CN111382330A

CN111382330A - Land property identification method and device, electronic equipment and storage medium

Info

Publication number: CN111382330A
Application number: CN202010164099.4A
Authority: CN
Inventors: 崔巍; 闫嘉; 张岩; 李振军; 王子芃; 张楠; 史红欣; 项波
Original assignee: Smartsteps Data Technology Co ltd
Current assignee: Smartsteps Data Technology Co ltd
Priority date: 2020-03-10
Filing date: 2020-03-10
Publication date: 2020-07-07

Abstract

The application provides a land property identification method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a plurality of POI data corresponding to an area to be identified; carrying out grid division on an area to be identified to obtain a plurality of corresponding block grids; marking the corresponding block grids according to POI data contained in each block grid to obtain first user properties of each block grid; obtaining a first human stream density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human stream density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model; and obtaining the target land property of the corresponding block grid according to the first land property and the second land property. According to the method and the device, the accuracy of recognizing the land property of the area to be recognized can be improved by adding the pedestrian flow density curve and the recognition model on the basis of POI data.

Description

Land property identification method and device, electronic equipment and storage medium

Technical Field

The application relates to the technical field of data processing, in particular to a land property identification method and device, electronic equipment and a storage medium.

Background

The method and the device for identifying the land property of the city are beneficial to better understanding the spatial structure of the city, and lay a foundation for analyzing the difference between the city development and the city planning in reality and determining the reasonable planning direction.

In the prior art, identification of land property of a city is usually performed by using a point of interest (POI), that is, counting the number of different types of POIs in each area in the city, and using the type with the largest number of POIs as the land property of the area. Due to the fact that POI data are updated late and not all sides, the land property identification through the POI is inaccurate.

Disclosure of Invention

An embodiment of the application aims to provide a land property identification method, a land property identification device, electronic equipment and a storage medium, which are used for solving the problem that the land property identification is inaccurate in the prior art.

In a first aspect, an embodiment of the present application provides a method for identifying land property, including: acquiring a plurality of POI data corresponding to an area to be identified; carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid; obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model; and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

According to the method and the device, the POI data are firstly utilized to preliminarily identify the area to be identified, the first land property of each block grid is obtained, then the identification model is utilized to analyze the pedestrian flow density curve of the block grid, the second land property is obtained, the second land property is utilized to modify the first land property of the corresponding block grid, the final target land property is obtained, and the accuracy of identifying the land property of the area to be identified is improved.

Further, the grid division of the area to be identified to obtain a plurality of corresponding block grids includes: obtaining a remote sensing image corresponding to the area to be identified; and acquiring streets with preset width based on the remote sensing images, and performing grid division on the area to be identified by using the streets with the preset width to obtain a plurality of block grids.

According to the embodiment of the application, the area to be recognized is better met by utilizing the street with the preset width.

Further, labeling the corresponding street grids according to the POI data corresponding to each street grid, and obtaining the first geographic property of each street grid, including: dividing a plurality of POI data corresponding to the area to be identified into special POI data and non-special POI data; if the block grids contain special POI data, marking the corresponding block grids according to the special POI data to obtain first geographical properties of each block grid; acquiring the weight of non-special POI data contained in the non-labeled block grid, and determining the first place property of the corresponding non-labeled block grid according to the weight of the non-special POI data; and the unmarked neighborhood grids are neighborhood grids in the area to be identified except for the neighborhood grids marked as the first place property.

In the embodiment of the application, because the influence of the special POI data is large, the corresponding block grids are marked by using the special POI data, the block grids without the special POI data are marked by adopting the weight of the non-special POI data, and the efficiency of marking the block grids is improved.

Further, before classifying the POI data corresponding to the area to be identified, the method further includes: and carrying out noise elimination on a plurality of POI data corresponding to the area to be identified. According to the POI data eliminating method and device, the POI data with small influence in the area to be recognized are eliminated, the subsequent calculated amount is reduced, and therefore recognition efficiency is improved.

Further, the determining a first geographic property of a corresponding neighborhood grid from the special POI data comprises: if the neighborhood grid contains one piece of special POI data, taking the property of land corresponding to the special POI data as a first property of land corresponding to the neighborhood grid; if the neighborhood grid contains a plurality of special POI data, acquiring the priority corresponding to each special POI data, and taking the property of the special POI data with the highest priority as the first property of the neighborhood corresponding to the neighborhood grid.

The higher the priority of the special POI data in the embodiment of the application is, the greater the influence of the special POI data is, so that the property of the land corresponding to the special POI data with the highest priority is used as the first property of the land grid, so that the marking is more accurate.

Further, the determining a first geographic property of a corresponding neighborhood grid according to the weight of the non-special POI data includes: and taking the land property corresponding to the non-special POI data with the largest weight as the first land property of the corresponding block grid.

In the embodiment of the application, the larger the weight of the non-specific POI data is, the larger the influence on the property of the land is, so that the property of the land corresponding to the POI data with the largest weight is used as the first property of the neighborhood grid, and the marking is more accurate.

Further, after determining the first geographic property of the corresponding neighborhood grid from the special POI data, the method further comprises: and if the number of the POIs of the neighborhood grids marked as the first geographical properties is greater than a first preset value and/or the POI density of the neighborhood grids is greater than a second preset value, deleting the first geographical properties corresponding to the neighborhood grids and marking as unmarked neighborhood grids.

According to the method and the device, the block grids with the POI quantity larger than the first preset value and/or the POI density larger than the second preset value are marked as the unmarked block grids so as to be marked again, and the marking accuracy of the block grids is improved.

Further, before obtaining a people flow density curve corresponding to the block grid with the first land property being the preset land property, the method further comprises: acquiring signaling data of mobile terminal users corresponding to a plurality of training block grids in a historical time period, and generating a corresponding second people flow density curve according to the signaling data of each training block grid; training the recognition model by using a second people stream density curve and a label corresponding to each training block grid to obtain a trained recognition model; the labels include third ground properties corresponding to the training neighborhood grid.

According to the method and the device, the second geological property corresponding to the block grid can be accurately predicted by utilizing the recognition model obtained by training the second people flow density curve and the corresponding third geological property.

Further, the obtaining a first human current density curve corresponding to a block grid with a first land property being a preset land property includes: acquiring signaling data of a mobile terminal user within a preset time period of a neighborhood grid with a first geographical property as a preset geographical property; and generating a corresponding first human current density curve according to the signaling data.

Because the signaling data of the mobile terminal user can well reflect the people flow data, a relatively objective people flow density curve can be generated according to the signaling data, and the second place property corresponding to the block grid can be accurately identified through the identification model.

Further, the obtaining the target land property of the corresponding neighborhood grid according to the first land property and the second land property comprises: if the first geological property is the same as the second geological property, the geological property of the target land is the first geological property or the second geological property; and if the first geological property is different from the second geological property, the target geological property is the second geological property.

According to the method and the device, on the basis of identifying the land property of the block grid by using POI data, the second land property obtained by analyzing the pedestrian flow density curve of the block grid by using the identification model is corrected, so that the accuracy of identifying the land property of the land can be improved.

In a second aspect, an embodiment of the present application provides another land property identification method, including: acquiring a region to be identified; carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; and acquiring a people flow density curve corresponding to the block grid with the first geographical property as the preset geographical property, inputting the people flow density curve into a pre-trained recognition model, and acquiring a fourth geographical property output by the recognition model.

In the embodiment of the application, the pedestrian flow density curves corresponding to the block grids with different land properties are different, so that the pedestrian flow density curves of the block grids are analyzed through the recognition model to obtain the fourth land property corresponding to the block grids, and the accuracy of recognizing the land property of the land is improved.

Further, the obtaining of the people flow density curve corresponding to the block grid with the first land property being the preset land property includes: acquiring signaling data of a mobile terminal user within a preset time period of a neighborhood grid with a first geographical property as a preset geographical property; and generating a corresponding first human current density curve according to the signaling data.

In a third aspect, an embodiment of the present application provides a land property identification device, including: the information acquisition module is used for acquiring a plurality of POI data corresponding to the area to be identified; the first grid division module is used for carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; the marking module is used for marking the corresponding block grids according to POI data contained in each block grid to obtain first place properties of each block grid; the first identification module is used for obtaining a first human stream density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human stream density curve into a pre-trained identification model and obtaining a second land property output by the identification model; and the correction module is used for obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

In a fourth aspect, an embodiment of the present application provides another land property identification device, including: the image acquisition module is used for acquiring a remote sensing image corresponding to the area to be identified; the second meshing module is used for meshing the area to be identified based on the remote sensing image to obtain a plurality of corresponding block meshes; and the second identification module is used for obtaining a people stream density curve corresponding to the block grid with the first land property being the preset land property, inputting the people stream density curve into a pre-trained identification model, and obtaining a fourth land property output by the identification model.

In a fifth aspect, an embodiment of the present application provides an electronic device, including: the system comprises a processor, a memory and a bus, wherein the processor and the memory are communicated with each other through the bus; the memory stores program instructions executable by the processor, the processor invoking the program instructions to be capable of performing the method of the first or second aspect.

In a sixth aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium, including: the non-transitory computer readable storage medium stores computer instructions that cause the computer to perform the method of the first or second aspect.

Additional features and advantages of the present application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the embodiments of the present application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.

Fig. 1 is a schematic view illustrating a process of identifying land property according to an embodiment of the present application;

FIG. 2 is a schematic view of a pedestrian flow density curve provided in an embodiment of the present application;

fig. 3 is a schematic diagram of a block grid obtained after division according to an embodiment of the present application;

FIG. 4 is a diagram illustrating a labeled result according to an embodiment of the present disclosure;

fig. 5 is a schematic view of POI distribution in a neighborhood grid according to an embodiment of the present disclosure;

fig. 6 is a second people flow density curve corresponding to the office-like block grid according to the embodiment of the present application;

fig. 7 is a second people flow density curve corresponding to the business-class block grid according to the embodiment of the present application;

FIG. 8 is a diagram illustrating model accuracy results provided by embodiments of the present application;

FIG. 9 is a schematic flow chart of another method for identifying a right-of-way property according to an embodiment of the present disclosure;

FIG. 10 is a schematic illustration of another annotation result provided in the examples of the present application;

FIG. 11 is a flowchart illustrating a method for identifying a geographical property according to an embodiment of the present disclosure;

fig. 12 is a schematic structural diagram of an identification device according to an embodiment of the present application;

fig. 13 is a schematic structural diagram of another identification device according to an embodiment of the present disclosure;

fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.

Fig. 1 is a schematic view of a process for identifying land property provided in an embodiment of the present application, and as shown in fig. 1, an execution main body of the identification method is an identification device, and the identification device may be an intelligent electronic device such as a desktop computer, a tablet computer, a notebook computer, a smart phone, and an intelligent wearable device, and the method includes:

step 101: acquiring a plurality of POI data corresponding to an area to be identified;

step 102: carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids;

step 103: marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid;

step 104: obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model;

step 105: and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

In step 101, the area to be identified may be a city, a certain administrative district in a city, a county or a certain area designated according to the requirement. POI are generally represented by bubble icons on an electronic map, such as scenic spots, government agencies, companies, shopping malls, restaurants and the like on the electronic map, which are all POI. One piece of POI data includes at least a name, genre, address, and latitude and longitude, where the genre indicates whether the POI is a sight, a government agency, a company, a mall, a restaurant, or the like. It will be appreciated that POI data to the area to be identified may be obtained via a third party map.

In step 102, since the area of the entire region to be identified is large, in order to improve the identification of the geographic property of the region to be identified, the area of the region to be identified may be gridded to obtain a plurality of relatively small street grids. It should be noted that, when performing grid division, the area to be identified may be divided into a plurality of block grids according to a preset size, or may be divided according to streets in the area to be identified.

In addition, the selection of the preset size or the street may be limited according to the actual situation, taking the preset size as an example, if the preset size is too small, a plurality of block grids may be obtained by division, which may cause a problem of large workload in the subsequent identification of the block grids, and a too small block grid may not include the POI, which may result in that the block grid cannot be identified, and the purpose of statistical combing the land property in the area to be identified may not be achieved. If the size of the block grid is too large, areas in the block grid may include multiple right-of-way properties, and the block grid may be inaccurately identified because the right-of-way properties are set for the block grid.

In step 103, in the embodiment of the present application, the right-of-way property of the area to be identified is divided into four types, which are: business class, residential class, service class, and office class. Although the types in the POI data include many kinds, each can correspond to one of the above four types. Therefore, the corresponding block grid can be labeled according to the POI data included in each block grid, and it can be understood that the labeling is to determine the first geographic property corresponding to the block grid, and the first geographic property of one block grid is one of the four geographic properties. It should be noted that the specific types of dividing the region to be identified may be more than the above four types, and the division may also be performed with finer granularity. Or, for the area to be identified, if it is known in advance that there is no area with certain land property in the area to be identified, for example, there is no area with a residential class, the land property of the area to be identified is classified into a business class, a service class and an office class. The specific classification of the region to be identified may be determined according to actual conditions, and this is not specifically limited in the embodiment of the present application.

In step 104, the first people flow density curve refers to the people flow situation in the corresponding block grid within a certain time period. As shown in fig. 2, the abscissa represents the time from 0 to 23 points in a day, and the ordinate represents the number of people in the neighborhood network at the corresponding time. It should be noted that the time period corresponding to the first human flow density curve may be one day, two days, or one week, and may be determined specifically according to actual situations.

The inventor observes and summarizes the pedestrian flow density curve for a long time to obtain the incidence relation between the pedestrian flow density curve and the land property, namely, different land properties correspond to different pedestrian flow density curves. Therefore, the recognition model can be trained in advance, the first human flow density curve corresponding to the block network is input into the recognition model, and the recognition model can output the corresponding second terrain property. It is understood that the nature of the predetermined right of way may be the commercial, residential, service, and office categories mentioned above, or any one or more of the four categories mentioned above. If the predetermined property of the right of way is a business class, a residential class, and an office class, then this step only identifies the second property of the right of way for a block grid where the first property of the right of way is a business class, a residential class, and an office class.

In step 105, after obtaining the first and second geographic properties of the neighborhood grid, a target geographic property of the corresponding neighborhood grid is determined based on the first and second geographic properties. And the target land property is a final land property corresponding to the block grid.

When obtaining the target site property of the corresponding neighborhood grid according to the first site property and the second site property, if the first site property and the second site property are the same, the target site property is the first site property or the second site property, for example: the first land property and the second land property are both of a residential type, and then the target land property is also of a residential type. If the first and second geocharacteristics are different, then the first geocharacteristics are modified using the second geocharacteristics, for example: the first land property is a residential class, the second land property is a commercial class, and then the target land property is a commercial class according to the second land property.

According to the method and the device for identifying the target land property, the POI data are firstly utilized to preliminarily identify the area to be identified, the first land property of each block grid is obtained, then the people flow density curve of the block grid is analyzed by utilizing the identification model, the second land property is obtained, the first land property of the corresponding block grid is modified by utilizing the second land property, and the final target land property is obtained. Therefore, the pedestrian flow density curve and the recognition model are added on the basis of the POI data, the first land property can be corrected more quickly, and the accuracy of recognizing the land property of the area to be recognized is improved.

On the basis of the above embodiment, the grid division of the area to be identified to obtain a plurality of corresponding street grids includes:

obtaining a remote sensing image corresponding to the area to be identified;

and acquiring streets with preset width based on the remote sensing images, and performing grid division on the area to be identified by using the streets with the preset width to obtain a plurality of block grids.

The remote sensing image corresponding to the area to be identified can be obtained by shooting through a remote sensing satellite, and then the shot remote sensing image is transmitted to the identification device, and in addition, the remote sensing image of the area to be identified uploaded by other people can also be obtained from a network. The embodiment of the application does not limit the specific mode of obtaining the remote sensing image.

The method and the device for obtaining the street information of the area to be identified can obtain the street information of the area to be identified in the remote sensing image, and the embodiment of the application utilizes the street with the preset width to divide the grid of the area to be identified to obtain a plurality of block grids. Where, in general, streets within a city may be divided into 3 meters, 6 meters, 8 meters and 12 meters, possibly with narrower or wider streets. And dividing the areas to be identified by adopting streets with different widths, wherein the obtained street grids are different in size. If the adopted street width is too narrow, the obtained street grids are more and small, the subsequent calculation amount is increased, and the statistics and the arrangement of the land use property of the area to be identified cannot be realized. If the street width used is too wide, the obtained block grid is small and large, and too large granularity can cause the problem of inaccurate land property identification of the block grid. Therefore, a suitable street width may be selected according to the specific situation of the area to be identified, for example, a street having a street width of 3 meters may be selected for division. Fig. 3 is a schematic diagram of a block grid obtained after division according to an embodiment of the present application.

In another embodiment, if the area to be identified includes physical barriers such as railways, rivers, mountains and the like, the area to be identified can be divided together with streets with preset widths, so that the accuracy of dividing the street grids is higher, and the method has universality.

Because the traditional grid division rule and the complicated and changeable land types have natural contradictions, the natural block division courtyards based on the street level have higher consistency of the internal land properties of each block grid, and the anti-interference performance is strong in the subsequent analysis link.

On the basis of the above embodiment, labeling a corresponding neighborhood grid according to POI data corresponding to each neighborhood grid to obtain a first geographic property of each neighborhood grid, includes:

dividing a plurality of POI data corresponding to the area to be identified into special POI data and non-special POI data;

if the block grids contain special POI data, marking the corresponding block grids according to the special POI data to obtain first geographical properties of each block grid;

acquiring the weight of non-special POI data contained in the non-labeled block grid, and determining the first place property of the corresponding non-labeled block grid according to the weight of the non-special POI data; and the unmarked neighborhood grids are neighborhood grids in the area to be identified except for the neighborhood grids marked as the first place property.

In a specific implementation process, the types of POIs belonging to the special POI data are preset, and the special POI data are non-special POI data except the special POI data. Generally, the special POI data is too weighted, and once the special POI data is included in a neighborhood grid, the property of the neighborhood grid is the property of the corresponding place of the special POI data. For example, a POI of the university type has a large corresponding area, which results in a large influence on the street grid where the POI is located, i.e., a large weight.

The special POI data mainly comprises a service class and a residence class; the special POI data has corresponding priority, the first-level priority is a service class, and the priority is highest, wherein the service class comprises comprehensive Hospital, colleges and universities, national and provincial scenic spots and the like; the second-level priority is the residential category, including house number information and the like, and the priority level is the second level; the third level of priority is also a service class, including elementary school, middle school, etc. It should be noted that which POI data are empirically determined in advance as the special POI data, and the special POI data may be divided into more priority levels, and the types of POIs included in each priority level are also empirically determined in advance.

Therefore, the POI data in the area to be identified can be divided into special POI data and non-special POI data.

If the special POI data are included in the neighborhood grid, the first place property of the corresponding neighborhood grid is determined according to the special POI data because the influence of the special POI data on the place property is larger than that of the non-special POI data. And the neighborhood grid labeled with the first locality property may be referred to as a labeled neighborhood grid, with the remainder being unlabeled.

While the unlabeled neighborhood grid is mostly a neighborhood grid containing only non-specific POI data, there may also be a neighborhood grid that has been labeled as first place property, but is reset. Which street grids are reset are explained in the following embodiments. For an unlabeled neighborhood grid, the non-special POI data contained in the neighborhood grid is used to determine the first geographic property corresponding to the neighborhood grid.

It should be noted that, if the area to be identified does not include special POI data, the corresponding neighborhood grids may be directly labeled with the non-special POI data corresponding to each neighborhood grid, so as to obtain the first geographical property of the neighborhood grids.

On the basis of the above embodiment, some POI types having a small influence on land use properties may be included in the area to be identified, for example: the system comprises road auxiliary facilities, transportation facility services, public facilities, motorcycle services, automobile maintenance, event activities, indoor facilities, traffic facilities and the like, and can also comprise other POI types, and the POI types can be preset to have small influence according to actual conditions. Therefore, in order to reduce the calculation workload, POI data belonging to the POI type is removed as noise.

According to the POI data eliminating method and device, the POI data with small influence in the area to be recognized are eliminated, the subsequent calculated amount is reduced, and therefore recognition efficiency is improved.

On the basis of the above embodiment, the determining a first geographic property of a corresponding neighborhood grid according to the special POI data includes:

if the neighborhood grid contains one piece of special POI data, taking the property of land corresponding to the special POI data as a first property of land corresponding to the neighborhood grid;

if the neighborhood grid contains a plurality of special POI data, acquiring the priority corresponding to each special POI data, and taking the property of the special POI data with the highest priority as the first property of the neighborhood corresponding to the neighborhood grid.

In a specific implementation process, the number of special POI data that may be included in some neighborhood grids is only one, and the number of special POI data that may be included in some neighborhood grids may be multiple. Then, for a neighborhood grid containing a specific POI data, the property of right of way corresponding to the specific POI data is used as the first property of right of way corresponding to the neighborhood grid. For a neighborhood grid containing a plurality of special POI data, the property of right of way corresponding to the special POI data with the highest priority can be used as the first property of right of way of the neighborhood grid. For example, two special POI data are included in a block grid, namely a university and a cell, and since the university has higher priority than the cell, the service class corresponding to the university is used as the first place property of the block grid. FIG. 4 is a diagram illustrating the labeled results according to an embodiment of the present invention, as shown in FIG. 4, the block grid filled with left-diagonal lines representing the business class; block grids filled with right slashes and representing office classes; a block grid representing a house class filled with a diagonal grid; the rectangular grid is populated with a street grid representing a class of service.

In another embodiment, for a neighborhood grid containing a plurality of special POI data, the first geographic property of the neighborhood grid can be determined according to the number, and if the number is the same, the first geographic property of the neighborhood grid can be determined by using the priority. For example: the first place property of a neighborhood grid is a residential category because the number of POI data of residential cells is greater than that of the university, including POI data of 1 university and POI data of 3 residential cells. If the neighborhood grid comprises POI data of 1 university and POI data of 1 residential district, the number of the POI data of the service class and the number of the POI data of the residential class are the same, and the first place property of the neighborhood grid is the service class because the priority of the place property of the service class corresponding to the university is higher than the priority of the residential class of the residential district.

In another embodiment, the first locality property of the neighborhood grid may also be determined in terms of area, for example, the neighborhood grid includes POI data for 1 university, POI data for 5 residential cells, and POI data for 1 elementary school; schools belong to the service class, residential communities belong to the residential class, the total area of universities and primary schools is 4 square kilometers, the total area of 5 residential communities is 3.5 square kilometers, and the area of a school is larger than that of a residential community, so that the first place property of the street grid is the service class.

On the basis of the above embodiment, after determining the first locality property of the corresponding neighborhood grid from the special POI data, the method further comprises:

and if the number of the POIs of the neighborhood grids marked as the first geographical properties is greater than a first preset value and/or the POI density of the neighborhood grids is greater than a second preset value, deleting the first geographical properties corresponding to the neighborhood grids and marking as unmarked neighborhood grids.

In a specific implementation process, after a plurality of experiments, the inventor finds that, for the block grids marked with the first geographical properties, the block grids belonging to the business class and the office class have a clustering effect, that is, the number and/or density of POIs in one block grid are abnormally high, as shown in fig. 5, the black dots represent POIs, and it can be seen from fig. 5 that there are more POIs in several block grids than in other block grids. It is understood that the number of POIs refers to the number of POIs in a neighborhood grid, and the POI density refers to the number of POIs in a neighborhood grid and the area of the neighborhood grid. If the number of the POIs is greater than the first preset value and/or the POI density is greater than the second preset value, the block grid is considered to be abnormal, the marked first place property needs to be deleted and marked as an unmarked block grid, so that the block grid can be marked again in the following process, and the accuracy of marking the block grid is improved.

It can be understood that both the first preset value and the second preset value can be determined according to actual situations, and this is not specifically limited in the embodiment of the present application.

On the basis of the above embodiment, the determining a first geographic property of a corresponding neighborhood grid according to the weight of the non-special POI data includes:

and taking the land property corresponding to the non-special POI data with the largest weight as the first land property of the corresponding block grid.

In a specific implementation process, weights corresponding to the POI data are preset, and the influence on the block grid is different due to different weights.

The following table shows weights corresponding to some POI types provided in the embodiment of the present application, and it can be understood that the weights corresponding to the POI types may be adjusted according to actual situations, and weights of POI types possibly corresponding to different areas to be identified may be different, which is not specifically limited in the embodiment of the present application.

For the non-labeled neighborhood grid, the POI data with the maximum weight in the neighborhood grid is obtained through the preset weight corresponding to the POI data, and the land property corresponding to the POI data with the maximum weight is used as the first land property of the neighborhood grid.

It should be noted that, in the above embodiment, the labeling is performed according to the special POI data, and then the block grid that is not labeled is reset, and when the block grid is labeled again, the special POI data in the block grid needs to be removed, and only the non-special POI data is weighted the most.

In another embodiment, when the non-labeled neighborhood grid is labeled, the sum of the weights of the non-specific POI data belonging to the same property of land in the neighborhood grid can be counted, and the property of land with the largest sum of the weights can be used as the first property of land corresponding to the neighborhood grid.

In another embodiment, when the unmarked neighborhood grid is marked, marking can be performed according to the number of POIs, that is, the number of POIs belonging to the same type of property in the neighborhood grid is counted, and the property of the site of the type with the largest number of POIs is used as the first property of the neighborhood grid.

In another embodiment, for the labeling of the unlabeled neighborhood grid, the labeling can be performed according to the area occupied by the POI, for example: taking the property of the land corresponding to the POI with the largest occupied area in the block grid as the first property of the block grid; or counting the total area occupied by POI belonging to the same type of land property in the block grid, and taking the land property with the maximum total area as the first land property of the block grid.

On the basis of the above embodiment, before obtaining the people flow density curve corresponding to the block grid with the first land property being the preset land property, the method further includes:

acquiring signaling data of mobile terminal users corresponding to a plurality of training block grids in a historical time period, and generating a corresponding second people flow density curve according to the signaling data of each training block grid;

training the recognition model by using a second people stream density curve and a label corresponding to each training block grid to obtain a trained recognition model; the labels include third ground properties corresponding to the training neighborhood grid.

In a specific implementation process, the training neighborhood grid refers to a neighborhood grid labeled with the property of the right place, and may be a neighborhood grid except for an area to be identified, and the labels corresponding to the training neighborhood grid may be obtained by manually labeling the neighborhood grid, and it can be understood that the labels are the property of the third place.

The historical time period can be one day, two days or one week, and can also be determined according to actual conditions. When the mobile terminal is powered on, signaling data may be continuously exchanged with a nearby base station, where the signaling data includes time information and location information, and of course, the signaling data may also include other data, which is not specifically limited in this embodiment of the present application.

After the signaling data of the mobile terminal users corresponding to the training neighborhood grids in the historical time period are obtained, a corresponding second people stream density curve can be generated according to the time information and the position information in the signaling data. Fig. 6 is a second people flow density curve corresponding to the office-class neighborhood grid provided in the embodiment of the present application, and fig. 7 is a second people flow density curve corresponding to the business-class neighborhood grid provided in the embodiment of the present application. Fig. 2 is a man current density curve of a house, and as can be seen from fig. 2, the man current density curve waveform of the house shows a form in which the peak is a double peak and the rear peak is higher than the front peak. As can be seen from fig. 6, the office class traffic density curve is generally unimodal and peak forward morphology. As can be seen from fig. 7, the stream density curve of the commercial type is a form in which the peak is not obvious, but the whole broken line is high in a large period of time. It will be appreciated that after obtaining the signaling data, a second people flow density curve, such as numpucy and pandas and matplotlib, may be obtained by drawing through some of the packets in python.

The recognition model may adopt a Convolutional Neural Network (CNN), which is a kind of feed forward Neural Network (fed Neural Networks) containing convolution calculation and having a deep structure, and is one of the representative algorithms of deep learning (deep learning). Convolutional neural Networks have a feature learning (representation) capability, and can perform Shift-Invariant classification (Shift-Invariant classification) on input information according to a hierarchical structure thereof, and are also called Shift-Invariant artificial neural Networks (SIANN).

The CNN comprises 3 layers of convolution pooling layers, 2 layers of full connection layers and an activation function ReLU, and a dropout function and a softmax function are adopted as classifiers. Wherein, the convolutional layer: given a new image, CNN does not know exactly which parts of the original image these features match, so it tries every possible position in the original image, which is equivalent to turning the feature into a filter. This process for matching is known as convolution. A pooling layer: to effectively reduce the computational load, another effective tool used by CNN is called "Pooling (Pooling)". Pooling is to reduce the size of the input image, reduce pixel information, and only retain important information. Full connection layer: the full-connection layer plays a role of a classifier in the whole convolutional neural network, namely, the result is identified and classified through the full-connection layer after the deep network such as convolution, activation function, pooling and the like. A classifier: the function of the activation function is to add a nonlinear factor and to perform nonlinear mapping on the convolution layer output result.

It should be noted that, in addition to the above structure, the recognition model may also be a CNN with other structures, and may also be other kinds of networks, for example: a support vector machine, a Long Short Term memory network (LSTM), etc.

After the recognition model is trained, the accuracy of the recognition model can be tested, and fig. 8 is a model accuracy result diagram provided by the embodiment of the application, and after multiple times of training, the accuracy of the recognition model is averagely more than 85%.

Fig. 9 is a schematic flow chart of another method for identifying a right-of-way property according to an embodiment of the present application, as shown in fig. 9, the method includes:

the method comprises the steps that firstly, signaling data of a plurality of training block grids in a historical time period are obtained from a database, and a signaling data table is obtained, wherein the signaling data corresponding to each training block grid corresponds to one signaling data table;

secondly, generating an original image according to the signaling data in the signaling data table, wherein the original image comprises a second people stream density curve;

and thirdly, preprocessing the original image, wherein the preprocessing comprises dimensionality reduction and abbreviation, the original image is 4-dimensional data, the data volume of the original image is large, and the original image can be reduced into a 3-dimensional image in order to reduce the calculation amount of the identification model. The reduction is to reduce the number of pixels of the original image, and the purpose is to reduce the calculation amount of the recognition model.

And fourthly, training the recognition model by utilizing the preprocessed image and the third geographical property of the training block grid corresponding to the image to obtain the trained recognition model. The training process comprises the following steps: inputting the preprocessed image into the CNN, outputting a prediction result by the CNN, calculating a loss value according to the prediction result and a third ground property corresponding to the preprocessed image, and reversely optimizing parameters in the CNN by using the loss value until the loss value is smaller than a preset value or reaches a preset iteration number.

And fifthly, acquiring signaling data of a mobile terminal user within a preset time period of the block grid with the first geographical property being the preset geographical property, generating a corresponding first human flow density curve according to the signaling data, inputting the first human flow density curve into a trained recognition model, and outputting a prediction result, namely the second geographical property corresponding to the block grid by the recognition model. Fig. 10 is a schematic diagram of another labeling result provided in the embodiment of the present application, as shown in fig. 10, after the first geographic property is corrected by the second geographic property, it can be found that the block grid originally labeled as the residential block is corrected to the office block, and some other block grids labeled with errors before are also corrected in time.

In addition, after the first place property and the second place property corresponding to each block grid in the area to be identified are obtained, the labels of the first place property and the second place property can be verified manually.

Fig. 11 is a schematic flowchart of a method for identifying a right-of-way property according to an embodiment of the present application, and as shown in fig. 11, the method includes:

step 1101: acquiring a region to be identified;

step 1102: carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids;

step 1103: and acquiring a people flow density curve corresponding to the block grid with the first geographical property as the preset geographical property, inputting the people flow density curve into a pre-trained recognition model, and acquiring a fourth geographical property output by the recognition model.

In step 1102, the method for performing mesh division on the region to be identified is the same as the above embodiment, and is not described herein again.

In step 1103, the first people flow density curve refers to the people flow situation in the corresponding block grid within a certain time period. As shown in fig. 2, the abscissa represents the time from 0 to 23 points in a day, and the ordinate represents the number of people in the neighborhood network at the corresponding time. It should be noted that the time period corresponding to the first human flow density curve may be one day, two days, or one week, and may be determined specifically according to actual situations.

The inventor observes and summarizes the pedestrian flow density curve for a long time to obtain the incidence relation between the pedestrian flow density curve and the land property, namely, different land properties correspond to different pedestrian flow density curves. Therefore, the recognition model may be trained in advance, the first human current density curve corresponding to the neighborhood network may be input to the recognition model, and the recognition model may output the corresponding fourth geographic property.

On the basis of the above embodiment, the obtaining a pedestrian flow density curve corresponding to a block grid with a first land property being a preset land property includes:

acquiring signaling data of a mobile terminal user within a preset time period of a neighborhood grid with a first geographical property as a preset geographical property;

and generating a corresponding first human current density curve according to the signaling data.

It should be noted that the obtaining manner of the first human flow density curve and the training process of the recognition model are consistent with the above embodiments, and are not described herein again.

In the embodiment of the application, as the people flow density curves with different land properties are different, the land properties of each block grid in the area to be identified can be accurately obtained by using the people flow density curves and the identification model.

Fig. 12 is a schematic structural diagram of an identification device according to an embodiment of the present application, where the identification device may be a module, a program segment, or a code on an electronic device. It should be understood that the identification device corresponds to the above-mentioned embodiment of the method of fig. 1, and can perform various steps related to the embodiment of the method of fig. 1, and the specific functions of the identification device can be referred to the above description, and the detailed description is appropriately omitted here to avoid redundancy. The identification device comprises: an information obtaining module 1201, a first mesh division module 1202, a labeling module 1203, a first identification module 1204, and a modification module 1205, where:

the information acquisition module 1201 is used for acquiring a plurality of POI data corresponding to an area to be identified; the first meshing module 1202 is configured to perform meshing on the area to be identified to obtain a plurality of corresponding street grids; the labeling module 1203 is configured to label the corresponding street grids according to the POI data included in each street grid, so as to obtain first geographic properties of each street grid; the first identification module 1204 is configured to obtain a first personal stream density curve corresponding to a neighborhood grid with a first geographic property being a preset geographic property, input the first personal stream density curve into a pre-trained identification model, and obtain a second geographic property output by the identification model; the modifying module 1205 is used for obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

On the basis of the foregoing embodiment, the first meshing module 1202 is specifically configured to:

obtaining a remote sensing image corresponding to the area to be identified;

On the basis of the foregoing embodiment, the labeling module 1203 is specifically configured to:

On the basis of the above embodiment, the recognition apparatus further includes a denoising module, configured to:

and carrying out noise elimination on a plurality of POI data corresponding to the area to be identified.

On the basis of the above embodiment, the identification apparatus further includes a reset module configured to:

On the basis of the above embodiment, the recognition apparatus further includes a first model training module, configured to:

On the basis of the foregoing embodiment, the first identifying module 1204 is specifically configured to:

On the basis of the foregoing embodiment, the modification module 1205 is specifically configured to:

if the first geological property is the same as the second geological property, the geological property of the target land is the first geological property or the second geological property;

and if the first geological property is different from the second geological property, the target geological property is the second geological property.

Fig. 13 is a schematic structural diagram of another identification device according to an embodiment of the present application, where the identification device may be a module, a program segment, or a code on an electronic device. It should be understood that the identification device corresponds to the above-mentioned embodiment of the method of fig. 11, and can perform the steps related to the embodiment of the method of fig. 11, and the specific functions of the identification device can be referred to the description above, and the detailed description is appropriately omitted here to avoid redundancy. The identification device comprises: an image acquisition module 1301, a second meshing module 1302, and a second identification module 1303, wherein:

the image obtaining module 1301 is used for obtaining a region to be identified; the second meshing module 1302 is configured to perform meshing on the area to be identified, so as to obtain a plurality of corresponding block meshes; the second identification module 1303 is configured to obtain a pedestrian flow density curve corresponding to the street grid with the first geographic property being the preset geographic property, input the pedestrian flow density curve into a pre-trained identification model, and obtain a fourth geographic property output by the identification model.

On the basis of the foregoing embodiment, the second identifying module 1303 is specifically configured to:

On the basis of the above embodiment, the recognition apparatus further includes a second model training module, configured to:

Fig. 14 is a schematic structural diagram of an entity of an electronic device provided in an embodiment of the present application, and as shown in fig. 14, the electronic device includes: a processor (processor)1401, a memory (memory)1402, and a bus 1403; wherein the content of the first and second substances,

the processor 1401 and the memory 1402 communicate with each other via the bus 1403;

the processor 1401 is configured to invoke the program instructions in the memory 1402 to perform the methods provided by the above-mentioned method embodiments, for example, including: acquiring a plurality of POI data corresponding to an area to be identified; carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid; obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model; and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

Processor 1401 may be an integrated circuit chip having signal processing capabilities. The processor 1401 may be a general-purpose processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; but may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. Which may implement or perform the various methods, steps, and logic blocks disclosed in the embodiments of the present application. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The Memory 1402 may include, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Read Only Memory (EPROM), an electrically Erasable Read Only Memory (EEPROM), and the like.

The present embodiment discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method provided by the above-mentioned method embodiments, for example, comprising: acquiring a plurality of POI data corresponding to an area to be identified; carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid; obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model; and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

The present embodiments provide a non-transitory computer-readable storage medium storing computer instructions that cause the computer to perform the methods provided by the above method embodiments, for example, including: acquiring a plurality of POI data corresponding to an area to be identified; carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids; marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid; obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model; and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.

In addition, units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

Furthermore, the functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A land property identification method is characterized by comprising the following steps:

acquiring a plurality of POI data corresponding to an area to be identified;

carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids;

marking the corresponding block grids according to POI data contained in each block grid, and obtaining first user properties of each block grid;

obtaining a first human flow density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human flow density curve into a pre-trained recognition model, and obtaining a second land property output by the recognition model;

and obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

2. The method according to claim 1, wherein the gridding the area to be identified to obtain a plurality of corresponding neighborhood grids comprises:

obtaining a remote sensing image corresponding to the area to be identified;

3. The method of claim 1, wherein labeling each street block grid according to the POI data corresponding to the street block grid to obtain the first locality property of each street block grid comprises:

4. The method of claim 3, wherein prior to classifying the POI data corresponding to the area to be identified, the method further comprises:

5. The method of claim 3, wherein determining the first user property of the corresponding neighborhood grid based on the special POI data comprises:

6. The method of claim 3, wherein determining the first geographic property of the corresponding neighborhood grid according to the weight of the non-special POI data comprises:

7. The method of claim 3, wherein after determining the first user property of the corresponding neighborhood grid from the special POI data, the method further comprises:

8. The method of claim 1, wherein prior to obtaining the current density profile corresponding to the neighborhood grid having the first geographic property being a predetermined geographic property, the method further comprises:

9. The method as claimed in claim 1, wherein the obtaining the first human current density curve corresponding to the block grid with the first land property being the preset land property comprises:

10. The method as claimed in any one of claims 1-9, wherein the obtaining the target right of way property of the corresponding neighborhood grid according to the first right of way property and the second right of way property comprises:

11. A land property identification method is characterized by comprising the following steps:

acquiring a region to be identified;

and obtaining a first personal stream density curve corresponding to the block grid with the first land property being the preset land property, inputting the first personal stream density curve into a pre-trained recognition model, and obtaining a fourth land property output by the recognition model.

12. The method as claimed in claim 11, wherein the obtaining the current density curve corresponding to the block grid with the first geographic property being the preset geographic property comprises:

13. The method of claim 11, wherein prior to obtaining the current density profile corresponding to the neighborhood grid having the first geographic property being a predetermined geographic property, the method further comprises:

14. A land property identification device, comprising:

the information acquisition module is used for acquiring a plurality of POI data corresponding to the area to be identified;

the first grid division module is used for carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids;

the marking module is used for marking the corresponding block grids according to POI data contained in each block grid to obtain first place properties of each block grid;

the first identification module is used for obtaining a first human stream density curve corresponding to a block grid with a first land property being a preset land property, inputting the first human stream density curve into a pre-trained identification model and obtaining a second land property output by the identification model;

and the correction module is used for obtaining the target land property of the corresponding block grid according to the first land property and the second land property.

15. A land property identification device, comprising:

the image acquisition module is used for acquiring a region to be identified;

the second grid division module is used for carrying out grid division on the area to be identified to obtain a plurality of corresponding block grids;

and the second identification module is used for obtaining a people stream density curve corresponding to the block grid with the first land property being the preset land property, inputting the people stream density curve into a pre-trained identification model, and obtaining a fourth land property output by the identification model.

16. An electronic device, comprising: a processor, a memory, and a bus, wherein,

the processor and the memory are communicated with each other through the bus;

the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any one of claims 1-13.

17. A non-transitory computer-readable storage medium storing computer instructions which, when executed by a computer, cause the computer to perform the method of any one of claims 1-13.