustering distance, setting the dmn field of the inr as the current dmn_number, and setting the chk field as false.
After the inner layer cycle is completed, the chk field of the outer layer record outr is modified to TRUE.
This outer layer cycle is ended.
Querying records in tmp table with dmn=dmn_number and chk=false, and if there is no query result, jumping out of the loop.
Inquiring dmn empty data in the tmp table, and if an inquiring result exists, executing 1 adding operation on the dmn_number; and searching a record with dmn being NULL and gid being minimum in the tmp table, updating dmn of the record to the latest dmn_number, and updating the chk field to FALSE.
Wherein,
and a resa data table, wherein the database table is created in the PostGIS and is used for storing imported house surface data.
gid, house face unique identification ID, a primary key field in a database for uniquely identifying each house face record.
geom, representing house surface geographical location information, is a space geometrical object, and may use point coordinates, line coordinates or surface coordinates to represent the position shape of the house roof.
dmn, (integer ), natural village number, for marking the natural village to which each house face belongs.
chk (bool, boolean) marks whether the house face has been judged, avoiding repeated judgment.
innr represents the atrial roof to be determined and outer represents the nuclear atrial roof of the current natural village. And calculating the distance between the innr and the outer, if the distance is smaller than the preset clustering distance, judging the innr to belong to the current natural village, setting dmn as the number of the current natural village, and adding the number into the range of the natural village. So inr and outr play an extremely important role in the algorithm, inr ensures that each house face can be judged whether to belong to a certain natural village or not, and outr is taken as a core of the natural village and is compared with the inr distance to determine the range of the natural village. The cooperation of inr and outr facilitates iterative expansion of the natural village range until all house roofs are determined to belong to a certain natural village.
It can be understood that the concave bag is generated by clustering with the preset distance threshold, on one hand, the house surface distribution mode in the natural village can be more accurately depicted, and the spatial relevance between house roofs is reflected by distance clustering. Compared with the method of directly dividing the boundary from the remote sensing image, the method can more finely describe the aggregation range of the house roof. On the other hand, the granularity of the clustering can be flexibly controlled by adjusting the distance threshold parameter, so that the multi-granularity house surface clustering is realized. On the other hand, the clustering mode is adopted to replace manual interpretation, thereby realizing the intelligence and automation of the natural village range judgment, and being suitable for the analysis of large-scale remote sensing images.
It should be noted that, in general, those skilled in the art will prefer to use the kmeans algorithm or an algorithm similar thereto for clustering because the number and location of POIs are already determined. However, the kmeans-like algorithm is not applicable in the application, and because the technical problem to be solved in the application is how to obtain the accurate natural village boundary suitable for sewage pipe network planning, when clustering, not only the problem of clustering self performance, but also the problem of whether suitable for sewage pipe network planning is needed to be considered, and in consideration of the problem, the roof position belonging to the same natural village obtained by adopting the distance clustering-based algorithm can be more compact. The distance clustering-based algorithm can be closely matched with the subsequent adjustment step, so that a large number of roof position attribution adjustment or excessive wrong roof position attribution adjustment cannot occur, and the accuracy of the natural village boundary problem is improved.
Example five
Corresponding to the method for acquiring the natural village boundary shown in fig. 1, fig. 16 shows an apparatus M100 for acquiring the natural village boundary according to the embodiment of the invention, which includes:
the target remote sensing image acquisition module M110 is used for acquiring a target remote sensing image containing a target administrative village region;
The house face position acquisition module M120 is used for extracting each house face position in the target remote sensing image by adopting a pre-trained deep learning model;
a Thiessen polygon generation module M130, configured to generate Thiessen polygons according to pre-labeled natural village POIs;
the first concave packet generating module M140 is configured to generate concave packets by clustering according to the positions of the house surfaces and a preset distance;
a first concave packet traversing module M150, configured to traverse each concave packet one by one, determine, for each traversed target concave packet, whether the target concave packet intersects more than one tessellated polygon, and if not, set a natural village attribution attribute of a house surface position in the target concave packet as a natural village corresponding to a POI, where the target concave packet is marked as processed;
the first concave packet traversing module M150 is further configured to determine whether the target concave packet contains more than 1 POI if the target concave packet intersects more than one taylon polygon, and if not, set a natural village attribution attribute of a house surface position in the target concave packet as a natural village corresponding to the POI to which the taylon polygon with the largest area occupied by the target concave packet belongs;
the first concave packet traversing module M150 is further configured to cut the target concave packet by using a Thiessen polygon to generate N corresponding concave steamed stuffed bun regions if the target concave packet contains more than 1 POI, where N is an integer greater than 1;
The first concave packet traversing module M150 is further configured to determine whether each target concave packet sub-area in the N concave packet sub-areas includes a POI one by one, if yes, set a natural village attribution attribute of a house roof position in the target concave packet sub-area as a natural village corresponding to the POI, and if not, merge a house face into a processed natural village nearest to the target concave packet sub-area;
and the second concave packet generating module M160 is used for obtaining concave packets from the roofing positions belonging to the same natural village by utilizing the natural village attribute of the classified house surface positions after traversing each concave packet one by one, so as to obtain the range of each natural village.
Still further, the natural village boundary acquiring device further includes:
the training remote sensing image acquisition module is used for acquiring a training remote sensing image of a training area, wherein the training remote sensing image is marked with a house roof;
the training data set slicing module is used for slicing the training remote sensing image and the house surface label according to a preset size to obtain a slicing training data set;
and the deep learning model training module is used for training the deep learning model to be trained by using the slice training data set to obtain a pre-trained deep learning model.
Still further, the house face position acquisition module includes:
the sliding window processing module is used for carrying out sliding window processing in the target remote sensing image by adopting a preset path and extracting the position of a house roof in each sliding window;
and the house face position integrating module is used for integrating the house face positions in all sliding windows to obtain each house face position in the target remote sensing image.
Still further, the first concave packet generation module includes:
the roof position interval calculation module is used for calculating the distance between every two house surface positions;
the roof position clustering module is used for classifying the house face position to be classified, which is smaller than the preset distance, from the reference house face position into the same category, and setting the natural village attribution attribute of the house face position to be classified into the same natural village as the reference house face position, wherein the reference house face position is the house face position to which the natural village attribution attribute is assigned;
the first concave packet solving module is used for solving concave packets for house surface positions of which the attribute of the natural village is the same natural village.
It will be appreciated that various implementations and combinations of implementations and advantageous effects thereof in the above embodiments are equally applicable to this embodiment, and will not be described here again.
Example six
Fig. 17 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. As shown in fig. 17, the electronic device D10 of this embodiment includes: at least one processor D100 (only one is shown in fig. 17), a memory D101 and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, the processor D100 implementing the steps in any of the various method embodiments described above when executing the computer program D102.
The electronic device D10 may be a computing device such as a desktop computer, a notebook computer, a palm computer, a cloud server, etc. The electronic device may include, but is not limited to, a processor D100, a memory D101. It will be appreciated by those skilled in the art that fig. 17 is merely an example of the electronic device D10 and is not meant to be limiting of the electronic device D10, and may include more or fewer components than shown, or may combine certain components, or different components, such as may also include input-output devices, network access devices, etc.
The processor D100 may be a central processing unit (Central Processing Unit, CPU), the processor D100 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory D101 may in some embodiments be an internal storage unit of the electronic device D10, such as a hard disk or a memory of the electronic device D10. The memory D101 may also be an external storage device of the electronic device D10 in other embodiments, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the electronic device D10. Further, the memory D101 may also include both an internal storage unit and an external storage device of the electronic device D10. The memory D101 is used for storing an operating system, an application program, a boot loader (BootLoader), data, other programs, etc., such as program codes of the computer program. The memory D101 may also be used to temporarily store data that has been output or is to be output.
It is understood that those skilled in the art can combine the various embodiments of the above embodiments to obtain technical solutions of the various embodiments under the teachings of the above embodiments.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (10)
1. A method for acquiring natural village boundaries, comprising the steps of:
acquiring a target remote sensing image containing a target administrative village region;
extracting each house surface position in the target remote sensing image by adopting a pre-trained deep learning model;
generating Thiessen polygons according to pre-marked natural village POIs;
clustering according to the positions of the house surfaces and preset distances to generate concave bags;
traversing each concave bag one by one, judging whether the target concave bag is intersected with more than one Thiessen polygon according to each traversed target concave bag, if not, setting the natural village attribution attribute of the house surface position in the target concave bag as a natural village corresponding to the POI, and marking the target concave bag as processed;
if the target concave bag is intersected with more than one Thiessen polygon, judging whether the target concave bag contains more than 1 POI, if not, setting the natural village attribution attribute of the house surface position in the target concave bag as the natural village corresponding to the POI of the Thiessen polygon with the largest area occupied by the target concave bag;
if the target concave packet contains more than 1 POI, cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas, wherein N is an integer greater than 1;
Judging whether each target ladle sub-region in the N ladle sub-regions contains POIs one by one, if so, setting the natural village attribution attribute of the house roof position in the target ladle sub-region as the natural village corresponding to the POIs, and if not, merging the house surface to the processed natural village nearest to the target ladle sub-region;
after traversing each concave bag one by one, the concave bags are obtained by utilizing the natural village attribute of the classified house surface positions and the roofing positions belonging to the same natural village, so as to obtain the range of each natural village.
2. The method for acquiring natural village boundaries as defined in claim 1, wherein before the step of acquiring the target remote sensing image comprising the target administrative village region, further comprising:
acquiring a training remote sensing image of a training area, wherein the training remote sensing image is marked with a house roof;
slicing the training remote sensing image and the house surface label according to a preset size to obtain a slice training data set;
and training the deep learning model to be trained by using the slice training data set to obtain a pre-trained deep learning model.
3. The method of acquiring natural village boundaries as defined in claim 2, wherein extracting each house surface position in the target remote sensing image using a pre-trained deep learning model comprises:
Carrying out sliding window processing in the target remote sensing image by adopting a preset path, and extracting the position of a house roof in each sliding window;
and integrating the house surface positions in all sliding windows to obtain each house surface position in the target remote sensing image.
4. The method for acquiring natural village boundaries according to claim 1, wherein clustering to generate concave packets at a predetermined distance according to each house surface position comprises:
calculating the distance between every two house surface positions;
classifying a house face position to be classified, which is less than a preset distance from a reference house face position, and the reference house face position into the same category, and setting the natural village attribution attribute of the house face position to be classified into the same natural village as the reference house face position, wherein the reference house face position is the house face position to which the natural village attribution attribute is assigned;
and solving the concave bag for the house surface position of which the home attribute of the natural village is the same natural village.
5. An acquisition apparatus for a natural village boundary, wherein the acquisition apparatus for a natural village boundary comprises:
the target remote sensing image acquisition module is used for acquiring a target remote sensing image containing a target administrative village region;
The house face position acquisition module is used for extracting each house face position in the target remote sensing image by adopting a pre-trained deep learning model;
the Thiessen polygon generation module is used for generating Thiessen polygons according to the pre-marked natural village POIs;
the first concave packet generation module is used for generating concave packets in a clustering mode according to the positions of the house surfaces and preset distances;
the first concave packet traversing module is used for traversing each concave packet one by one, judging whether the target concave packet is intersected with more than one Thiessen polygon according to each traversed target concave packet, if not, setting the natural village attribution attribute of the house surface position in the target concave packet as a natural village corresponding to the POI, and marking the target concave packet as processed;
the first concave packet traversing module is further configured to determine whether the target concave packet contains more than 1 POI if the target concave packet intersects more than one taylon polygon, and if not, set a natural village attribution attribute of a house surface position in the target concave packet as a natural village corresponding to the POI to which the taylon polygon with the largest area occupied by the target concave packet belongs;
the first concave packet traversing module is further used for cutting the target concave packet by using a Thiessen polygon to generate N corresponding concave packet sub-areas if the target concave packet contains more than 1 POI, wherein N is an integer greater than 1;
The first concave packet traversing module is further used for judging whether each target concave packet sub-area in the N concave packet sub-areas contains POIs one by one, if so, setting the natural village attribution attribute of the house roof position in the target concave packet sub-area as the natural village corresponding to the POIs, and if not, merging the house surface to the processed natural village nearest to the target concave packet sub-area;
and the second concave packet generation module is used for solving concave packets of roof positions belonging to the same natural village by utilizing the natural village attribute of the classified house surface positions after traversing each concave packet one by one, so as to obtain the range of each natural village.
6. The natural village boundary acquisition device as defined in claim 5, wherein the natural village boundary acquisition device further comprises:
the training remote sensing image acquisition module is used for acquiring a training remote sensing image of a training area, wherein the training remote sensing image is marked with a house roof;
the training data set slicing module is used for slicing the training remote sensing image and the house surface label according to a preset size to obtain a slicing training data set;
and the deep learning model training module is used for training the deep learning model to be trained by using the slice training data set to obtain a pre-trained deep learning model.
7. The natural village boundary acquisition device as defined in claim 5, wherein the house face position acquisition module comprises:
the sliding window processing module is used for carrying out sliding window processing in the target remote sensing image by adopting a preset path and extracting the position of a house roof in each sliding window;
and the house face position integrating module is used for integrating the house face positions in all sliding windows to obtain each house face position in the target remote sensing image.
8. The natural village boundary acquisition device as defined in claim 7, wherein the first concave packet generation module comprises:
the roof position interval calculation module is used for calculating the distance between every two house surface positions;
the roof position clustering module is used for classifying the house face position to be classified, which is smaller than the preset distance, from the reference house face position into the same category, and setting the natural village attribution attribute of the house face position to be classified into the same natural village as the reference house face position, wherein the reference house face position is the house face position to which the natural village attribution attribute is assigned;
the first concave packet solving module is used for solving concave packets for house surface positions of which the attribute of the natural village is the same natural village.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of acquiring natural village boundaries as claimed in any one of claims 1 to 4 when the computer program is executed by the processor.
10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements a method of acquiring a natural village boundary as claimed in any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311657980.8A CN117351368B (en) | 2023-12-06 | 2023-12-06 | Natural village boundary acquisition method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311657980.8A CN117351368B (en) | 2023-12-06 | 2023-12-06 | Natural village boundary acquisition method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117351368A CN117351368A (en) | 2024-01-05 |
| CN117351368B true CN117351368B (en) | 2024-02-23 |
Family
ID=89371467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311657980.8A Active CN117351368B (en) | 2023-12-06 | 2023-12-06 | Natural village boundary acquisition method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117351368B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521973A (en) * | 2011-12-28 | 2012-06-27 | 昆明理工大学 | Road matching method for mobile phone switching positioning |
| CN115311576A (en) * | 2022-08-09 | 2022-11-08 | 中山大学 | Natural village boundary extraction method based on rural house distribution |
| CN117152604A (en) * | 2023-07-31 | 2023-12-01 | 河北师范大学 | Building contour extraction method and device, electronic equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7399220B2 (en) * | 2002-08-02 | 2008-07-15 | Kriesel Marshall S | Apparatus and methods for the volumetric and dimensional measurement of livestock |
-
2023
- 2023-12-06 CN CN202311657980.8A patent/CN117351368B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521973A (en) * | 2011-12-28 | 2012-06-27 | 昆明理工大学 | Road matching method for mobile phone switching positioning |
| CN115311576A (en) * | 2022-08-09 | 2022-11-08 | 中山大学 | Natural village boundary extraction method based on rural house distribution |
| CN117152604A (en) * | 2023-07-31 | 2023-12-01 | 河北师范大学 | Building contour extraction method and device, electronic equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| Yaping Chen ; Akot Deng.Using POI Data and Baidu Migration Big Data to Modify Nighttime Light Data to Identify Urban and Rural Area.《 IEEE Access》.2022,第93513-93524页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117351368A (en) | 2024-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Aljumaily et al. | Urban point cloud mining based on density clustering and MapReduce | |
| JP6032467B2 (en) | Spatio-temporal data management system, spatio-temporal data management method, and program thereof | |
| CN110413816B (en) | Color Sketch Image Search | |
| CN109815993B (en) | GPS track-based regional feature extraction, database establishment and intersection identification method | |
| El‐Sayed et al. | Plane detection in 3D point cloud using octree‐balanced density down‐sampling and iterative adaptive plane extraction | |
| Wei et al. | On the spatial distribution of buildings for map generalization | |
| Galvanin et al. | Extraction of building roof contours from LiDAR data using a Markov-random-field-based approach | |
| US20150178594A1 (en) | Point cloud simplification | |
| CN113223042B (en) | Intelligent acquisition method and equipment for remote sensing image deep learning sample | |
| Bastani et al. | Machine-assisted map editing | |
| CN111090712A (en) | Data processing method, device and equipment and computer storage medium | |
| CN111339336B (en) | Analysis method and system for house type data | |
| CN111291811A (en) | Region dividing and classifying method, device and computer readable storage medium | |
| CN111738319B (en) | Clustering result evaluation method and device based on large-scale samples | |
| CN110956213A (en) | Method and device for generating remote sensing image feature library and method and device for retrieving remote sensing image | |
| Eum et al. | Vehicle detection from airborne LiDAR point clouds based on a decision tree algorithm with horizontal and vertical features | |
| Yu et al. | Change detection for 3D vector data: a CGA-based Delaunay–TIN intersection approach | |
| Ozdemir et al. | A novel algorithm for regularization of building footprints using raw LiDAR point clouds | |
| CN113298042A (en) | Method and device for processing remote sensing image data, storage medium and computer equipment | |
| CN117351368B (en) | Natural village boundary acquisition method and device, electronic equipment and storage medium | |
| Ahmed et al. | Choosing thresholds for density-based map construction algorithms | |
| CN105740878A (en) | Symbol recognition method and device of geographic information system | |
| CN110532437B (en) | Electronic certificate prompting method, electronic certificate prompting device, computer equipment and storage medium | |
| Beilschmidt et al. | A linear-time algorithm for the aggregation and visualization of big spatial point data | |
| CN110188156A (en) | A kind of work transmission line three dimensional design achievement key message extracting method and system |
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 |