WO2018113787A1

WO2018113787A1 - Region division method and device, and storage medium

Info

Publication number: WO2018113787A1
Application number: PCT/CN2017/118096
Authority: WO
Inventors: 赵娟娟; 罗圣美; 范小朋; 刘丽霞; 文韬; 吉锋
Original assignee: 中兴通讯股份有限公司; 中国科学院深圳先进技术研究院
Priority date: 2016-12-23
Filing date: 2017-12-22
Publication date: 2018-06-28
Also published as: CN108242203A

Abstract

Embodiments of the present invention provide a region division method and device. The method comprises: obtaining a region boundary of a region to be divided, and calculating a rectangular boundary of the region to be divided; obtaining location information of traffic hubs in the region to be divided, and drawing, by using the traffic hubs as discrete points, a shortest-path polygon; performing region combination processing on the shortest-path polygon and the rectangular boundary of the region to be divided, and obtaining a preliminary region division result; obtaining a feature vector calculation parameter of a point of interest in the region to be divided, and calculating, according to the feature vector calculation parameter, a feature vector of each point of interest in each region in the preliminary region division result; calculating, according to the feature vector of each point of interest in each region, a region similarity between each region and a spatially adjacent region thereof; and combining, according to the region similarity, regions in the preliminary region division result, and generating a final region division result. The embodiments of the present invention further disclose a computer storage medium.

Description

Method and device for dividing area and storage medium

The present application is filed on the basis of the Chinese Patent Application No. No. No. No. No. No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No

Technical field

The present invention relates to the field of information technology, and in particular, to a method and device for partitioning a region and a storage medium.

Background technique

In recent years, with the rapid development of LBS (Location Based Service) and GIS (Geographic Information System), urban area research has become more and more important in many fields, such as path planning and transportation. Traffic analysis, regional traffic analysis, regional population assessment, regional attributes, population movements, inter-regional relationships, etc., before these analyses are performed, the city must first be divided into regions.

In order to make the area division simple, the most common method is the grid-based map segmentation method, which divides the map into fixed-size squares. This mechanical division leads to the division of the original location with great relevance. In different areas, location-based services such as location services are not well provided, resulting in poor location-based service experience.

Summary of the invention

The embodiment of the invention provides a region dividing method and device to solve the problem of mechanical division of the existing grid map.

In one aspect, an embodiment of the present invention provides a method for area division, including:

Obtaining an area boundary of the area to be divided, and calculating a rectangular boundary of the area to be divided;

Obtaining the location information of the transportation hub point in the area to be divided, and drawing the shortest path polygon by using the transportation pivot point as a discrete point;

Combining the shortest path polygon with the rectangular boundary of the area to be divided, and obtaining the preliminary area division result, the preliminary area division result includes multiple areas;

Obtaining and calculating a parameter according to the feature vector of the interest point in the region to be divided, and calculating a feature vector of each interest point in each region in the preliminary region division result;

Calculating the regional similarity between each region and its spatially adjacent region according to the feature vector of each interest point in each region;

According to the regional similarity, the regions in the preliminary regional division result are combined to generate the final regional division result.

In one aspect, an embodiment of the present invention provides an area dividing apparatus, including: a boundary module, a point of interest module, and an optimization module, where

The boundary module is configured to obtain a boundary of the area to be divided, calculate a rectangular boundary of the area to be divided, obtain position information of the transportation pivot point in the area to be divided, and draw a shortest path polygon by using the transportation pivot point as a discrete point; and the shortest path polygon and the to be divided The regional rectangular boundary is subjected to regional combination processing to obtain preliminary regional division result, and the preliminary regional division result includes multiple regions;

a point of interest module, configured to acquire and calculate a parameter according to a feature vector of the point of interest in the area to be divided, and calculate a feature vector of each point of interest in each area in the preliminary area division result;

The optimization module is configured to calculate the regional similarity between each region and its spatial neighboring region according to the feature vector of each interest point in each region; according to the regional similarity, the regions in the preliminary region segmentation result are combined to generate a final region segmentation result. .

In another aspect, an embodiment of the present invention provides a computer storage medium. The computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the foregoing area division method.

The embodiment of the invention provides a method and device for dividing a region and a computer storage medium. Firstly, by using the distribution characteristics of a transportation hub such as a bus stop, the path of the segmentation site is first divided into a shortest path polygon region such as a Tyson Path-Voronoi polygon. Then, by constructing POI feature vectors of interest points to merge adjacent regions, the road network and regional POI characteristics can be fully taken into account, so that the two regions with sufficiently high POI feature vector similarity in the preliminary division are merged, so that the POI feature vector is made. The same or similar positions are divided into one area, instead of the mechanical division of the equal area of the rectangular division in the prior art, obviously avoiding the problem of unreasonable division of the area caused by mechanical division, and having the characteristics of high regional division rationality. In addition, when a service is provided based on such a divided area, a higher quality service can be provided. For example, when the location-based service is provided, when the user holds the device in an area, all the POI information with similarity can be fully provided to the user, thereby realizing the full delivery of the POI information, thereby Reasonable user complaints or low usage satisfaction caused by the lack of POI information.

DRAWINGS

FIG. 1 is a flowchart of a method for dividing a region according to an embodiment of the present invention;

2 is a schematic structural diagram of an area dividing apparatus according to an embodiment of the present invention;

3 is a flowchart of a method for dividing an urban area according to an embodiment of the present invention;

4 is a diagram of a city boundary and a rectangular boundary according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of preliminary division results according to an embodiment of the present invention; FIG.

6 is a schematic diagram of a final division result according to an embodiment of the present invention;

7 is a partial enlarged view of a preliminary division result according to an embodiment of the present invention;

FIG. 8 is an enlarged partial view of a partial division result according to an embodiment of the present invention; FIG.

FIG. 9 is a schematic diagram of a path-Tyson polygon calculation according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is understood that the preferred embodiments described below are only used to illustrate and explain the present invention, and are not intended to limit this invention.

FIG. 1 is a flowchart of a method for dividing a region according to an embodiment of the present invention. As shown in FIG. 1 , the method for dividing a region provided by this embodiment includes:

S101: Acquire an area boundary of the area to be divided, and calculate a rectangular boundary of the area to be divided;

S102: Obtain location information of a transportation hub point in the area to be divided, and draw a shortest path polygon by using the transportation pivot point as a discrete point;

S103: Combine the shortest path polygon with the rectangular boundary of the to-be-divided area to obtain a preliminary area division result, where the preliminary area division result includes multiple areas;

S104: Acquire and calculate a parameter according to a feature vector of the point of interest in the area to be divided, and calculate a feature vector of each point of interest in each area in the preliminary area division result;

S105: Calculate a regional similarity between each region and a spatially adjacent region according to a feature vector of each interest point in each region;

S106: Combine the regions in the preliminary region division result according to the regional similarity to generate a final region division result.

In some embodiments, the obtaining location information of the transportation hub point in the area to be divided in the foregoing embodiment includes:

Obtaining a location data set of all traffic sites in the area to be divided; the traffic site includes at least one of a bus stop, a subway station, a dock, and an airport. In some embodiments, the transportation site may further include: a plurality of road junctions, or may further include: a location where the traffic volume or the traffic volume is greater than a preset value.

The location data of the plurality of traffic stations having the predetermined association relationship is merged to obtain the location information of the transportation hub point.

The plurality of transportation stations having the predetermined association relationship in the embodiment of the present invention may include: a transportation station having the same name. For example, the bus route includes an upward route from the first direction to the second direction, and further includes: a downward route from the second direction to the first direction. The up route and the down route are round trips. In some cases, the bus routes with the same name are included in the uplink route and the downlink route, but the specific locations are different. For this case, the sites with the same names on the two routes are merged to form one of the traffic hub points. For example, take the midpoints of two bus stops of the same name in the up and down routes.

For another example, in some embodiments, there are 4 bus stops or 3 subway stations at an intersection, and the distribution is very close to the edge of the intersection. At this time, the four bus stops can also be combined. / or 3 subway stations, using the intersection as the transportation hub. The method of merging includes: connecting two traffic stations having a predetermined association relationship, and taking a midpoint of the connection line as the transportation hub point. When there are more than two traffic stations with a predetermined association relationship, the traffic stations are sequentially connected to form a polygon, and the geometric center of the polygon or the midpoint of a certain side coincident with the road is taken as the location of the transportation pivot point. In some cases, a separate traffic site with no other site and its predetermined association relationship may also be referred to as the transportation hub point. Therefore, the transportation hub point may be one of a bus stop, a subway, a dock or an airport, or may be a location point formed by combining a plurality of traffic stations having a predetermined association relationship. Of course, the above is only an example, and the specific implementation is not limited to any of the above.

In this embodiment, the point may be a Point Of Interest (POI). The corresponding POI is generally provided with POI information. The POI information is an information element in the geographic information, and is information of a website based on geographic information, a public service station, a bus station, or the like, or a service station capable of providing a service. Typically each of the POI information may include information such as the name of the service site and/or corresponding code, the type of service provided, and traffic conditions.

In some embodiments, the shortest path polygon in the above embodiment is a path-Tyson polygon, and the traffic point is a discrete point, and drawing the shortest path polygon includes:

Obtaining the shortest road path between the first transportation hub point and the second transportation hub point;

Obtaining a center of a minimum circle covering the shortest road path between the first transportation hub point and the second transportation hub point;

Connecting the center of the circle with the midpoint of the first transportation hub point and the second transportation hub point as the boundary between the first transportation hub point and the second transportation hub point;

Loop through all the traffic hub points and get the shortest path polygon.

Typically, the minimum circle completely covers the mid-circle radius of the shortest path connecting the first traffic junction point and the transportation hub point or performs the smallest circle. That is, the shortest path between the first traffic junction point and the second traffic junction point is located within the minimum circle.

In some embodiments, the obtaining preliminary region division result in the foregoing embodiment includes:

Obtaining the administrative area of the smallest administrative unit in the rectangular boundary of the area to be divided;

The administrative region and the region in the shortest path polygon are intersected to generate a preliminary region division result.

In some embodiments, the feature vector calculation parameters in the foregoing embodiment include location information, a radius of radiation, and a topic weight; and calculating feature vectors of each point of interest in each region in the preliminary region division result includes:

Obtain the subject radiation range and theme weight of the point of interest;

Statistics on the number of points of interest for each topic in each region;

Statistics acquire the feature vectors of all points of interest in each region.

In some embodiments, the method for dividing a region in the foregoing embodiment, before calculating the feature vector of each point of interest in each region in the preliminary region segmentation result, further includes:

Setting a threshold number of regions to be divided into regions;

Determining whether the number of regions in the preliminary region division result is greater than a threshold number of regions;

If greater than, the feature vector of each interest point in each region in the preliminary region division result is calculated;

If it is not greater than, the preliminary region division result is used as the final region division result.

In some embodiments, the regions in the preliminary region division result are combined according to the regional similarity in the foregoing embodiment, and the final region partitioning result is generated by:

Setting an area threshold of the area to be divided;

Selecting two regions in each region that have the largest similarity and the area of the region and the threshold that is not greater than the region threshold, and merge the two regions into one region;

Determining whether the number of merged regions is less than a threshold number of regions;

If the number of the merged regions is not greater than the threshold number of the regions, the partitioning region task ends, and the final region partitioning result is output;

If the number of merged regions is greater than the number of regions threshold, then the merge is continued through the points of interest.

2 is a schematic structural diagram of an area dividing apparatus according to an embodiment of the present invention. As shown in FIG. 2, the area dividing apparatus provided in this embodiment includes: a boundary module 21, a point of interest module 22, and an optimization module 23, where

The boundary module 21 is configured to obtain a boundary of the area to be divided, calculate a rectangular boundary of the area to be divided, obtain position information of the transportation pivot point in the area to be divided, and draw a shortest path polygon by using the transportation pivot point as a discrete point; Dividing the rectangular boundary of the area to perform regional combination processing, and obtaining preliminary regional division result, and the preliminary regional division result includes multiple regions;

The point of interest module 22 is configured to acquire and calculate a parameter according to a feature vector of the point of interest in the area to be divided, and calculate a feature vector of each point of interest in each area in the preliminary area division result;

The optimization module 23 is configured to calculate the regional similarity between each region and its spatial neighboring region according to the feature vector of each interest point in each region; and combine the regions in the preliminary region segmentation result according to the region similarity to generate a final region segmentation. result.

In some embodiments, the boundary module 21 in the foregoing embodiment is configured to acquire a location data set of all traffic stations in the area to be divided, merge location data of multiple traffic stations having a predetermined association relationship, and obtain a location of the traffic pivot point. Information; the transportation site includes at least one of a bus stop, a subway station, a dock, and an airport. In the embodiment of the present invention, the traffic site may be simply referred to as a site.

In some embodiments, the boundary module 21 in the foregoing embodiment is configured to acquire a shortest road path between the first transportation junction point and the second transportation pivot point; and acquire the first transportation hub point and the second transportation pivot point. The center of the smallest circle of the shortest road path; the line connecting the center of the circle with the midpoint of the first traffic junction point and the second traffic junction point as the boundary between the first traffic junction point and the second traffic junction point; The traffic hub point, the shortest path polygon is obtained, and the shortest path polygon is the path-Tyson polygon.

In some embodiments, the boundary module 21 in the foregoing embodiment is configured to acquire an administrative area of a minimum administrative unit in a rectangular boundary of the area to be divided; and intersect the administrative area with the area in the shortest path polygon to generate a preliminary area division result.

In some embodiments, the feature vector calculation parameters in the foregoing embodiment include location information, a radius of radiation, and a topic weight; the point of interest module 22 is configured to acquire a subject radiation range and a topic weight of the point of interest; and statistically obtain each topic of each region. The number of points of interest; statistics obtain the feature vectors of all points of interest in each region.

In some embodiments, the optimization module 23 in the foregoing embodiment is further configured to: before calculating the feature vector of each interest point in each region in the preliminary region division result, set a threshold number of regions to be divided; and determine a preliminary region division result. Whether the number of regions in the region is greater than the threshold number of the region; if greater, the feature vector of each point of interest in each region in the preliminary region segmentation result is calculated; if not greater, the preliminary region segmentation result is used as the final region segmentation result.

In some embodiments, the optimization module 23 in the above embodiment is configured to set an area threshold of the area to be divided; and select two areas in which the area similarity is the largest and the area is not larger than the area threshold. The two regions are merged into one region; the number of the merged regions is determined to be smaller than the threshold number of the region; if the number of merged regions is not greater than the threshold for the number of regions, the task for dividing the region ends, and the final region partition result is output; if the number of merged regions is If it is greater than the number of regions threshold, then continue to merge through the points of interest.

In practical applications, all the functional modules in the embodiment shown in FIG. 2 can be implemented by using a processor, an editing logic device, or the like.

In this embodiment, the area to be divided is taken as an example. In the actual application, the area to be divided may be any national, provincial, and village level map. The present invention is further explained in conjunction with a specific application scenario.

Regional POI (Point of Interest) is a very important indicator of division. Regional POI refers to points of interest in the area, such as restaurants, shopping malls, stadiums, schools, companies, and so on. Each POI contains at least four aspects of information: name, category, longitude, and latitude. The POI information reflects the location semantics of the region and is the meta-characteristic of the region. The human traffic is attracted by the POI. For this reason, understanding the POI distribution of the region is very important for calculating the traffic of the region. The Tyson Polygon (Voronoi Polygon) is a method proposed by the Dutch meteorologist A.H. Thiessen to partition the plane by using discrete points in the plane. The method first joins the discrete points in the plane into triangles and then pairs the triangles. The vertical bisector of each side and the polygon formed by the intersection of the vertical bisectors are Voronoi polygons. This method ensures that there is only one discrete point in each Voronoi polygon. Originally, the method was used to calculate regional rainfall based on the rainfall of discretely distributed weather stations. The method is now widely used in base station coverage calculation, crystal structure, and geographical division. However, the shortcoming of using the method for traffic area division is that the method does not consider the degree of association between regions, which may lead to the case where dense business circles are divided into multiple regions, which brings difficulties for upper-layer applications based on region division. .

In this embodiment, by using the distribution characteristics of the bus station, the Path-Voronoi polygon region of the segmentation site is first divided, and then the POI feature vector is constructed to merge the adjacent regions, so that the road network and the regional POI characteristics can be fully considered.

Optionally, the present embodiment aims to construct a general traffic area dividing method, and complete the traffic area division of various different requirements by merging the small areas with different features from the bottom up.

In this paper, based on the shortcomings of grid map segmentation method, combined with the advantages of Voronoi polygon partitioning method, a simple, efficient and adaptable traffic area division method based on bus station and POI (Boundary-Tyson) is proposed.

In practical applications, large and medium-sized cities have developed public transportation systems, and basically cover the entire range of urban areas. Therefore, using the bus station as the basic discrete point, combined with the POI information of each topic near the bus station, the Path-Voronoi polygon algorithm is used. The city is divided into regions.

Optionally, as shown in FIG. 3, the area division method provided in this embodiment includes the following main steps:

1. Obtain the location data set of the city bus station.

Assuming that there are q bus stops in the target city, each bus station has an uplink and a downlink. The target city has 2q bus station location data, which are represented as L{L11, L12, L21, L22, ... Lq1, Lq2}. The element Li1 in the set represents the uplink station latitude and longitude (lngi1, lati1) of the i-th bus station, and Li2 represents the downlink station latitude and longitude (lngi2, lati2) of the i-th bus station.

2. Combine the location data of the uplink and downlink stations of the bus.

The average of the above latitude and longitude is taken as the location of the site after the merger. For example, the merged position of the uplink and downlink sites of the i-th bus station is Li=(lngi,lati), where

Lngi=(lngi1+lngi2)/2

Lati=(lati1+lati2)/2

3. Get the city boundary.

The boundary of a city is an irregular polygon composed of multiple points.

Its main sources are:

Buy a vector map directly and get a vector polygon from the city border

Obtain a vector map from a web service for business users provided by a map service provider such as Baidu and Gaode, and obtain a vector polygon of the city boundary from

Use a special GIS software such as ArcGIS to draw a vector map according to the scale of the existing paper grid map, and obtain the vector polygon of the city boundary from it.

If none of the above methods are available, please configure the person to collect the map data and draw a vector map to obtain the vector polygon of the city boundary.

4. Get the rectangular boundary of the city.

The maximum longitude lngmax of all boundary points on the city boundary, the maximum latmax of the latitude is taken as the coordinates of the upper right vertex of the rectangular boundary (lngmax, latmax), and the longitude of the longitude of all the boundaries of the city is lngmin, the minimum value of latitude is latmin, as the boundary of the rectangle Left lower vertex coordinates (lngmin, latmin);

5. Set the threshold of the number of urban areas and the threshold of the area.

The threshold N' of the number of urban areas is set according to the granularity requirements of the user of the different regions for the division of the area. The specific thresholds are set as follows:

The threshold of the number of urban areas is in principle not greater than the number of sites merged in the second step;

The threshold of the number of urban areas is not less than the number of areas required by the regional users, and should generally be equal to the number of areas required by the regional users;

If the number of areas required by the area consumer is greater than the number of stations merged in step 2, the threshold of the number of urban areas should obey the first principle. In general, the number of bus stops in the next-line cities will be as many as several thousand, which is enough to meet the requirements of regional users.

In order to ensure that the size of the divided area is maintained at a certain level, and the result of the division of the interference area of the very large area is avoided, the urban area threshold S' is preset. The specific thresholds are set as follows:

The urban area area threshold S' is in principle not greater than the urban area Scity;

The urban area threshold S' should generally not be greater than the minimum of the urban secondary administrative division (district level). For example, the city has n district-level administrative divisions, the areas of which are S1, S2...Sn, and the smallest area Smin (min belongs to 1...n), the urban area threshold S' should not generally be Greater than Smin;

The urban area area threshold S' is in principle not less than the area of the triangle formed by any three of the merged stations in the step 2). For example, if there are n stations in the city and their positions are L1, L2...Ln, then the urban area area threshold S' should not be larger than the area of the triangle LaLbLc (a is not equal to b is not equal to c, and abc is greater than 1 and less than n). ).

6. Generate a Path-Voronoi polygon by taking the bus station as a discrete point.

As shown in FIG. 9, point A and point B in the figure are bus stations, and the curve path from point A to point B is the shortest road path from point A to point B. The traditional Voronoi polygon is a boundary between the AB and the two perpendicular points A and B. The Path-Voronoi method uses the center D of the smallest circle covering the shortest path between AB, and the line of point C in the same line AB is the boundary between A and B. Normally, the line segment AB and its shortest road path form a closed polygon; when the shortest road path from point A to B coincides with the line segment AB, the vertical line of the line segment AB is taken as the boundary between A and B; When the line segment AB is the smallest round chord, the CD line is the vertical line in the line segment AB. In this case, the division method degenerates into the Voronoi polygon method. The polygon constructed by the Path-Voronoi method is called a Path-Voronoi polygon.

7. Divide the polygon area within the city boundary.

The region after the Voronoi polygon partition is intersected with the corresponding administrative region of the city, and the polygon outside the boundary is subtracted to obtain a preliminary region division result;

8. Determine if the number of regions is less than the threshold.

If the number of regions is smaller than the threshold number N' of the number of urban regions, it is considered to satisfy the region division condition and jump to the last step. On the contrary, it is necessary to continue to further merge the regions through the POI and proceed to the next step.

9, get POI information

Divide POI information into different topics according to categories, and collect POI information of cities according to the theme;

10. Set the POI theme radiation radius.

The subject radiation range Ri (Ri belongs to R1, R2 ... Rn) is set according to the theme i of the POI (i belongs to 1, 2...n).

Calculation formula:

Where k is the radiation constant, and the range of values is generally between 1.2 and 3.

The specific setting principles are as follows:

Calculate the average area of the area occupied by each topic category point in the city Si (Si∈S1, S2, ....Sn). Under normal circumstances, due to the large sample size and insufficient information, it is difficult to make statistics on all individual areas, and the subject area estimation is carried out by sampling hypothesis test.

11. Set the POI theme weight.

The theme weights Pi(Pi∈P1, P2...Pn) are set according to the theme 1, 2, .

The setting of this weight depends on the industry in which the regional user belongs, and the POI weight with higher industry association is larger. For example, the traffic control department generally does not care about the POI point of the Internet cafe theme. In this application, the weight of the Internet cafe can be set to zero.

12. Count the POI quantity factors for each topic in each region.

Assume that the number of topics 1 in area A is IA1, and the number of topics 1 in the surrounding area is OA1.

Then the quantity factor NA1 of the subject 1 of the area A is:

Wherein rA1i is the distance of the i-th region A in the POI of the subject 1 category of the peripheral region; R1 is the radius of the radiation range of the subject 1; and the length rA1i is the same as the unit of R1;

The radiation factor from the i-th to the region A in the POI of the topic 1 category of the surrounding area.

13. Count POI feature vectors for each region.

Assuming that the number factor of the subject 1 in the polygon area A is NA1, the area of the polygon area A is SA, and the weight of the subject 1 is P1, the feature factor DA1 of the subject 1 of the area A is:

It follows that the POI feature vector of region A is:

14. The similarity between the statistical area and the adjacent area.

Assume that each topic POI feature vector of area A

The subject POI feature vector of a neighboring region B

Since the Euclidean distance only reflects the relative distance between the vectors, and the cosine distance only reflects the vector angle information, the valley distance (also known as the Jaccard distance), which can simultaneously represent the vector angle and the relative distance information, is used to measure the vector relationship. Two vectors

with

The valley distance measure is taken as the distance d _AB between the two regions, and the formula is

Then its similarity is:

Iterate through all regions and all its neighbors, calculate the similarity and put it into the neighborhood relationship set M.

It is worth noting that the similarity between the area A and the area B is equal to the similarity between the area B and the area A, that is, mAB=mBA. Therefore, the similarity between the two regions is only calculated once.

15. Similar areas merge.

The two regions in the neighborhood relationship set M having the largest regional similarity and the area area and not larger than the area threshold S' are selected, and the two regions are merged into one region X.

16. Determine whether the number of regions is less than a threshold.

If the number of regions is smaller than the threshold number N' of the number of urban regions, it is considered that the region dividing target is satisfied, and the dividing region task ends. On the contrary, it is necessary to continue to further merge the regions through the POI and proceed to the next step.

17. Update the relationship between the merged area and the neighborhood,

Recalculate the POI quantity factor of all topics in the merged area X. For the calculation method, see step 12.

Recalculate the POI feature vector of the merged region X. For the calculation method, see step 13.

The similarity of the merged region X to the adjacent region is updated to the neighborhood relationship set M.

Enter the zone merge step 15.

18. Iterate over the sub-administrative area.

When the regional user is an administrative agency or enterprise classified according to the administrative division, if there is a need for the division of the traffic area under the corresponding sub-administrative division, the existing polygons may be further divided by the sub-administrative division to facilitate the relevant organizations. management. See step 7 for the method of segmentation.

Optionally, taking the traffic area division of Shenzhen as an example, taking the Shenzhen bus station as a discrete point, using the improved Voronoi polygon algorithm combined with the POI information of Shenzhen to divide the city, the key steps are as follows:

Get the city bus site location data set. Obtained 11,710 bus route stations in Shenzhen and deposited them in the database.

Combine the location data of the bus uplink and downlink sites. The bus station is merged into one site, and the 11,710 sites obtained in the first step are merged according to the uplink and downlink respectively, and 5,865 sites are obtained; the name, location and related information of the site are stored in the spatial database.

Get the city border. Obtain the Shenzhen city boundary and store it as a polygon to the spatial database.

Get the rectangular border of the city. Set the rectangular boundary of Shenzhen, which must be larger than the Shenzhen city boundary, that is, take the largest and smallest latitude and longitude of Shenzhen, and conclude that the smallest circumscribed rectangle of Shenzhen is stored in the database, as shown in Figure 4.

Set the city area number threshold and area area threshold. According to the current needs, the city area number threshold and the area area threshold are set by the configuration file.

The Path-Voronoi polygon is generated by taking the bus station as a discrete point.

Taking the minimum circumscribed rectangle of the fourth step above as the original area before the partition, the 5865 bus stations merged in the second step are the basic discrete points, and the Path-Voronoi method is used to generate the equal-path Path-Voronoi polygon based on the shortest path division; The polygon area within the city boundary intersects all the Path-Voronoi polygon areas with the corresponding city administrative area to obtain the preliminary area division result, as shown in Fig. 5.

Save the acquired area polygons to the spatial database.

Determine if the number of regions is less than the threshold

The threshold of the number of urban areas in the configuration file is read, and it is determined whether the current number of all the stations is less than the threshold. If not, the process proceeds to the next step.

Get information about the POI.

The POI categories are divided into the following 20 topics, as shown in Table 1 below:

Table 1

Obtain all the POIs of the above 20 categories, and store the information (location, name, city, street, classification, etc.) of each POI point into the database.

Set the POI theme radiation radius. The POI topic radiating radius database for each topic is set according to the basic principles.

Set the POI theme weights. The POI theme weights of each topic are set in the database according to the basic principles.

Count the POI quantitative factors for each topic in each region. The POI quantity factor is set for all topics of all regions delineated according to the method described in step 12.

The POI feature vectors of each region are counted. The POI feature vector in all regions delineated is calculated according to the method described in the thirteenth step.

The similarity between the statistical area and the adjacent area. According to the method described in the 14th step of the Summary of the Invention, the valley distance is calculated for the feature vector of each pair of neighborhoods as the neighborhood similarity.

Similar areas are merged. According to the method described in the fifteenth step, similar region merging is performed with the similarity as the main criterion.

Determine whether the number of regions is less than or equal to the threshold. According to the method described in the 16th step, here is the exit decision of the area division task. If the partitioning requirements are not met, proceed to the next step.

Update the relationship between the merged region and the neighborhood. According to the method described in the 17th step, the relationship between the merged area and the neighborhood is updated and the process proceeds to step 15 to continue the merge.

After repeated iterative merging of the regions, the final result can be obtained as shown in FIG. 6.

As shown in Fig. 7 and Fig. 8 is an enlarged version of a certain area, as shown in Fig. 7 and Fig. 8, comparing the preliminary partitioning result and the final partitioning result, it can be seen from the area of the ellipse circle in the figure that the final partitioning result will be similar to the POI. The very high-level areas have been merged to provide a complete and independent basis for analysis such as the popularity index of the business circle and passenger flow.

In summary, through the implementation of the embodiments of the present invention, at least the following beneficial effects exist:

The embodiment of the invention provides a region division method, which firstly divides the path of the segmentation site - the shortest path polygon region such as the Tyson Path-Voronoi polygon by using the distribution characteristics of the traffic hub such as the bus stop, and then constructs the interest point by constructing the interest point. The POI feature vector merges adjacent regions, which can fully take into account the road network and regional POI characteristics, making the division of urban regions more reasonable, and solving the problem that the existing regional division based on grid maps is unreasonable.

Those skilled in the art will appreciate that embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention can take the form of a hardware embodiment, a software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) including computer usable program code.

The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or FIG. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine for the execution of instructions for execution by a processor of a computer or other programmable data processing device. Means for implementing the functions specified in one or more of the flow or in a block or blocks of the flow chart.

The computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device. The apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.

These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

The above is only a specific embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modifications made in accordance with the principles of the present invention should be construed as falling within the scope of the present invention.

Industrial applicability

In the embodiment of the present invention, when the area is divided, the mechanical area is no longer divided into rectangles. First, the position information of the pivot point is selected, and the preliminary division of the area is performed based on the position information of the transportation pivot point, and then is located around the transportation pivot point. The feature vector of the POI, the region is merged with the preliminary divided regions having the same POI attribute or the similar POI attribute, so that the POIs having the same or similar POI attributes can be divided into the same area, obviously not in the mechanical division area, thereby The forced unreasonable division of the POI attribute in the mechanical division area is avoided, and the area-based service based on the division can provide a higher quality service, thereby having a positive industrial effect. In the process of implementation, the area division can be realized by the execution of computer code, etc., and has the characteristics of strong achievability.

Claims

A method of area division, comprising:

Obtaining an area boundary of the area to be divided, and calculating a rectangular boundary of the area to be divided;

Obtaining location information of the transportation hub point in the area to be divided, and drawing the shortest path polygon by using the transportation pivot point as a discrete point;

Performing a combination processing on the shortest path polygon and the rectangular boundary of the to-be-divided area to obtain a preliminary area division result, where the preliminary area division result includes multiple areas;

Obtaining and calculating a parameter according to the feature vector of the point of interest in the area to be divided, and calculating a feature vector of each point of interest in each area in the preliminary area division result;

Calculating a regional similarity between each region and a spatially adjacent region according to the feature vector of each interest point in each region;

According to the regional similarity, the regions in the preliminary region division result are combined to generate a final region division result.
The area dividing method according to claim 1, wherein the obtaining the location information of the transportation hub point in the area to be divided comprises:

Obtaining a location data set of all traffic stations in the area to be divided; the transportation station includes at least one of a bus station, a subway station, a dock, and an airport;

The location data of the plurality of traffic stations having the predetermined association relationship is merged to obtain the location information of the transportation hub point.
The area dividing method according to claim 1, wherein the shortest path polygon is a path-Tyson polygon, and the traffic point is a discrete point, and drawing the shortest path polygon comprises:

Obtaining the shortest road path between the first transportation hub point and the second transportation hub point;

Obtaining a center of a minimum circle covering the shortest road path between the first transportation hub point and the second transportation hub point;

Connecting the center of the circle with the midpoint of the first transportation hub point and the second transportation hub point as a boundary between the first transportation hub point and the second transportation pivot point;

Loop through all the traffic hub points to get the shortest path polygon.
The area dividing method according to claim 1, wherein the obtaining the preliminary area dividing result comprises:

Obtaining an administrative area of a minimum administrative unit in a rectangular boundary of the area to be divided;

And intersecting the administrative region with the region of the shortest path polygon to generate the preliminary region dividing result.
The area dividing method according to claim 1, wherein the feature vector calculation parameter comprises position information, a radius of radiation, and a topic weight; and the calculating the feature vector of each point of interest in each region in the preliminary region division result comprises:

Obtain the subject radiation range and theme weight of the point of interest;

Statistics on the number of points of interest for each topic in each region;

Statistics acquire the feature vectors of all points of interest in each region.
The area dividing method according to any one of claims 1 to 5, further comprising: before calculating a feature vector of each point of interest in each area in the preliminary area dividing result,

Setting a threshold number of regions to be divided into regions;

Determining whether the number of regions in the preliminary region division result is greater than the threshold number of the region;

If greater than, calculating feature vectors of each interest point in each region in the preliminary region division result;

If not greater than, the preliminary region division result is used as the final region division result.
The area dividing method according to claim 6, wherein the merging the areas in the preliminary area dividing result according to the area similarity, and generating the final area dividing result comprises:

Setting an area threshold of the area to be divided;

Selecting two regions having the largest regional similarity in each region and the area and the area not exceeding the area threshold, and combining the two regions into one region;

Determining whether the number of merged regions is less than a threshold number of the regions;

If the number of the merged regions is not greater than the threshold number of the regions, the partitioning region task ends, and the final region partitioning result is output;

If the number of merged regions is greater than the number of regions threshold, then the merge is continued through the points of interest.
A region dividing device includes: a boundary module, a point of interest module, and an optimization module, wherein

The boundary module is configured to obtain a boundary of the area to be divided, calculate a rectangular boundary of the area to be divided, obtain position information of the transportation pivot point in the area to be divided, and draw the shortest path polygon by using the transportation pivot point as a discrete point; The shortest path polygon is combined with the rectangular boundary of the to-be-divided area to obtain a preliminary area division result, where the preliminary area division result includes multiple areas;

The point of interest module is configured to acquire and calculate a parameter according to a feature vector of a point of interest in the area to be divided, and calculate a feature vector of each point of interest in each area in the preliminary area division result;

The optimization module is configured to calculate a region similarity between each region and a spatial neighboring region according to the feature vector of each interest point in each region; and partition the region in the preliminary region according to the region similarity Merge to generate final zoning results.
The area dividing apparatus according to claim 8, wherein the boundary module is configured to acquire a location data set of all traffic stations in the area to be divided, and merge location data of a plurality of traffic stations having a predetermined association relationship, to obtain the Location information of a transportation hub point; the transportation site includes at least one of a bus stop, a subway station, a dock, and an airport.
The area dividing device according to claim 8, wherein the boundary module is configured to acquire a shortest road path between the first transportation junction point and the second transportation pivot point; and acquire the first transportation hub point and the second traffic a center of a minimum circle of the shortest road path between the pivot points; a connection between the center of the circle and the midpoint of the first transportation hub point and the second transportation hub point as a first transportation hub point and a second transportation hub point The boundary is processed by looping through all of the traffic hub points to obtain the shortest path polygon, the shortest path polygon being the path-Tyson polygon.
The area dividing device according to claim 8, wherein the boundary module is configured to acquire an administrative area of a minimum administrative unit among the rectangular boundaries of the area to be divided; and the area in the administrative area and the shortest path polygon The intersection is taken to generate the preliminary region division result.
The area dividing apparatus according to claim 8, wherein the feature vector calculation parameter comprises position information, a radius of radiation, and a topic weight; the point of interest module is configured to acquire a subject radiation range and a topic weight of the point of interest; The number of points of interest for each topic in the region; statistically obtain the feature vectors of all points of interest in each region.
The area dividing apparatus according to any one of claims 8 to 12, wherein the optimization module is further configured to set a to-be-divided area before calculating a feature vector of each point of interest in each area in the preliminary area division result. a threshold of the number of regions; determining whether the number of regions in the preliminary segmentation result is greater than the threshold of the number of regions; if greater than, calculating a feature vector of each point of interest in each region of the preliminary region segmentation result; if not greater than Then, the preliminary region division result is used as a final region division result.
The area dividing device according to claim 13, wherein the optimization module is configured to set an area threshold of the area to be divided; and select the area similarity in each area to be the largest and the area of the area is not larger than the area of the area. The two regions of the threshold are merged into one region; the number of the merged regions is determined to be smaller than the threshold number of the region; if the number of merged regions is not greater than the threshold for the number of regions, the task of dividing the region ends, and the output is ended. The result of the final region division; if the number of the merged regions is greater than the threshold for the number of regions, the merge is continued through the points of interest.
A computer storage medium having stored therein computer executable instructions for performing the method of any one of claims 1 to 7.