Method, device and storage medium for determining population distribution heat data

Abstract

Embodiments of the present application disclose a method and device for determining population distribution heat data and a storage medium, and belong to the technical field of computers. The method comprises the following steps: obtaining regional attribute data of a target region; determining first population weight data of the target region based on the regional attribute data of the target region; and determining second population distribution heat data of the target region based on first population distribution heat data of the target region and the first population weight data of the target region. The embodiments of the present application can improve the accuracy of population distribution heat data.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus and storage medium for determining thermal data of population distribution. Background Technology

[0002] Population distribution heatmap data describes the distribution of population within a geographical area. Typically, a grid is drawn across the geographical area, dividing it into multiple equal-sized unit areas (usually squares). Population distribution heatmap data consists of the number of people in each unit area. With the development of information technology, the applications of population distribution heatmap data are becoming increasingly widespread. For example, it can be used to guide the construction and operation of communication network infrastructure, store location selection, and advertising placement.

[0003] The common method for obtaining population distribution heat map data in related technologies is as follows: the network-side server collects the global positioning system (GPS) location data of each terminal, performs statistical calculations on the large amount of location data collected, and obtains the population distribution heat map data.

[0004] In the process of developing this application, the inventors discovered that the related technology has at least the following problems:

[0005] Because GPS location data has low accuracy, the resolution of population distribution heat map data collected using the above methods is also low. The resolution of population distribution heat map data is reflected in the size of the unit area; the larger the unit area, the lower the resolution, and vice versa. The resolution of population distribution heat map data obtained through GPS location data statistics is typically 200m × 200m. This resolution is insufficient for practical applications, and there is an urgent need for a method to improve the accuracy of population distribution heat map data. Summary of the Invention

[0006] This application provides a method for determining population distribution heat map data, which can solve the problem that the accuracy of population distribution heat map data is too low to meet the needs of practical applications. The technical solution is as follows:

[0007] Firstly, a method for determining population distribution heat map data is provided. The method includes: acquiring regional attribute data of a target area; determining primary population weight data of the target area based on the regional attribute data of the target area; and determining secondary population distribution heat map data of the target area based on the primary population distribution heat map data of the target area and the primary population weight data of the target area.

[0008] Among these, population distribution heatmap data describes the population distribution at different locations within a geographic region. Geographic attribute data describes the geographic attributes of different locations within a geographic region. Population weight data describes the probability of population distribution at different locations within a geographic region.

[0009] The geographic attribute data includes attribute values ​​for at least one geographic attribute corresponding to each unit area within the target region. The unit area is determined based on the target resolution. Geographic attributes are those that influence the population. The first-level population weight data includes the population weight value corresponding to each unit area within the target region. The first-level population distribution heatmap data includes the number of people corresponding to each initial unit area within the target region. The second-level population distribution heatmap data includes the number of people corresponding to each unit area within the target region. The initial unit area is determined based on the initial resolution. The target resolution is greater than the initial resolution. The initial resolution is the resolution of the initial population distribution heatmap data; it is a lower resolution. The target resolution is the desired resolution; it is a higher resolution.

[0010] The regional attribute data, population weight data, and second population distribution heat map data all correspond to the target resolution, while the first population distribution heat map data corresponds to the initial resolution.

[0011] The solution shown in this application embodiment allows the regional attribute data to be obtained from a map database. If the attribute values ​​of each regional attribute obtained from the map database are vector data, the vector data is converted into raster data. If the attribute values ​​of each regional attribute obtained from the map database are raster data, they can be used directly.

[0012] When determining the second population distribution heat map based on the first population distribution heat map data and the first-level population weight data, one approach is to first interpolate the first population distribution heat map data, then filter the interpolation result based on the first-level population weight data. The filtered result can then be used as the second population distribution heat map data. Alternatively, the filtered result can be further processed to obtain the second population distribution heat map data. Alternatively, one can directly interpolate the first population distribution heat map data based on the first-level population weight data, obtaining the interpolated result as the second population distribution heat map data. Or, the interpolated result can be further processed to obtain the second population distribution heat map data.

[0013] In one possible implementation, at least one geographic attribute that influences the population includes: geographic type and / or number of building floors. These two geographic attributes are more likely to affect the number of people in a geographic area, so the population weight values ​​determined using the attribute values ​​of these two geographic attributes are more accurate.

[0014] In one possible implementation, the population weight value for each unit area in the target area is determined based on the attribute values ​​of at least one regional attribute corresponding to each unit area, and the correspondence between the attribute values ​​of the regional attributes and the population weight values. For each initial unit area in the target area, the population weight values ​​corresponding to all unit areas in the initial unit area are normalized to obtain the normalized population weight value for each unit area. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the first-level population weight data for the target area.

[0015] The correspondence between regional attribute values ​​and population weight values ​​can be established using a correspondence table. After obtaining the regional attribute data, for each unit region, based on the attribute value of at least one regional attribute corresponding to that unit region, the corresponding population weight value is looked up in the correspondence table to obtain the population weight value for that unit region. Alternatively, the correspondence between regional attribute values ​​and population weight values ​​can be established using a calculation formula. After obtaining the regional attribute data, for each unit region, the attribute value of at least one regional attribute corresponding to that unit region is input into the calculation formula to obtain the population weight value for that unit region.

[0016] Further normalization of the obtained population weight values ​​using the initial unit area can make the number of people in the same initial unit area as the first population distribution heat map data as similar as possible, so that the output high-resolution population distribution heat map data can be as faithful as possible to the input low-resolution population distribution heat map data.

[0017] In one possible implementation, the correspondence between the attribute value of the regional attribute and the population weight value is expressed by the calculation formula for the attribute value of the regional attribute and the population weight value:

[0018] z i =p1×x i1 +p2×x i2 +......+p N ×x iN +θ; where x i1 x i2 ...x iN Let z be the attribute value of each geographic attribute in unit region i within the target region. i Let p1, p2, ..., p be the population weight values ​​for region i. N And θ are constants.

[0019] p1, p2, ... p NThe values ​​of θ can be set by technicians based on experience, or they can be learned from precisely detected sample population distribution heat map data. By using a calculation formula and determining the constants in the formula through learning, the population weight values ​​can be calculated more accurately, thereby improving the accuracy of the second population distribution heat map data.

[0020] In one possible implementation, the first population distribution heat map data of the target area is interpolated to obtain the third population distribution heat map data of the target area. Based on the third population distribution heat map data of the target area and the first-level population weight data of the target area, the second population distribution heat map data of the target area is determined.

[0021] The third population distribution heat map includes the number of people in each unit area within the target region.

[0022] Interpolation of the first population distribution heat map data can be performed using cubic spline interpolation or average interpolation. To determine the third population distribution heat map data, the first-level population weight data can be used to filter the third population distribution heat map data to obtain the second population distribution heat map data. Alternatively, the first-level population weight data and the third population distribution heat map data can be multiplied symmetrically to obtain the second population distribution heat map data.

[0023] Using cubic spline interpolation can better reflect the edge transition between adjacent initial unit regions and has a higher degree of matching with the actual population distribution.

[0024] In one possible implementation, the third population distribution heat map data of the target region is filtered based on the first-level population weight data of the target region, the target filtering window size, and the target window sliding step size to obtain a first filtering result. Based on the first filtering result, the second population distribution heat map data of the target region is determined.

[0025] The side length of the target filtering window can be the ratio of the side length of the initial unit region to the side length of the current unit region. The target window sliding step size can be 1.

[0026] After obtaining the first filtering result, the first filtering result can be used as the second population distribution heat map data of the target area, or the first filtering result can be further processed to obtain the second population distribution heat map data of the target area.

[0027] Using filtering can better reflect the mutual influence of the number of people in different areas and has a higher degree of matching with the actual population distribution.

[0028] In one possible implementation, geographic sub-region type data of the target sub-region within the target region is obtained. Based on the geographic sub-region type data of the target sub-region, secondary population weight data of the target sub-region is determined. Based on the first filtering result and the secondary population weight data of the target sub-region, secondary population distribution heat map data of the target region is determined.

[0029] The regional subdivision data includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region. The secondary population weight data includes the population weight value corresponding to each unit area in the target sub-region.

[0030] A mapping table between regional subdivision type values ​​and population weight values ​​can be used. For each regional subdivision type value, the corresponding population weight value is looked up to obtain the population weight value for each regional unit. Then, the population weight data is determined based on the population weight value for each regional unit. Alternatively, a calculation formula between regional subdivision type values ​​and population weight values ​​can be used. For each regional subdivision type value, the corresponding population weight value is calculated to obtain the population weight value for each regional unit. Then, the secondary population weight data is determined based on the population weight value for each regional unit.

[0031] Furthermore, the first filtering result is processed using secondary population weight data. Specifically, the second filtering result can be used to filter the first filtering result again, and the second filtering result can be used as the second population distribution heat map data of the target area. Alternatively, the second filtering result can be further processed to obtain the second population distribution heat map data of the target area.

[0032] By using regionally segmented data to determine secondary population weight data and further processing the first filtering results, the population distribution in a specific region (target sub-region) can be calculated more accurately, resulting in a higher degree of matching between the calculated second population distribution heat map data and the actual population distribution.

[0033] In one possible implementation, the population weight value for each unit region in the target sub-region is determined based on the type value corresponding to each unit region and the correspondence between the type value and the population weight value. For each unit region in the target sub-region, the influence value of each of the other unit regions on the population weight of the unit region is determined based on the population weight values ​​of multiple other unit regions outside the unit region and the distances between the unit region and each of the other unit regions. Based on the influence value of the other unit regions on the population weight of the unit region, the population weight value corresponding to the unit region is adjusted to obtain the adjusted population weight value of the unit region. For each initial unit region in the target sub-region, the adjusted population weight values ​​corresponding to all unit regions in the initial unit region are normalized to obtain the normalized population weight value corresponding to each unit region. The normalized population weight values ​​corresponding to all unit regions in the target sub-region are determined as the secondary population weight data of the target sub-region.

[0034] The correspondence can be a correspondence table or a calculation formula.

[0035] The above process first adjusts the population weight value for each unit region, and then performs normalization. The following explanation uses the adjustment process of the population weight value for a single unit region as an example:

[0036] The unit region can be any unit region within the target sub-region, and this unit region can be referred to as the unit region to be adjusted. Other unit regions outside the unit region to be adjusted are identified within the target sub-region as reference unit regions. At this point, all other unit regions outside the unit region to be adjusted within the target sub-region can be selected as reference unit regions. Alternatively, unit regions meeting specified conditions can be selected from the other unit regions outside the unit region to be adjusted within the target sub-region as reference unit regions. These specified conditions can be arbitrarily set according to actual needs; for example, the specified condition is that the distance between the reference unit region and the unit region to be adjusted is less than or equal to a distance threshold, which can be preset based on empirical values.

[0037] Furthermore, the distance between each reference unit region and the unit region to be adjusted can be determined. This distance can be a straight-line distance, or it can include both longitude and latitude distances. Then, based on the distance between each reference unit region and the unit region to be adjusted, and the population weight value corresponding to each reference unit region, the population weight influence value of each reference unit region on the unit region to be adjusted can be determined. The specific methods for determining the weight influence value can vary, such as querying a correspondence table or calculating using a formula. Below is a usable formula form:

[0038]

[0039] The distance between the reference unit region and the unit region to be adjusted can include the longitude distance and the latitude distance, d x d represents the distance in the longitude direction. y Let H be the latitudinal distance, H be the population weight of the reference unit area, and W be the influence of the reference unit area on the population weight of the unit area to be adjusted. This formula is based on a Gaussian function, where the standard deviation σ corresponds to the longitude direction. x Standard deviation σ corresponding to the latitudinal direction y The width can be determined based on the first width in the longitude direction and the second width in the latitude direction of the target sub-region. Specifically, the standard deviation corresponding to the larger of the first and second widths can be set as 1, and then calculated according to σ... x With σ y The ratio is equal to the ratio of the first width to the second width, and another standard deviation is calculated using this ratio.

[0040] After determining the impact value of the population weight of each reference unit area on the unit area to be adjusted, the population weight value of the unit area to be adjusted can be added to these population weight impact values ​​to obtain the adjusted population weight value of the unit area to be adjusted.

[0041] Because in reality, the number of people at any location has a radiating effect on the surrounding area, the above method can better restore the actual population distribution by calculating the mutual influence of population weight values ​​between unit areas.

[0042] In one possible implementation, based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, the portion of the first filtering result corresponding to the target sub-region is filtered again to obtain the second filtering result. The portions of the second filtering result and the first filtering result that do not correspond to the target sub-region are determined as the second population distribution heat map data of the target region.

[0043] The filter window size for secondary filtering can be the same as or different from that for primary filtering, and the sliding step size can also be the same or different. Secondary filtering applies the portion of the first filtering result that corresponds to the target sub-region. After obtaining the second filtering result through secondary filtering, the second filtering result can be combined with the portions of the first filtering result that do not correspond to the target sub-region to obtain the second population distribution thermal data for the target region.

[0044] Using filtering can better reflect the mutual influence of the number of people in different areas, and the calculated second population distribution data has a higher degree of matching with the actual situation.

[0045] In one possible implementation, the target sub-region is a multi-story building area. The regional subdivision type data includes sub-regional subdivision type data corresponding to each layer, and the sub-regional subdivision type data corresponding to each layer includes the type value of the regional subdivision type for each unit area in the target sub-region at that layer. The secondary population weight data includes sub-population weight data corresponding to each layer, and the sub-population weight data corresponding to each layer includes the population weight value for each unit area in the target sub-region at that layer. Accordingly, the secondary filtering process can be as follows: For each layer, based on the sub-population weight data of the layer, the target filtering window size, and the target window sliding step, the portion of the first filtering result corresponding to the target sub-region is filtered again to obtain the layer-specific filtering result. The combination of the layer-specific filtering results determines the second filtering result.

[0046] Filtering is performed on multiple layers separately, and then the filtering results from the multiple layers are simply combined to obtain the second filtering result. This second filtering result is the three-dimensional population distribution heat map data of the target sub-region. This processing method can yield relatively accurate three-dimensional population distribution heat map data.

[0047] In a second aspect, an apparatus for determining thermal data of population distribution is provided, the apparatus comprising one or more modules for implementing the method of the first aspect and its possible implementations.

[0048] Thirdly, a computer device is provided, comprising a memory and a processor, the memory for storing computer instructions; the processor executes the computer instructions stored in the memory to cause the computer device to perform the method of the first aspect and its possible implementations.

[0049] Fourthly, a computer-readable storage medium is provided, which stores computer program code, such that when the computer program code is executed by a computer device, the computer device performs the method of the first aspect and its possible implementations.

[0050] Fifthly, a computer program product is provided, comprising computer program code, wherein when the computer program code is executed by a computer device, the computer device executes the method of the first aspect and its possible implementations.

[0051] The beneficial effects of the technical solutions provided in this application are:

[0052] In this embodiment, high-resolution geographic attribute data is used to determine high-resolution population weight data. Then, based on the high-resolution population weight data, low-resolution population distribution heat map data is processed to obtain high-resolution population distribution heat map data, thereby improving the accuracy of the population distribution heat map data. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application;

[0054] Figure 2 This is a schematic diagram of population distribution heat map data provided in an embodiment of this application;

[0055] Figure 3 This is a schematic diagram of a regional attribute data provided in an embodiment of this application;

[0056] Figure 4 This is a schematic diagram of population weighting data provided in an embodiment of this application;

[0057] Figure 5 This is a schematic diagram of a type of regional subdivision data provided in an embodiment of this application;

[0058] Figure 6 This is a flowchart illustrating a method for determining population distribution thermal data provided in an embodiment of this application;

[0059] Figure 7 This is a schematic diagram of an interpolation process provided in an embodiment of this application;

[0060] Figure 8 This is a schematic diagram of a filtering process provided in an embodiment of this application;

[0061] Figure 9 This is a flowchart illustrating a method for determining population distribution thermal data provided in an embodiment of this application;

[0062] Figure 10 This is a schematic diagram illustrating the relationship between a target sub-region and a target region, provided in an embodiment of this application.

[0063] Figure 11 This is a flowchart illustrating a method for determining population distribution thermal data provided in an embodiment of this application;

[0064] Figure 12 This is a schematic diagram of a device for determining thermal data of population distribution provided in an embodiment of this application. Detailed Implementation

[0065] This application provides a method for determining population distribution heatmap data, which can be implemented by a computer device. The computer device can be used to statistically analyze population distribution heatmap data. It can be used by an entity with population distribution statistical needs; for example, a subway company might need to statistically analyze urban population distribution for subway line planning, or a chain store might need to statistically analyze urban population distribution for store location decisions. The computer device can also be used by an entity providing third-party population distribution statistical services. The computer device can be a terminal, server, or server group, etc., located in the computer room or office of such entities.

[0066] like Figure 1 As shown, a computer device may include a processor 1, a memory 2, and a communication component 3, etc.

[0067] The processor can be a central processing unit (CPU). The processor can be used to acquire regional attribute data of the target area, determine population weights based on regional attribute data, filter population heat map data based on population weight data, and so on.

[0068] The memory can be various types of volatile or non-volatile memory, such as solid-state drives (SSDs) and dynamic random access memory (DRAM). The memory can be used to store pre-stored data, intermediate data, and result data during the processing of population distribution heat map data, such as population distribution heat map data, geographic attribute data, and population weight data.

[0069] Communication components can be wired network connectors, wireless fidelity (WiFi) modules, Bluetooth modules, cellular communication modules, etc. These components can be used to transmit data with other devices. For example, they can receive first-order population distribution heat map data and regional attribute data from other devices, and can also send calculated second-order population distribution heat map data to a specified device, and so on.

[0070] The method for determining population distribution heat map data provided in this application can be used to determine high-resolution population distribution heat map data based on low-resolution population distribution heat map data. This application primarily improves the resolution of population distribution heat map data in the spatial domain, without limiting the processing in the time domain.

[0071] The following explains some important terms used in the processing:

[0072] Population distribution heat map data

[0073] Population distribution heatmap data is a type of data that describes the distribution of population in different locations within a geographic region.

[0074] The following is combined with Figure 2 This section describes the population distribution heatmap data for the target area. The target area is divided into many grids, each with the same size and shape (e.g., square). Each grid can be called a unit area. Each unit area corresponds to a population, such as... Figure 2 The population density of unit area A1 in the target area is 11. The population density of all unit areas in the target area constitutes the population distribution heat map of the target area. Visually, population distribution heat map data uses a raster format, and can be simply called raster data. Population distribution heat map data has a certain resolution, which is denoted by the size of the unit area. This resolution can be the side length of a square unit area; for example, the resolution could be 10m × 10m. Because the size of the unit area is directly related to the number and density of unit areas contained in the target area, the size of the unit area is used to represent the resolution.

[0075] Population heat map data can be stored as an array, which contains one or more values, each representing the number of people in a unit area within the target region.

[0076] Regional attribute data

[0077] Geographic attribute data describes the geographic attributes of different locations within a geographical area. It can correspond to at least one geographic attribute that influences the population, such as geographic type or number of building floors. Geographic type can be coarse-grained, such as intersection, road, park, shopping mall, residential building, theater, etc. Regarding the number of building floors, if the area is not divided into floors, such as a park or road, it can be considered as one floor. It is evident that different geographic types have different degrees of influence on the population. For example, the population density at intersections is generally higher than that on roads, and the population density in shopping malls is generally higher than that in parks. Different numbers of building floors also have different degrees of influence on the population; the more floors, the larger the population is likely to be.

[0078] The following is combined with Figure 3 This section describes the geographic attribute data of the target area. The target area is divided into many grids, each with the same size and shape (e.g., square). Each grid can be called a unit area. Each unit area corresponds to a geographic attribute value, such as... Figure 3The unit area A1 in the target area corresponds to a shopping mall type and a building with 5 floors. The attribute values ​​of the regional attributes corresponding to all unit areas in the target area constitute the regional attribute data of the target area. From a visual perspective, the regional attribute data uses raster data, which can be simply referred to as raster data. The regional attribute data has a certain resolution, which is denoted as the size of the unit area. It can be the side length of a square unit area, for example, the resolution can be 10m × 10m.

[0079] In terms of storage format, regional attribute data can be an array containing one or more values. Each value represents the attribute value of a regional attribute corresponding to a unit area in the target region. If there are multiple regional attributes, each value can be a vector, with each element of the vector being the attribute value of a regional attribute.

[0080] Population weight data

[0081] Population weighting data describes the probability of population distribution at different locations within a geographic region. This probability is influenced by geographic attribute data; for example, shopping malls have a higher weight than parks.

[0082] The following is combined with Figure 4 The population weight data for the target area is explained below. The target area is divided into many grids, each with the same size and shape (e.g., squares). Each grid can be called a unit area. Each unit area corresponds to a population weight value, which is a factor representing the probability of a high or low population density within the unit area. Figure 4 The population weight value corresponding to unit area A1 in the middle is 0.8. The population weight values ​​corresponding to all unit areas in the target area constitute the population weight data of the target area. From a visual perspective, the population weight data adopts a raster format, which can be simply referred to as raster data. The population weight data has a certain resolution, which is denoted by the size of the unit area. It can be the side length of a square unit area, for example, the resolution can be 10m×10m.

[0083] Population weight data can be stored as an array, which contains one or more values, each representing the population weight value of a unit area within the target region.

[0084] Regional segmentation data

[0085] Geographic segmentation type data describes the segmentation of different locations within a geographic region, providing a more granular representation of geographic classification. Geographic segmentation types are generally used to further subdivide sub-regions of a specific geographic type within a target region. For example, a park can be further subdivided into lawns, roads, pavilions, ponds, etc., while a shopping mall can be further subdivided into shops, corridors, restrooms, elevator lobbies, etc.

[0086] The following is combined with Figure 5 This section explains the geographic sub-region data within the target area. The target sub-region is further divided into grids, resulting in multiple unit areas, each with its own geographic sub-region type, such as... Figure 5 The geographic subdivision type corresponding to unit area A1 in the middle is "corridor". The geographic subdivision types corresponding to all unit areas within the target sub-region constitute the geographic subdivision type data for the target sub-region. Visually, this geographic subdivision type data uses a raster format and can be simply referred to as raster data. This data has a certain resolution, denoted as the size of the unit area, which can be the side length of a square unit area; for example, a resolution of 10m × 10m.

[0087] Geographic segmentation data can be stored as an array, which contains one or more values, each representing a geographic segmentation type corresponding to a unit area in the target sub-region.

[0088] initial resolution

[0089] The processing in this embodiment calculates high-resolution second population distribution heat map data based on low-resolution first population distribution heat map data. The low resolution can be referred to as the initial resolution, which is the resolution of the first population distribution heat map data input to the method of this embodiment. The first population distribution heat map data can be obtained through general channels, for example, population distribution heat map data obtained through GPS data statistics, with a resolution of 200m × 200m.

[0090] Target resolution

[0091] The aforementioned high resolution can be referred to as the target resolution, which is the resolution of the second population distribution heat map data output by the method in this embodiment of the application. The target resolution can be set based on the application requirements of the population heat map data, or determined based on the resolution of the available geographic attribute data. For example, the target resolution can be 10m×10m, 5m×5m, etc.

[0092] The following will describe the specific implementation methods. Figure 6 The processing flow shown is explained in detail below:

[0093] 601, retrieve the geographic attribute data of the target area.

[0094] The target area can be any geographical region, such as a city, a country, an administrative region, or a region artificially defined on a map. The geographic attribute data obtained in this step can be the attribute values ​​of at least one geographic attribute corresponding to each unit area within the target area. The unit area is determined based on the target resolution; for example, if the target resolution is 10m × 10m, then the unit area is a 10m × 10m square. The target resolution, i.e., the resolution of the desired population heat map data, can be preset by technical personnel.

[0095] Map applications typically maintain a map database on their backend servers. This database stores data on various geographic attributes (such as region type, building number of floors, etc.) corresponding to the geographical areas covered by the map. This data can be raster or vector data. If it is raster data and its resolution equals the target resolution, the computer can directly acquire this raster data. If the resolution of the raster data is higher than the target resolution, it is down-resolution processed. If it is vector data, it needs to be converted to raster data.

[0096] The following example uses vector data of a geographic type to illustrate the process of converting vector data into raster data.

[0097] For example, the vector data of a shopping mall might include the coordinates of each vertex of its planar outline (assuming the outline is an arbitrary polygon) and a mall type identifier. The mall is located within a target area. Based on the vertex coordinates, the corresponding polygon can be determined as the mall's planar outline. Furthermore, all unit areas within the scope of the planar outline in the target area can be identified, and the corresponding region type for these unit areas can be determined as "shopping mall."

[0098] Using a similar approach, vector data with various geographic attributes can be converted into raster data.

[0099] 602. Based on the attribute values ​​of the regional attributes corresponding to each unit region in the regional attribute data, and the correspondence between the attribute values ​​of the regional attributes and the population weight values, determine the population weight value corresponding to each unit region in the target region.

[0100] Among them, the population weight value is a factor that represents the probability of a high or low population density in a unit area.

[0101] Depending on the form of the correspondence, this step can be handled in several ways. Some of these methods are described below.

[0102] Method 1 uses the formula for calculating the relationship between the attribute value of the regional attribute and the population weight value.

[0103] The calculation formula can be expressed as follows:

[0104] z i =p1×x il +p2Xx i2 +......+p w ×x iN +θ

[0105] Where, x i1 x i2 ...x iN These represent the attribute values ​​for each geographic attribute in unit region i within the target region. i Let p1, p2, ..., p be the population weight values ​​for unit region i. N These are the weighting coefficients for each regional attribute, and can be constants. θ can also be a constant. p1, p2, ... p N The values ​​of θ can be learned from accurately detected population distribution heatmap data, or they can be set by technicians based on experience. The learning methods will be explained in detail later. i can be considered as the unit area number; for any numbered unit area, the population weight value can be calculated based on this formula.

[0106] Method 2 uses a table that maps the attribute values ​​of the regional attribute to the population weight values.

[0107] The correspondence table can be set by technical personnel based on experience. After obtaining the regional attribute data, for each unit area, based on the attribute value of at least one regional attribute corresponding to that unit area, the corresponding population weight value is looked up in the correspondence table, thus obtaining the population weight value corresponding to that unit area.

[0108] 603. For each initial unit region in the target region, normalize the population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region.

[0109] Based on the target resolution, the target region can be divided into several unit regions. Based on the initial resolution, the target region can be divided into several initial unit regions, and each initial unit region can include multiple unit regions. For example, if the initial resolution is 200m × 200m and the target resolution is 10m × 10m, then the size of each initial unit region is 200m × 200m, and the size of each unit region is 10m × 10m. Thus, each initial unit region includes 400 unit regions.

[0110] Based on the above characteristics, the population weight values ​​of unit regions within an initial unit region can be normalized. For any initial unit region, multiple unit regions within that initial unit region are identified, and their population weight values ​​are obtained. These population weight values ​​are then normalized to obtain the normalized population weight value for each unit region within the multiple unit regions. This process is repeated for each initial unit region to obtain the normalized population weight value for each unit region within the target region.

[0111] 604. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the population weight data of the target area.

[0112] 605. Interpolate the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area.

[0113] The first population distribution heat map includes the number of people corresponding to each initial unit area within the target region. The initial unit area is determined based on the initial resolution; for example, if the initial resolution is 200m × 200m, then the initial unit area is a 200m × 200m square. The third population distribution heat map includes the number of people corresponding to each unit area within the target region. The target resolution is greater than the initial resolution.

[0114] Interpolation processing of the first population distribution heat map data can be performed using cubic spline interpolation or average interpolation, among other methods. The interpolation process can be found in [reference needed]. Figure 7 .

[0115] 606. Based on the population weight data of the target area, the target filter window size, and the target window sliding step size, the third population distribution heat map data of the target area is filtered to obtain the second population distribution heat map data of the target area.

[0116] The second population distribution heat map includes the number of people in each unit area within the target region. This second population distribution heat map is the result of filtering the third population distribution heat map. The target filter window size can be set according to actual needs. For example, the ratio of the initial unit area's side length to the unit area's side length can be set as the filter window's side length. If the initial unit area is a 200m × 200m square, and the unit area's side length is a 10m × 10m square, then the target filter window size is 20 × 20. The target window sliding step size can be set by technicians based on experience, such as 1, 2, etc.

[0117] The filtering process can be found in [reference needed]. Figure 8Taking a square target area as an example, the values ​​in the figure are represented by X and Y, which do not mean that the values ​​are the same in different unit areas. The actual data can be any value. The third population distribution heat map data and population weight data are both data of the target area and are based on the target resolution, so they have the same spatial distribution. As shown in the figure, they can both be represented as 8×8 matrices.

[0118] A filter window can be used to select corresponding portions of the third-order population distribution heatmap data and population weight data. The target filter window size in the figure is 3×3. This can be achieved by using the filter window to select 3×3 matrices at the top left corner of both the third-order population distribution heatmap data and the population weight data, and then calculating the inner product Y of the two matrices. Additionally, a filtering result is created, which has the same data format as the third-order population distribution heatmap data—an 8×8 matrix. Initially, its contents are empty; the inner product Y calculated each time will be added to it later.

[0119] Next, we determine the placement of the current inner product Y in the filtering result. The filtering window is currently located at the top left corner of the third population distribution heat map. Therefore, the filtering window can also be placed at the top left corner of the filtering result. Then, we determine the corresponding position of the specified position in the filtering window within the filtering result as the result position, which is the placement of the inner product Y in the filtering result. The specified position can be arbitrarily set. Generally, for filtering windows with odd side lengths, the specified position can be the center of the filtering window; for filtering windows with even side lengths, the specified position is the top left (or top right, etc.) position among the four positions of the center of the filtering window. After determining the result position, we place the current inner product Y at this result position. As shown in the figure, if the 3×3 filtering window is located at the top left corner of the filtering result, then the corresponding position of the center position of the filtering window in the filtering result is the 2nd row, 2nd column position. This is the result position, and the value of the current inner product Y can be filled into the 2nd row, 2nd column position in the filtering result.

[0120] After the above processing, the first matrix inner product calculation of the filtering process is completed. Further, the filtering window can be slid to the right, the sliding distance being the target window's sliding step size. As shown in the figure, if the target window's sliding step size is 1, then it slides one unit to the right. This is then used to calculate the second matrix inner product. This process continues until the filtering window moves to the far right and the matrix inner product is calculated. Then, the filtering window is slid to the far left and down, the down sliding distance being the target window's sliding step size. This process continues until the filtering window moves to the bottom right corner and the matrix inner product is calculated.

[0121] At this point, the filtered results are still missing data from the outer ring, such as positions 1 / 1, 1 / 2, and 1 / 3 in the graph. This data can be supplemented by filling in the missing data in the outer ring using various methods. For example, for each missing data position in the outer ring, data from the nearest already filled position can be used to fill the gap. Alternatively, preset values ​​can be entered into the missing data positions, and so on.

[0122] After obtaining the second population distribution heat map data for the target area, the data can be sent to the target device for display. Alternatively, the second population distribution heat map data obtained over a certain period can be statistically analyzed, and based on the statistical results, a target advertising space can be selected from multiple pending advertising spaces for ad placement. Similarly, based on the statistical results, a target store location can be selected from multiple pending store locations for store establishment, and so on.

[0123] Population distribution heatmap data changes continuously over time, so the first population distribution heatmap data obtained above also changes continuously over time. Therefore, for the target area, steps 601-604 can be executed only once within a certain period, during which time the regional attribute data of the target area can be approximated as unchanged. Alternatively, steps 601-604 can be re-executed each time a change in the regional attribute data of the target area is detected. Alternatively, steps 601-604 can be re-executed each time the second population distribution heatmap data is calculated. After step 604 is completed, the population weight data of the target area is obtained. The first population distribution heatmap data can be obtained periodically. Each time the first population distribution heatmap data is obtained, steps 605-606 can be executed by calling the population weight data of the target area.

[0124] In the above processing, for low-resolution population distribution heat map data, an interpolation-then-filtering approach is used to obtain high-resolution population distribution heat map data. Optionally, after determining the population weight data, interpolation can also be performed directly on the low-resolution population distribution heat map data based on the population weight data to obtain high-resolution population distribution heat map data.

[0125] The specific steps can be as follows:

[0126] For each initial unit region, the population size of that initial unit region in the first population distribution heatmap is determined. The population weight value of each unit region contained within that initial unit region is determined in the population weight data, and the sum of these population weight values ​​is determined. For each unit region within the initial unit region, the ratio of that unit region's population weight value to the sum is determined. Furthermore, the product of this ratio and the population size of that initial unit region is determined to obtain the population size corresponding to that unit region. Using this method, the population size of all unit regions within the target region can be obtained, thus completing the interpolation and obtaining high-resolution population distribution heatmap data.

[0127] In the above processing, population weight data is used to filter the third population distribution heat map data. Optionally, filtering can be omitted, and the population weight data and the third population distribution heat map data can be directly multiplied digit-wise to obtain the second population distribution heat map data. The specific processing can be as follows:

[0128] For each unit area, obtain the population population of that unit area in the third population heat map data, and obtain the population weight value of that unit area in the population weight data. Calculate the product of the population population and the population weight value to determine the updated population population of that unit area. Use the updated population population of all unit areas in the target area to determine the final high-resolution population distribution heat map data.

[0129] In the above processing, step 603 normalizes the population weight values, which makes the number of people in the same initial unit area as the first population distribution heat map data and the second population distribution heat map data obtained by filtering as similar as possible. This ensures that the output high-resolution population distribution heat map data is as faithful as possible to the input low-resolution population distribution heat map data. Optionally, the normalization process in step 603 can be omitted. The population weight values ​​of all unit areas output in step 602 can be directly used as the population weight data of the target area. Then, after filtering, the filtering result can be scaled and adjusted so that the number of people in the same initial unit area as the first population distribution heat map data and the second population distribution heat map data are the same. The specific processing can be as follows:

[0130] For each initial unit area, the population corresponding to that initial unit area (referred to as the initial population) is obtained from the first population distribution heat map. The total population of all unit areas included in that initial unit area is determined from the filtering results. The ratio of the initial population to the total population is then determined. In the filtering results, the population of each unit area included in that initial unit area is multiplied by this ratio to obtain the adjusted population. The adjusted population of all unit areas in the target area is then used to determine the second population distribution heat map for the target area.

[0131] In the above processing, p1, p2, ... p N The values ​​of θ can be learned based on accurately detected population distribution heat map data. The corresponding learning method is as follows:

[0132] First, obtain the sample population distribution heat map data of the precisely detected sample area. This data includes the number of people in each unit area within the sample area, determined by the target resolution, such as 10m × 10m. Precise detection can be achieved in various ways, such as through analysis of surveillance video images or by detecting terminal locations via base stations, etc. Next, the sample population distribution heat map data is down-resolution processed. Specifically, for each initial unit area in the sample area, the number of people in all unit areas included in that initial unit area is added together to obtain the initial unit area's population. The population of all initial unit areas in the sample area constitutes the down-resolution population distribution heat map data. The initial unit area is determined by the initial resolution, such as 200m × 200m. Then, using the down-resolution population distribution heat map data as the first population distribution heat map data, the above process is applied to it to obtain the second population distribution heat map data. In the processing, the calculation formula used in step 602 uses p1, p2, ... p... N Since θ is unknown, the second population distribution heat map obtained at this time is an expression using these unknowns. The correct values ​​for the second population distribution heat map should be the sample population distribution heat map. Therefore, this expression should be equal to the sample population distribution heat map. In this way, p1 is established.

[0133] p2、……p N The relationship between θ and θ can be established through a large number of samples, resulting in a large number of p1, p2, ... p N The relationships between θ and p1, p2, ..., p2 can be obtained by performing regression analysis on these relationships. N The values ​​of θ.

[0134] The above process uses a single filter to obtain high-resolution population distribution heatmap data. The following processing flow will acquire geographic sub-region data within the target area, and further apply a second filter to the first filter result to obtain high-resolution population distribution heatmap data. The corresponding processing flow can be found in [reference needed]. Figure 9 It includes the following steps:

[0135] 901, retrieve the geographic attribute data of the target area.

[0136] 902. Based on the attribute values ​​of the regional attributes corresponding to each unit region in the regional attribute data, and the correspondence between the attribute values ​​of the regional attributes and the population weight values, determine the population weight value corresponding to each unit region in the target region.

[0137] 903. For each initial unit region in the target region, normalize the population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region.

[0138] 904. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the first-level population weight data of the target area.

[0139] The primary population weight data and the secondary population weight data used in subsequent steps are population weight data determined based on different parameters. They are used for the first and second filtering processes, respectively. While their numerical values ​​may differ, their essential meaning is the same: both describe the probability of population distribution at different locations within a geographical area. Accordingly, these population weight values ​​can also be called primary population weight values.

[0140] 905. Interpolate the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area.

[0141] 906. Based on the primary population weight data of the target region, the target filtering window size, and the target window sliding step size, the third population distribution heat map data of the target region is filtered to obtain the first filtering result.

[0142] The processing in steps 901-906 is similar to that in steps 601-606, and can be found in the relevant descriptions in the above process. The difference is that the filtering result in step 606 is directly used as the final output of the second population distribution heat map, while the filtering result in step 906 requires further processing to obtain the final output of the second population distribution heat map.

[0143] 907, retrieve the geographic sub-region type data of the target area.

[0144] Here, the target sub-region is a part or all of the target region, and the relationship between the target sub-region and the target region can be as follows: Figure 10 As shown. The geographic subdivision type data has been explained at the beginning of the embodiments of this application. Here, the geographic subdivision type data includes the type value of the geographic subdivision type corresponding to each unit area in the target sub-region.

[0145] Computer equipment can define one or more target sub-regions within a target area. There are various ways to define target sub-regions; several are listed here:

[0146] Method 1: Manually select a target sub-region within the target region.

[0147] Method Two: First, identify unit areas of a specified geographic type within the target area. The specified geographic type can be set by technical personnel based on actual needs. For example, the specified geographic type can be one or more types such as office buildings, parks, and shopping malls. Then, determine the target sub-regions composed of unit areas of the specified geographic type. Specifically, within the unit areas of the specified geographic type, adjacent unit areas of the same geographic type can be identified as a single target sub-region.

[0148] Optionally, for each identified target sub-region, determine whether the target sub-region is an orthogonally placed rectangle. An orthogonally placed rectangle is one whose sides are parallel to a meridian or parallel. If the target sub-region is not an orthogonally placed rectangle, it can be adjusted by determining the smallest inscribed rectangle of the orthogonal sub-region and adjusting it to fit the region corresponding to that rectangle. If the target sub-region is an orthogonally placed rectangle, no adjustment is needed. If the target sub-region can be divided into multiple orthogonally placed rectangles when it is not an orthogonally placed rectangle, another processing method can be used: dividing the target sub-region into multiple orthogonally placed rectangles, with each orthogonally placed rectangle serving as a target sub-region.

[0149] For each identified target sub-region, we can obtain its regional sub-type data for subsequent processing. Compared to regional type data, regional sub-type data provides a more granular classification of regional types. For example, a park can be further subdivided into grassland, roads, pavilions, ponds, etc. Regional sub-type data can be obtained from the map application's backend server using a method similar to step 601, or it can be manually collected and recorded, i.e., by staff visiting the site to observe and record it.

[0150] 908. Based on the type value of the regional subdivision type corresponding to each unit area in the target sub-region, and the correspondence between the type value and the population weight value, determine the population weight value corresponding to each unit area in the target sub-region.

[0151] The population weight value here can also be called the secondary population weight value.

[0152] This step is similar to step 602. It can also be done by using a correspondence table or a calculation formula. Please refer to the above embodiment. The only difference is that the attribute used to determine the population weight value here is the regional subdivision type.

[0153] 909. For each unit region in the target sub-region, based on the population weight values ​​of multiple other unit regions outside the target sub-region and the distances between the unit region and the multiple other unit regions, determine the population weight influence values ​​of the multiple other unit regions on the unit region. Based on the population weight influence values ​​of the multiple other unit regions on the unit region, adjust the population weight value corresponding to the unit region to obtain the adjusted population weight value of the unit region.

[0154] This step takes into account the mutual influence of population weight values ​​between different unit areas and adjusts the population weight value of each unit area within the target sub-region. The adjustment process for the population weight value of a single unit area is illustrated below:

[0155] The unit region can be any unit region within the target sub-region, and this unit region can be referred to as the unit region to be adjusted. Other unit regions outside the unit region to be adjusted are identified within the target sub-region as reference unit regions. At this point, all other unit regions outside the unit region to be adjusted within the target sub-region can be selected as reference unit regions. Alternatively, unit regions meeting specified conditions can be selected from the other unit regions outside the unit region to be adjusted within the target sub-region as reference unit regions. These specified conditions can be arbitrarily set according to actual needs; for example, the specified condition is that the distance between the reference unit region and the unit region to be adjusted is less than or equal to a distance threshold, which can be preset based on empirical values.

[0156] Furthermore, the distance between each reference unit region and the unit region to be adjusted can be determined. This distance can be a straight-line distance, or it can include both longitude and latitude distances. Then, based on the distance between each reference unit region and the unit region to be adjusted, and the population weight value corresponding to each reference unit region, the population weight influence value of each reference unit region on the unit region to be adjusted can be determined. The specific methods for determining the weight influence value can vary, such as querying a correspondence table or calculating using a formula. Below is a usable formula form:

[0157]

[0158] The distance between the reference unit region and the unit region to be adjusted can include the longitude distance and the latitude distance, d x d represents the distance in the longitude direction. yLet H be the latitudinal distance, H be the population weight of the reference unit area, and W be the influence of the reference unit area on the population weight of the unit area to be adjusted. This formula is based on a Gaussian function, where the standard deviation σ corresponds to the longitude direction. x Standard deviation σ corresponding to the latitudinal direction y The width can be determined based on the first width in the longitude direction and the second width in the latitude direction of the target sub-region. Specifically, the standard deviation corresponding to the larger of the first and second widths can be set as 1, and then calculated according to σ... x With σ y The ratio is equal to the ratio of the first width to the second width, and another standard deviation is calculated using this ratio.

[0159] After determining the impact value of the population weight of each reference unit area on the unit area to be adjusted, the population weight value of the unit area to be adjusted can be added to these population weight impact values ​​to obtain the adjusted population weight value of the unit area to be adjusted.

[0160] Based on the above method, the adjusted population weight value of each unit area in the target sub-region can be determined.

[0161] Optionally, the standard deviation σ mentioned above x and standard deviation σ y You can also use the same value, such as 1.

[0162] Alternatively, other formulas can be used, such as the population weight impact value being equal to the product of the straight-line distance and a fixed coefficient, etc.

[0163] Optionally, the adjustment of the population weight value in step 909 can be skipped, and the population weight value corresponding to each unit area in the target sub-region output in step 908 can be directly processed.

[0164] 910. For each initial unit region in the target sub-region, normalize the adjusted population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region.

[0165] This step is similar to step 603; please refer to the relevant instructions.

[0166] 911. The normalized population weight values ​​corresponding to all unit areas in the target sub-region are determined as the secondary population weight data of the target sub-region.

[0167] The secondary population weight data includes the population weight value corresponding to each unit area in the target region.

[0168] 912. Based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the second filtering result.

[0169] The first filtering result corresponds to the data for the entire target region. Step 912 can only perform filtering on the portion of the first filtering result that corresponds to the target sub-region. For the portion of the first filtering result that does not correspond to the target sub-region, this step can skip processing. The secondary population weight data corresponds to the target sub-region. Moreover, both the secondary population weight data and the portion of the first filtering result corresponding to the target sub-region are based on the target resolution, so these two data have the same spatial distribution; for example, they are both 6×6 matrices. The corresponding filtering process is similar to the filtering process in step 606; please refer to the relevant instructions.

[0170] 913. The portions of the second and first filtering results that do not correspond to the target sub-region are determined as the second population distribution heat map data of the target region.

[0171] In the above processing, for low-resolution population distribution heat map data, an interpolation-then-filtering approach is used to obtain the first filtered result. Optionally, after determining the first-level population weight data, interpolation can also be performed directly on the low-resolution population distribution heat map data based on the first-level population weight data to obtain high-resolution population distribution heat map data. For specific processing details, please refer to the relevant content above.

[0172] In the above processing, the primary population weight data is used to filter the secondary population distribution heat map data. Optionally, filtering can be omitted, and the primary population weight data and the secondary population distribution heat map data can be directly multiplied digit-wise to obtain the primary intermediate data, which corresponds to the primary filtering result. For specific processing details, please refer to the relevant content above.

[0173] In the above processing, the second-level population weight data is used to filter the first filtering result. Optionally, filtering can be omitted, and the second-level population weight data can be directly multiplied digit-wise with the first filtering result to obtain the second population distribution heat map data. For specific processing details, please refer to the relevant content above.

[0174] In the above processing, steps 903 and 910 involve normalization. Optionally, these two normalization steps can be omitted. Instead, the population weight value output in step 902 can be directly determined as the primary population weight data of the target area, and the population weight value output in step 909 can be determined as the secondary population weight data of the target area. Then, after the second filtering, the portion of the first filtering result that does not correspond to the target sub-region is identified as a combination of the first filtering result and the second filtering result. This combination data is then scaled and adjusted to obtain the second population distribution heat map data, ensuring that the final output of the second population distribution heat map data has the same number of people in the same initial unit area as the first population distribution heat map data. For specific processing details, please refer to the relevant content above.

[0175] In the above processing, step 902 will also involve p1, p2, ... p N The learning of the values ​​of θ can be based on the learning methods described above, still relying on... Figure 6 The process involves processing the down-resolution population distribution heat map data to ultimately determine p1, p2, ... p N And the value of θ. Additionally, if step 908 also uses a calculation formula to calculate the population weight value, it also involves pre-learning the constants in the calculation formula. This can be done by determining p1, p2, ... p... N After determining the values ​​of θ, a similar learning method is used for further learning. Alternatively, the learning method described above can be used to learn the values ​​of the constants involved in steps 902 and 908 together.

[0176] The following details the handling of cases where the target sub-region is a multi-story building area. The corresponding handling methods are as follows: Figure 11 As shown, it includes the following steps:

[0177] 1101, Obtain the geographic attribute data of the target area.

[0178] 1102. Based on the attribute values ​​of the regional attributes corresponding to each unit region in the regional attribute data, and the correspondence between the attribute values ​​of the regional attributes and the population weight values, determine the population weight value corresponding to each unit region in the target region.

[0179] 1103. For each initial unit region in the target region, normalize the population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region.

[0180] 1104. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the first-level population weight data of the target area.

[0181] 1105. Interpolate the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area.

[0182] 1106. Based on the primary population weight data of the target region, the target filtering window size, and the target window sliding step size, the third population distribution heat map data of the target region is filtered to obtain the first filtering result.

[0183] The processing in steps 1101-1106 is similar to that in steps 601-606, and can be found in the relevant descriptions in the above process. The difference is that the filtering result in step 606 is directly used as the final output of the second population distribution heat map, while the filtering result in step 1106 needs to undergo further processing to obtain the final output of the second population distribution heat map.

[0184] 1107, retrieve the geographic sub-region type data of the target area.

[0185] The geographic segmentation data includes sub-geographic segmentation data for each layer. The sub-geographic segmentation data for each layer includes the type value of the geographic segmentation type for each unit region within the target sub-region at that layer.

[0186] The process of determining the target sub-region and obtaining regional subdivision data is similar to step 907. For specific processing details, please refer to the relevant content above.

[0187] 1108. For each floor of a multi-story building, based on the type value of the regional subdivision type corresponding to each unit area in the target sub-region at that floor, and the correspondence between the type value and the population weight value, determine the population weight value corresponding to each unit area in the target sub-region at that floor.

[0188] 1109. In each floor of a multi-story building, for each unit area in the target sub-region, based on the population weight values ​​corresponding to multiple other unit areas outside the target sub-region and the distances between the unit area and the multiple other unit areas, determine the population weight influence values ​​of the multiple other unit areas on the unit area. Based on the population weight influence values ​​of the multiple other unit areas on the unit area, adjust the population weight value corresponding to the unit area to obtain the adjusted population weight value of the unit area.

[0189] The processing of each layer in steps 1108-1109 is the same as in steps 908-909. For details, please refer to the relevant content above.

[0190] 1110. For each initial unit region in the target sub-region, normalize the adjusted population weight values ​​of all unit regions in the initial unit region at all layers to obtain the normalized population weight value of each unit region at each layer.

[0191] To illustrate the normalization process in this step, consider an initial unit area size of 200m × 200m, a unit area size of 10m × 10m, and a building with 10 floors. In this case, each floor of an initial unit area corresponds to 400 adjusted population weight values, and the 10 floors together contain 4000 adjusted population weight values. For an initial unit area, these 4000 population weight values ​​need to be normalized.

[0192] 1111, the normalized population weight values ​​corresponding to all unit regions in the target sub-region of each layer are determined as the sub-population weight data corresponding to each layer, thus obtaining the secondary population weight data of the target sub-region.

[0193] The secondary population weight data includes the sub-population weight data corresponding to each level.

[0194] 1112. For each floor of a multi-story building, based on the sub-population weight data of that floor, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the filtering result corresponding to that floor.

[0195] The first filtering result is filtered again using the sub-population weight data of each layer to obtain the corresponding filtering result for each layer. The filtering process of each layer is similar to the filtering process in step 606. For details, please refer to the above-mentioned content.

[0196] 1113. Combine the filtering results corresponding to each layer to determine the second filtering result.

[0197] The second filtering result includes multiple layers of data.

[0198] 1114. The portions of the second and first filtering results that do not correspond to the target sub-region are determined as the second population distribution heat map data of the target region.

[0199] The second population distribution heat map data consists of multi-layered data in the target sub-region and single-layered data in the area outside the target sub-region.

[0200] In the above processing, for low-resolution population distribution heat map data, an interpolation-then-filtering approach is used to obtain the first filtered result. Optionally, after determining the first-level population weight data, interpolation can also be performed directly on the low-resolution population distribution heat map data based on the first-level population weight data to obtain high-resolution population distribution heat map data. For specific processing details, please refer to the relevant content above.

[0201] In the above processing, the primary population weight data is used to filter the secondary population distribution heat map data. Optionally, filtering can be omitted, and the primary population weight data and the secondary population distribution heat map data can be directly multiplied digit-wise to obtain the primary intermediate data, which corresponds to the primary filtering result. For specific processing details, please refer to the relevant content above.

[0202] In the above processing, the second-level population weight data is used to filter the first filtering result. Optionally, filtering can be omitted, and the sub-population weight data of each layer can be directly multiplied with the first filtering result digit by digit. For details, please refer to the relevant content above.

[0203] In the above processing, steps 1103 and 1110 involve normalization. Optionally, these two normalization steps can be omitted. Instead, the population weight value output in step 1102 can be directly determined as the primary population weight data for the target area, and the population weight value output in step 1109 can be determined as the secondary population weight data for the target area. Then, after the second filtering, the combination data of the portions of the second and first filtering results that do not correspond to the target sub-region is determined. This combination data is then scaled and adjusted to obtain the second population distribution heat map data, ensuring that the final output of the second population distribution heat map data has the same number of people in the same initial unit area as the first population distribution heat map data. For specific processing details, please refer to the relevant content above.

[0204] In the above processing, step 1102 will also involve p1, p2, ... p N The learning of the values ​​of θ can be based on the learning methods described above, still relying on... Figure 6 The process involves processing the down-resolution population distribution heat map data to ultimately determine p1, p2, ... p N And the value of θ. Additionally, if step 1108 also uses a calculation formula to calculate the population weight value, it also involves pre-learning the constants in the calculation formula. This can be done by determining p1, p2, ... p... N After determining the values ​​of θ, a similar learning method is used for further learning. Alternatively, the learning method described above can be used to learn the values ​​of the constants involved in steps 1102 and 1108 together.

[0205] In this embodiment, high-resolution geographic attribute data is used to determine high-resolution population weight data. Then, based on the high-resolution population weight data, low-resolution population distribution heat map data is processed to obtain high-resolution population distribution heat map data, thereby improving the accuracy of the population distribution heat map data.

[0206] Based on the same technical concept, this application also provides an apparatus for determining population distribution thermal data. This apparatus can be applied to the computer equipment provided in the above embodiments, such as... Figure 12 As shown, the device includes:

[0207] The acquisition module 1210 is used to acquire regional attribute data of a target area. The regional attribute data includes attribute values ​​of at least one regional attribute corresponding to each unit area within the target area. The unit area is determined based on the target resolution, and the regional attribute is a regional attribute that has an impact on the population. Specifically, it can implement the acquisition functions in steps 601, 901, and 1101 above, as well as other implicit steps.

[0208] The weight determination module 1220 is used to determine the primary population weight data of the target area based on the regional attribute data of the target area. The primary population weight data includes the population weight value corresponding to each unit area within the target area. Specifically, it can implement the weight determination functions in steps 602-604, 902-904, and 1102-1104 above, as well as other implicit steps.

[0209] The resolution enhancement module 1230 is used to determine the second population distribution heat map of the target area based on the first population distribution heat map data and the first-level population weight data of the target area. The first population distribution heat map data includes the number of people corresponding to each initial unit area in the target area, where the initial unit area is determined according to an initial resolution. The second population distribution heat map data includes the number of people corresponding to each unit area in the target area, and the target resolution is greater than the initial resolution. Specifically, it can implement the resolution enhancement function in steps 605-606, 905-913, and 1105-1114 above, as well as other implicit steps.

[0210] In one possible implementation, the at least one geographic attribute that has an impact on the population includes: geographic type and / or number of building floors.

[0211] In one possible implementation, the weight determination module is configured to: determine the population weight value corresponding to each unit region in the target region based on the attribute value of at least one regional attribute corresponding to each unit region in the target region, and the correspondence between the attribute value of the regional attribute and the population weight value; for each initial unit region in the target region, normalize the population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region; and determine the normalized population weight values ​​corresponding to all unit regions in the target region as the first-level population weight data of the target region.

[0212] In one possible implementation, the correspondence between the attribute value of the regional attribute and the population weight value is defined by a calculation formula for the attribute value of the regional attribute and the population weight value, wherein the calculation formula is:

[0213] z i =p1×x i1 +p2×x i2 +......+p N ×x iN +θ; where x i1 x i2 ...x iN Let z be the attribute value of each geographic attribute in unit region i within the target region. i Let p1, p2, ..., p be the population weight values ​​for the unit region i. N And θ are constants.

[0214] In one possible implementation, the resolution enhancement module 1230 is configured to: interpolate the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area, wherein the third population distribution heat map data includes the number of people corresponding to each unit area in the target area; and determine the second population distribution heat map data of the target area based on the third population distribution heat map data of the target area and the first-level population weight data of the target area.

[0215] In one possible implementation, the resolution enhancement module 1230 is configured to: filter the third population distribution heat map data of the target region based on the first-level population weight data of the target region, the target filtering window size, and the target window sliding step size to obtain a first filtering result; and determine the second population distribution heat map data of the target region based on the first filtering result.

[0216] In one possible implementation, the resolution enhancement module 1230 is configured to: acquire geographic sub-regional subdivision type data of the target sub-region within the target region, wherein the geographic subdivision type data includes the type value of the geographic subdivision type corresponding to each unit region in the target sub-region; determine the secondary population weight data of the target sub-region based on the geographic subdivision type data of the target sub-region, wherein the secondary population weight data includes the population weight value corresponding to each unit region in the target sub-region; and determine the second population distribution heat map data of the target region based on the first filtering result and the secondary population weight data of the target sub-region.

[0217] In one possible implementation, the resolution enhancement module 1230 is configured to: determine the population weight value corresponding to each unit region in the target sub-region based on the type value corresponding to each unit region in the target sub-region and the correspondence between the type value and the population weight value; for each unit region in the target sub-region, determine the population weight influence value of each of the other unit regions on the unit region based on the population weight values ​​corresponding to multiple other unit regions outside the unit region in the target sub-region and the distance between the unit region and the multiple other unit regions respectively; adjust the population weight value corresponding to the unit region based on the population weight influence value of the multiple other unit regions on the unit region to obtain the adjusted population weight value of the unit region; for each initial unit region in the target sub-region, normalize the adjusted population weight values ​​corresponding to all unit regions in the initial unit region to obtain the normalized population weight value corresponding to each unit region; and determine the normalized population weight values ​​corresponding to all unit regions in the target sub-region as the secondary population weight data of the target sub-region.

[0218] In one possible implementation, the resolution enhancement module 1230 is configured to: based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, perform a second filtering process on the portion of the first filtering result that corresponds to the target sub-region to obtain a second filtering result; and determine the portion of the second filtering result and the portion of the first filtering result that does not correspond to the target sub-region as the second population distribution heat map data of the target region.

[0219] In one possible implementation, the target sub-region is a multi-story building area. The regional subdivision type data includes sub-regional subdivision type data corresponding to each layer, and the sub-regional subdivision type data corresponding to each layer includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region at that layer. The secondary population weight data includes sub-population weight data corresponding to each layer, and the sub-population weight data corresponding to each layer includes the population weight value corresponding to each unit area in the target sub-region at that layer.

[0220] The resolution enhancement module 1230 is configured to: for each layer, based on the sub-population weight data of the layer, the target filtering window size, and the target window sliding step size, perform a second filtering process on the part of the first filtering result corresponding to the target sub-region to obtain the filtering result corresponding to the layer; and combine the filtering results corresponding to each layer to determine the second filtering result.

[0221] It should be noted that the above-mentioned acquisition module 1210, weight determination module 1220 and resolution enhancement module 1230 can be implemented by a processor, or by a processor in conjunction with a memory.

[0222] It should be noted that the apparatus for determining population distribution thermal data provided in the above embodiments is only illustrated by the division of the above functional modules when performing the processing of determining population distribution thermal data. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the apparatus can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus for determining population distribution thermal data and the method embodiment for determining population distribution thermal data provided in the above embodiments belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0223] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a device, they generate, in whole or in part, the processes or functions described in the embodiments of this application. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic cable, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to the device or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, and magnetic tape), an optical medium (e.g., digital video disk (DVD), etc.), or a semiconductor medium (e.g., solid-state drive).

[0224] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0225] The above description is only one embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A method for determining population distribution thermal data, characterized in that, The method includes: Obtain the regional attribute data of the target area, wherein the regional attribute data includes the attribute value of at least one regional attribute corresponding to each unit area in the target area, the unit area is determined according to the target resolution, the regional attribute is a regional attribute that has an impact on the population, and the regional attribute includes the regional type; Based on the geographic attribute data of the target area, the primary population weight data of the target area is determined, wherein the primary population weight data includes the population weight value corresponding to each unit area in the target area; Interpolation processing is performed on the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area. The first population distribution heat map data includes the number of people corresponding to each initial unit area in the target area. The initial unit area is determined according to the initial resolution. The target resolution is greater than the initial resolution. The third population distribution heat map data includes the number of people corresponding to each unit area in the target area. Based on the primary population weight data of the target region, the third population distribution heat map data of the target region is filtered to obtain the first filtering result; Obtain the regional sub-regional subdivision type data of the target area, wherein the regional subdivision type data includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region, the target sub-region is the adjacent unit area of ​​the target area that corresponds to a specified regional type, and the regional subdivision type data is a further subdivision of the sub-regions of the specified regional type; Based on the regional subdivision data of the target sub-region, the secondary population weight data of the target sub-region is determined, wherein the secondary population weight data includes the population weight value corresponding to each unit area in the target sub-region; The first filtering result is filtered based on the secondary population weight data of the target sub-region to obtain the second population distribution heat map data of the target region.

2. The method according to claim 1, characterized in that, The at least one regional attribute that has an impact on the population includes: regional type and number of building floors.

3. The method according to claim 1 or 2, characterized in that, The determination of the primary population weight data of the target region based on the regional attribute data of the target region includes: Based on the attribute values ​​of at least one regional attribute corresponding to each unit area in the target area, and the correspondence between the attribute values ​​of the regional attributes and the population weight values, the population weight value corresponding to each unit area in the target area is determined. For each initial unit region in the target region, the population weight values ​​corresponding to all unit regions in the initial unit region are normalized to obtain the normalized population weight value corresponding to each unit region. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the first-level population weight data of the target area.

4. The method according to claim 3, characterized in that, The correspondence between the attribute value of the regional attribute and the population weight value is defined by the calculation formula for the attribute value of the regional attribute and the population weight value. The calculation formula is as follows: ； in, , ... Each is a unit region within the target region. The attribute value of each regional attribute in the text. For the unit region Population weight value, , ... and It is a constant.

5. The method according to claim 1, characterized in that, The determination of the secondary population weight data of the target sub-region based on the regional subdivision type data of the target sub-region includes: Based on the type value corresponding to each unit area in the target sub-region and the correspondence between the type value and the population weight value, the population weight value corresponding to each unit area in the target sub-region is determined. For each unit region in the target sub-region, based on the population weight values ​​corresponding to multiple other unit regions outside the unit region in the target sub-region, and the distances between the unit region and the multiple other unit regions, the population weight influence values ​​of the multiple other unit regions on the unit region are determined. Based on the population weight influence values ​​of the multiple other unit regions on the unit region, the population weight value corresponding to the unit region is adjusted to obtain the adjusted population weight value of the unit region. For each initial unit region in the target sub-region, the adjusted population weight values ​​corresponding to all unit regions in the initial unit region are normalized to obtain the normalized population weight value corresponding to each unit region. The normalized population weight values ​​corresponding to all unit regions in the target sub-region are determined as the secondary population weight data of the target sub-region.

6. The method according to claim 5, characterized in that, The filtering process, based on the secondary population weight data of the target sub-region, is applied to the first filtering result to obtain the second population distribution heat map data of the target region, including: Based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the second filtering result; The portions of the second filtering result and the first filtering result that do not correspond to the target sub-region are determined as the second population distribution thermal data of the target region.

7. The method according to claim 6, characterized in that, The target sub-region is a multi-story building area; the regional subdivision type data includes sub-region subdivision type data corresponding to each layer, and the sub-region subdivision type data corresponding to each layer includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region at that layer; the secondary population weight data includes sub-population weight data corresponding to each layer, and the sub-population weight data corresponding to each layer includes the population weight value corresponding to each unit area in the target sub-region at that layer. The second filtering result is obtained by further filtering the portion of the first filtering result corresponding to the target sub-region based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, including: For each layer, based on the sub-population weight data of the layer, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the filtering result corresponding to the layer. The second filtering result is determined by combining the filtering results corresponding to each layer.

8. An apparatus for determining thermal data of population distribution, characterized in that, The device includes: The acquisition module is used to acquire the regional attribute data of the target area, wherein the regional attribute data includes the attribute value of at least one regional attribute corresponding to each unit area in the target area, the unit area is determined according to the target resolution, the regional attribute is a regional attribute that has an impact on the population, and the regional attribute includes the regional type; The weight determination module is used to determine the primary population weight data of the target area based on the regional attribute data of the target area, wherein the primary population weight data includes the population weight value corresponding to each unit area in the target area; A resolution enhancement module is used to interpolate the first population distribution heat map data of the target area to obtain the third population distribution heat map data of the target area. The first population distribution heat map data includes the number of people corresponding to each initial unit area in the target area, where the initial unit area is determined based on an initial resolution, and the target resolution is greater than the initial resolution. The third population distribution heat map data includes the number of people corresponding to each unit area in the target area. The module also filters the third population distribution heat map data of the target area based on the first-level population weight data of the target area to obtain a first filtering result. Finally, it acquires the regional subdivision type data of the target sub-regions within the target area. The regional subdivision type data includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region. The target sub-region is a unit area that is adjacent to the target region and corresponds to a specified regional type. The regional subdivision type data is a further subdivision of the sub-regions of the specified regional type. Based on the regional subdivision type data of the target sub-region, the secondary population weight data of the target sub-region is determined, wherein the secondary population weight data includes the population weight value corresponding to each unit area in the target sub-region. Based on the secondary population weight data of the target sub-region, the first filtering result is filtered to obtain the second population distribution heat map data of the target region.

9. The apparatus according to claim 8, characterized in that, The at least one regional attribute that has an impact on the population includes: regional type and number of building floors.

10. The apparatus according to claim 8 or 9, characterized in that, The weight determination module is used for: Based on the attribute values ​​of at least one regional attribute corresponding to each unit area in the target area, and the correspondence between the attribute values ​​of the regional attributes and the population weight values, the population weight value corresponding to each unit area in the target area is determined. For each initial unit region in the target region, the population weight values ​​corresponding to all unit regions in the initial unit region are normalized to obtain the normalized population weight value corresponding to each unit region. The normalized population weight values ​​corresponding to all unit areas in the target area are determined as the first-level population weight data of the target area.

11. The apparatus according to claim 10, characterized in that, The correspondence between the attribute value of the regional attribute and the population weight value is defined by the calculation formula for the attribute value of the regional attribute and the population weight value. The calculation formula is as follows: ； in, , ... Each is a unit region within the target region. The attribute value of each regional attribute in the text. For the unit region Population weight value, , ... and It is a constant.

12. The apparatus according to claim 8, characterized in that, The resolution enhancement module is used for: Based on the type value corresponding to each unit area in the target sub-region and the correspondence between the type value and the population weight value, the population weight value corresponding to each unit area in the target sub-region is determined. For each unit region in the target sub-region, based on the population weight values ​​corresponding to multiple other unit regions outside the unit region in the target sub-region, and the distances between the unit region and the multiple other unit regions, the population weight influence values ​​of the multiple other unit regions on the unit region are determined. Based on the population weight influence values ​​of the multiple other unit regions on the unit region, the population weight value corresponding to the unit region is adjusted to obtain the adjusted population weight value of the unit region. For each initial unit region in the target sub-region, the adjusted population weight values ​​corresponding to all unit regions in the initial unit region are normalized to obtain the normalized population weight value corresponding to each unit region. The normalized population weight values ​​corresponding to all unit regions in the target sub-region are determined as the secondary population weight data of the target sub-region.

13. The apparatus according to claim 12, characterized in that, The resolution enhancement module is used for: Based on the secondary population weight data of the target sub-region, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the second filtering result; The portions of the second filtering result and the first filtering result that do not correspond to the target sub-region are determined as the second population distribution thermal data of the target region.

14. The apparatus according to claim 13, characterized in that, The target sub-region is a multi-story building area; the regional subdivision type data includes sub-region subdivision type data corresponding to each layer, and the sub-region subdivision type data corresponding to each layer includes the type value of the regional subdivision type corresponding to each unit area in the target sub-region at that layer; the secondary population weight data includes sub-population weight data corresponding to each layer, and the sub-population weight data corresponding to each layer includes the population weight value corresponding to each unit area in the target sub-region at that layer. The resolution enhancement module is used for: For each layer, based on the sub-population weight data of the layer, the target filtering window size, and the target window sliding step size, the part of the first filtering result corresponding to the target sub-region is filtered again to obtain the filtering result corresponding to the layer. The second filtering result is determined by combining the filtering results corresponding to each layer.

15. A computer device, characterized in that, The computer device includes a memory and a processor, the memory being used to store computer instructions; The processor executes computer instructions stored in the memory to cause the computer device to perform the method according to any one of claims 1 to 7.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program code, which, when executed by a computer device, performs the method according to any one of claims 1 to 7.

Images