A data processing method, apparatus and device

Abstract

The application discloses a data processing method, device and equipment, and relates to the technical field of data processing, to solve the problem that the existing POI data collection mode needs manual participation and cannot obtain the center point coordinates and contour information of POIs. The method comprises the following steps: acquiring dot matrix data of a target point of interest (POI); converting each point cloud coordinate in the dot matrix data into a geocentric coordinate system coordinate; determining a planar area mapped by the geocentric coordinate system coordinate of the dot matrix data; determining the contour information of the target POI according to the position distribution of the dot matrix data in the planar area, and determining the center point coordinate of the planar area as the center point coordinate of the target POI. The application can achieve the technical effect of obtaining the center point coordinate and the contour information of the target POI without manual participation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data processing method, apparatus, and device. Background Technology

[0002] Currently, the latest method for Points of Interest (POI) data providers to collect POI data is crowdsourced data updates. This method mainly involves mobilizing users to collect POI data. Although this method can collect POI data in real time, the collection process is relatively complex, requires manual participation, and usually only obtains information such as POI name, address, phone number, and image. Summary of the Invention

[0003] This invention provides a data processing method, apparatus, and device to solve the problem that existing POI data acquisition methods require manual intervention and cannot obtain the center point coordinates and contour information of the POI.

[0004] In a first aspect, embodiments of the present invention provide a data processing method, including:

[0005] Obtain the raster data of the target point of interest (POI);

[0006] Convert the coordinates of each point cloud in the matrix data to geocentric coordinates.

[0007] Determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data;

[0008] Based on the positional distribution of the dot matrix data in the planar region, the contour information of the target POI is determined, and the center point coordinates of the planar region are determined as the center point coordinates of the target POI.

[0009] Optionally, determining the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data includes:

[0010] Map the geocentric coordinates of the dot matrix data onto a nine-square grid in a plane.

[0011] For the first grid in the planar nine-grid that only maps a portion of the point cloud in the point matrix data, the first grid is divided into nine smaller grids.

[0012] The step of determining the contour information of the target POI based on the distribution of the dot matrix data in the planar region, and determining the center point coordinates of the planar region as the center point coordinates of the target POI, includes:

[0013] Based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids, the contour information of the target POI is determined;

[0014] The coordinates of the center point of the planar nine-square grid are determined as the coordinates of the center point of the target POI.

[0015] Optionally, the method further includes:

[0016] Obtain a panoramic image of the target POI;

[0017] Text recognition is performed on the panoramic image to obtain at least one of the target POI's name and telephone information.

[0018] Optionally, the text recognition of the panoramic image includes:

[0019] Text recognition is performed on the panoramic image to obtain the text information in the panoramic image;

[0020] The text information is filtered.

[0021] Based on the structural characteristics of telephone data, identify telephone information in filtered text information;

[0022] Based on the POI name structure characteristics, identify the name information in the remaining text information, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0023] Optionally, identifying the name information in the remaining text information according to the POI name structure features includes:

[0024] Identify the clustering of the remaining text information in the panoramic image;

[0025] If the remaining text information appears as a cluster in the panoramic image, then the remaining text information is determined to be the target text information.

[0026] When the remaining text information is presented as multiple clusters in the panoramic image, the target cluster among the multiple clusters is determined, and the text information at the target cluster is determined as the target text information;

[0027] Based on the POI name structure features, identify the name information in the target text information.

[0028] Optionally, when the remaining text information is presented as multiple clusters in the panoramic image, determining the target cluster among the multiple clusters includes:

[0029] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0030] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0031] Optionally, identifying the name information in the target text information according to the POI name structure features includes:

[0032] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0033] The identified name elements are combined in a preset order to obtain name information.

[0034] Optionally, after determining the center point coordinates of the planar region as the center point coordinates of the target POI, the method further includes:

[0035] The address information of the target POI is determined based on the center point coordinates of the target POI.

[0036] Optionally, determining the address information of the target POI based on the center point coordinates of the target POI includes:

[0037] Based on the center point coordinates of the target POI and the pre-constructed administrative region data, the administrative location information of the target POI is determined.

[0038] Optionally, after determining the administrative location information of the target POI, the method further includes:

[0039] Determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0040] Determine the road point on each of the N roads that is closest to the coordinates of the center point, thus obtaining N road points;

[0041] The target road containing the road point that is closest to the coordinates of the center point among the N road points is determined as the road information where the target POI is located.

[0042] Optionally, after determining the target road containing the road point among the N road points that is closest to the coordinates of the center point as the road information where the target POI is located, the method further includes:

[0043] Obtain information on M house numbers containing the target road, where M is an integer greater than or equal to 1;

[0044] The doorplate information that is closest to the center point coordinates of the target POI among the M doorplate information is determined as the doorplate information of the target POI.

[0045] Optionally, determining the target address information among the M address information that is closest to the center point coordinates of the target POI as the address information of the target POI includes:

[0046] Determine a first distance between a first road point and the center point coordinates of the target POI, and a second distance between a second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0047] If the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance, then the target address information is determined to be the address information of the target POI.

[0048] Secondly, embodiments of the present invention also provide a data processing apparatus, comprising:

[0049] The first acquisition module is used to acquire the dot matrix data of the target point of interest (POI);

[0050] The conversion module is used to convert the coordinates of each point cloud in the point cloud data into geocentric coordinates.

[0051] The first determining module is used to determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data;

[0052] The second determining module is used to determine the contour information of the target POI based on the positional distribution of the dot matrix data in the planar region, and to determine the center point coordinates of the planar region as the center point coordinates of the target POI.

[0053] Optionally, the first determining module includes:

[0054] A mapping unit is used to map the geocentric coordinates of the dot matrix data onto a planar nine-square grid.

[0055] A partitioning unit is used to divide the first grid in the planar nine-grid into nine smaller grids that only map a portion of the point cloud from the point matrix data.

[0056] The second determining module includes:

[0057] The first determining unit is used to determine the contour information of the target POI based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids;

[0058] The second determining unit is used to determine the center point coordinates of the planar nine-square grid as the center point coordinates of the target POI.

[0059] Optionally, the data processing apparatus further includes:

[0060] The second acquisition module is used to acquire a panoramic image of the target POI;

[0061] The recognition module is used to perform text recognition on the panoramic image to obtain at least one of the name and telephone information of the target POI.

[0062] Optionally, the identification module includes:

[0063] The first recognition submodule is used to perform text recognition on the panoramic image to obtain text information in the panoramic image;

[0064] The filtering submodule is used to filter the text information.

[0065] The second recognition submodule is used to identify telephone information in the filtered text information according to the characteristics of telephone data structure.

[0066] The third identification submodule is used to identify the name information in the remaining text information according to the POI name structure characteristics, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0067] Optionally, the third identification submodule includes:

[0068] The first recognition unit is used to identify the aggregation of the remaining text information in the panoramic image;

[0069] The third determining unit is used to determine the remaining text information as target text information when the remaining text information is presented as a cluster in the panoramic image;

[0070] The fourth determining unit is used to determine the target cluster among the multiple clusters when the remaining text information is presented as multiple clusters in the panoramic image, and to determine the text information at the target cluster as the target text information;

[0071] The second identification unit is used to identify the name information in the target text information according to the POI name structure features.

[0072] Optionally, the fourth determining unit is used for:

[0073] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0074] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0075] Optionally, the second identification unit is used for:

[0076] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0077] The identified name elements are combined in a preset order to obtain name information.

[0078] Optionally, the data processing apparatus further includes:

[0079] The third determining module is used to determine the address information of the target POI based on the center point coordinates of the target POI.

[0080] Optionally, the third determining module is used to determine the administrative location information of the target POI based on the center point coordinates of the target POI and pre-constructed administrative region data.

[0081] Optionally, the data processing apparatus further includes:

[0082] The fourth determining module is used to determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0083] The fifth determining module is used to determine the road point on each of the N roads that is closest to the coordinates of the center point, thereby obtaining N road points;

[0084] The sixth determining module is used to determine the target road where the road point closest to the center point's coordinates is located among the N road points as the road information where the target POI is located.

[0085] Optionally, the data processing apparatus further includes:

[0086] The third acquisition module is used to acquire M house number information containing the target road, where M is an integer greater than or equal to 1;

[0087] The seventh determining module is used to determine the target door information that is closest to the center point coordinates of the target POI among the M door information information as the door information of the target POI.

[0088] Optionally, the seventh determining module includes:

[0089] The fifth determining unit is used to determine a first distance between the first road point and the center point coordinates of the target POI, and a second distance between the second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0090] The sixth determining unit is used to determine the target address information as the address information of the target POI when the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance.

[0091] Thirdly, embodiments of the present invention also provide a data processing device, including: a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the data processing method described in the first aspect above.

[0092] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the data processing method described in the first aspect above.

[0093] In this embodiment of the invention, point cloud data of a target point of interest (POI) is acquired; the coordinates of each point cloud in the point cloud data are converted into geocentric coordinates; the planar region mapped by the geocentric coordinates of the point cloud data is determined; based on the positional distribution of the point cloud data in the planar region, the contour information of the target POI is determined, and the center point coordinates of the planar region are determined as the center point coordinates of the target POI. Thus, the center point coordinates and contour information of the target POI can be obtained in this way without manual intervention. Attached Figure Description

[0094] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0095] Figure 1 This is one of the flowcharts of the data processing method provided in the embodiments of the present invention;

[0096] Figure 2 This is a schematic diagram illustrating the determination of the center point coordinates and outline of a target POI using a dynamic nine-square grid, as provided in an embodiment of the present invention.

[0097] Figure 3 This is the second flowchart of the data processing method provided in the embodiments of the present invention;

[0098] Figure 4 This is a structural diagram of the data processing apparatus provided in an embodiment of the present invention;

[0099] Figure 5 This is a structural diagram of the data processing device provided in an embodiment of the present invention. Detailed Implementation

[0100] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0101] To make the embodiments of the present invention clearer, the relevant concepts involved in the embodiments of the present invention will be explained below:

[0102] Points of Interest (POIs) are the core data of location-based services. Each POI typically contains four pieces of information: name, category, coordinates, and classification. Comprehensive POI information is essential for enriching navigation maps. Timely POI information can remind users of road branching and detailed information about surrounding buildings, and also make it easy to find the places you need during navigation, allowing you to choose the most convenient and smooth route for route planning.

[0103] Currently, POI data providers primarily collect POI data through four methods: map provider data version updates, collaborative data updates, intelligence data updates, and crowdsourced data updates. Map provider data version updates are the standard method, involving map providers with Class A surveying and mapping qualifications for navigation electronic maps who collect POI data periodically (usually quarterly) using floating cars. Collaborative data updates are also a standard method, acquiring unique POI data from partners (data content providers) through partnerships and updating it periodically. Intelligence data updates are another standard method, collecting POI data via internet or mobile internet web crawlers. Crowdsourced data updates represent a newer model, involving encouraging users to collect POI data.

[0104] Of the above technical solutions, only the crowdsourcing data update method can collect POI data in real time, but the collection mode is complex, resulting in extremely low user participation. However, in this method, POI providers require users to provide information such as the POI's name, address, phone number, coordinates, and image before it can be considered a valid POI for data collection.

[0105] In summary, existing POI data acquisition solutions still have the following technical problems:

[0106] 1) Although crowdsourced data updates are highly up-to-date, fields such as name, address, and phone number all require manual input. This method increases the difficulty of user operation and reduces operability and user satisfaction.

[0107] 2) Unable to obtain the center point coordinates and outline information of the POI.

[0108] See Figure 1 , Figure 1 This is a flowchart of the data processing method provided in the embodiments of the present invention, such as... Figure 1 As shown, it includes the following steps:

[0109] Step 101: Obtain the dot matrix data of the target point of interest (POI).

[0110] A Point of Interest (POI) is a geospatial information point. A POI can be a residential area, park, company, shop, bus stop, etc., displayed on an electronic map. The target POI mentioned above can refer to the POI for which POI data needs to be collected; for example, it could be a merchant A whose data is to be collected.

[0111] Point cloud data is a dataset composed of point cloud data. Point cloud refers to the spatial data points on the surface of an object obtained using a 3D laser scanner. Point cloud data includes the 3D position information (including coordinates and Euler angles) of multiple points on the object's reflective surface in the lidar coordinate system and the reflection intensity information of multiple points related to the object's material.

[0112] In this step, the three-dimensional point cloud matrix of the map collected by the autonomous driving map collection vehicle can be obtained, and the point cloud data of non-target points of interest (POIs) in the matrix data can be filtered out to obtain the matrix data of the target POI.

[0113] Specifically, the acquisition of the target POI's dot matrix data may include: scanning the area where the target POI is located using a 3D laser scanner to obtain 3D point cloud data of the area where the target POI is located; selecting point clouds whose reflection intensity exceeds the preset intensity threshold from the 3D point cloud data of the area where the target POI is located according to a preset intensity threshold; and determining the dataset composed of all point clouds whose reflection intensity exceeds the preset intensity threshold as the dot matrix data of the target POI.

[0114] In some embodiments, the point cloud data of the target POI can also be determined from the point cloud data collected by the lidar by means of point cloud clustering, point cloud feature extraction, calculation of the geometric relationship of the point cloud, comparison of the reflection intensity of the point cloud, etc.

[0115] Step 102: Convert the coordinates of each point cloud in the point matrix data into geocentric coordinates.

[0116] After obtaining the dot matrix data of the target POI, the coordinate data of each point cloud can be extracted from the dot matrix data of the target POI. In order to complete the coordinate transformation, the Euler angles of each point cloud, namely heading φ, pitch θ, and roll γ, can also be obtained. The coordinate data of the point cloud can be calculated by the Beidou terminal.

[0117] The coordinates of each point cloud in the point cloud data obtained in step 101 are generally three-dimensional coordinates under the lidar coordinate system. However, the purpose of obtaining POI data is to build an electronic map and provide navigation for autonomous driving. Therefore, it is necessary to convert the relative coordinates of the point cloud into actual coordinates in the real world, that is, to convert the relative coordinates of the point cloud into geocentric coordinate system (World Geodetic System 1984, WGS84) coordinates.

[0118] Among them, the WGS84 coordinate system is a spatial rectangular coordinate system established with the Earth's center of mass as the origin, or a geodetic coordinate system established with the Earth's ellipsoid, whose center coincides with the Earth's center of mass, as the reference surface.

[0119] Specifically, point cloud coordinates can be converted into WGS84 coordinates using rotation and translation matrices based on the relative coordinates and Euler angles of the point cloud.

[0120] First, the rotation matrix R, converted to the WGS84 coordinate system, can be calculated based on the Euler angles of the point cloud and Formula 1. Assume the relative coordinates of each point cloud are (X... i Y i Z i The three Euler angles are the heading angles. Pitch (θ) and roll (γ). Formula 1 is as follows:

[0121]

[0122]

[0123] Then, according to Formula 2, the transformed WGS84 coordinates (X) can be obtained. i ′, Y i ′, Z i Formula 2 is as follows:

[0124]

[0125] Where T is the translation matrix for transforming the relative coordinates of the point cloud from the local coordinate system to the WGS84 coordinate system.

[0126] Step 103: Determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data.

[0127] In this step, the point data in the transformed geocentric coordinate system, i.e., WGS84 coordinates, can be mapped to a planar region, such as a rectangular region, thereby obtaining the projection of the point data in the planar region. It should be noted that the mapped planar region can be bounded by the edge point cloud of the point data; that is, the planar region can be the smallest region enclosing the point data.

[0128] Step 104: Based on the positional distribution of the dot matrix data in the planar region, determine the contour information of the target POI, and determine the center point coordinates of the planar region as the center point coordinates of the target POI.

[0129] In this step, the positional distribution of each point cloud in the dot matrix data in the planar area can be determined based on the projection of the dot matrix data in the planar area. Then, the contour information of the dot matrix data in the planar area can be determined based on the positional distribution, and the contour information can be determined as the contour information of the target POI.

[0130] Furthermore, since the relative positions of each point cloud remain unchanged after the dot matrix data is mapped to the planar region, the center point of the planar region can be determined, which is the center point of the dot matrix data. Therefore, the coordinates of the center point of the dot matrix data mapped to the planar region can be determined as the center point coordinates of the target POI.

[0131] To facilitate the storage and retrieval of POI data, the POI data, such as the center point coordinates and contour information of the target POI determined in the above steps, can be integrated to generate POI data corresponding to the target POI.

[0132] In this way, the outline information and center point coordinate information of the target POI can be obtained without the need for manual intervention by the user, which is highly efficient.

[0133] Optionally, step 103 includes:

[0134] Map the geocentric coordinates of the dot matrix data onto a nine-square grid in a plane.

[0135] For the first grid in the planar nine-grid that only maps a portion of the point cloud in the point matrix data, the first grid is divided into nine smaller grids.

[0136] Step 104 includes:

[0137] Based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids, the contour information of the target POI is determined;

[0138] The coordinates of the center point of the planar nine-square grid are determined as the coordinates of the center point of the target POI.

[0139] In one implementation, a planar nine-square grid can be used to determine the center point coordinates and contour information of the target POI through dynamic division of the nine-square grid.

[0140] Specifically, the geocentric coordinates of the target POI's point cloud data can be plotted in a nine-square grid, which means mapping the geocentric coordinates of the point cloud data onto a planar region and dividing the planar region into nine squares. This allows us to determine the distribution of the geocentric coordinates of each point cloud in the point cloud data within the nine-square grid.

[0141] For each cell in the planar nine-square grid, if a cell only maps to partial point cloud data, that cell can be further divided, such as into nine equal-proportion smaller cells, until each cell or smaller cell contains all point cloud data, or none contains any point cloud data at all; that is, there is no cell or smaller cell that only contains partial point cloud data. This method is also known as the planar dynamic nine-square grid method.

[0142] In this way, based on the distribution of the point data in the planar nine-square grid and the nine smaller squares, the contour information of the point data in the planar nine-square grid can be determined. This contour information is the contour information of the target POI. Specifically, the contour of the edge squares or edge smaller squares in the planar nine-square grid containing point clouds can be determined as the contour of the target POI. Furthermore, the coordinates of the center point of the planar nine-square grid can be determined as the coordinates of the center point of the target POI.

[0143] For example, such as Figure 2 As shown, the dynamic nine-square grid is a large square composed of nine squares with equal side lengths. If only a portion of a square contains the point cloud data of the target POI, this square is further subdivided into nine smaller squares. Then, the coordinates of the center point of the nine-square grid can be calculated, or the coordinates of the point cloud located at or closest to the center point of the nine-square grid can be determined as the coordinates of the center point. These center point coordinates are the coordinates of the center point of the target POI. Figure 2The pentagram shown is located at the center point of the target POI.

[0144] The outline formed by the grid cells or zigzag cells of a nine-square grid containing point clouds is the outline of the target POI, such as... Figure 2 As shown, the black grid represents the shape outline of the target POI defined by the dynamic nine-grid design, the thick black line represents the outline of the target POI, and the point cloud coordinates on the outline represent the outline coordinates of the target POI.

[0145] In this way, by using a dynamic nine-square grid in this implementation method, the center point coordinates and contour information of the target POI can be accurately and quickly determined, reducing the amount of calculation.

[0146] Optionally, the method further includes:

[0147] Obtain a panoramic image of the target POI;

[0148] Text recognition is performed on the panoramic image to obtain at least one of the target POI's name and telephone information.

[0149] In one embodiment, the name, phone number, and other information of the target POI can also be extracted by acquiring a panoramic image of the target POI and recognizing the text in the panoramic image.

[0150] The above-mentioned acquisition of the panoramic image of the target POI can be based on the specific acquisition time of the dot matrix data of the target POI, and the acquisition of the panoramic image captured by the high-precision camera at that time.

[0151] Specifically, the above-mentioned text recognition of the panoramic image can be performed by recognizing the text in the panoramic image, and then extracting the name and telephone information of the target POI from the recognized text based on the compositional features of the telephone number, such as recognizing the mobile phone number based on the first three digits and the total number of digits of a string of numbers, and the compositional features of the point of interest name, such as recognizing the merchant name based on the brand name, category name, etc.

[0152] To improve text recognition efficiency, a scene text recognition method based on dimensionality reduction can be used to identify the panoramic image. This method identifies text appearing within the scene, such as road signs, billboards, and numbers on athletes' jerseys. Scene text recognition is more complex than traditional manual text extraction but can quickly and effectively identify identification information and contact details within the image.

[0153] Thus, through this implementation method, the name and / or telephone information of the target POI can also be obtained, making the target POI data more detailed and comprehensive.

[0154] Optionally, the text recognition of the panoramic image includes:

[0155] Text recognition is performed on the panoramic image to obtain the text information in the panoramic image;

[0156] The text information is filtered.

[0157] Based on the structural characteristics of telephone data, identify telephone information in filtered text information;

[0158] Based on the POI name structure characteristics, identify the name information in the remaining text information, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0159] In this embodiment, to more accurately identify useful POI information in the panoramic image, text recognition can be performed on the panoramic image to obtain text information, and then the text information can be filtered to remove useless or meaningless text information.

[0160] Specifically, secondary POI information in the panoramic image can be filtered out based on the text clarity, i.e., text color difference, to obtain filtered text information. For example, text located at the corners of the panoramic image or text with a relatively scattered distribution can be filtered out.

[0161] Then, for the filtered text information, the presence of phone numbers can be identified based on the characteristics of telephone data structure. If present, the phone number information can be extracted. For example, phone numbers can be identified based on their number of digits. Phone numbers are categorized as landline, mobile, and toll-free. For landline identification, the area code and number of digits can be determined based on the user's location. For mobile numbers, the number of digits and the first three digits can be used to identify the mobile number. For toll-free numbers, the number of digits and the first digit can be used to identify the phone number.

[0162] After recognizing the telephone information, the remaining text information can be further recognized. Specifically, the POI name information in the remaining text information can be recognized according to the POI name structure features. Taking the merchant name as an example, the merchant name structure features can be composed of a name prefix, brand name, category words, and name suffix. Therefore, the name prefix, brand name, category words, and name suffix in the remaining text information can be recognized in sequence according to the merchant name structure features to obtain the recognized merchant name information.

[0163] In this way, the POI information in the panoramic image can be quickly and accurately identified through this implementation method.

[0164] Optionally, identifying the name information in the remaining text information according to the POI name structure features includes:

[0165] Identify the clustering of the remaining text information in the panoramic image;

[0166] If the remaining text information appears as a cluster in the panoramic image, then the remaining text information is determined to be the target text information.

[0167] When the remaining text information is presented as multiple clusters in the panoramic image, the target cluster among the multiple clusters is determined, and the text information at the target cluster is determined as the target text information;

[0168] Based on the POI name structure features, identify the name information in the target text information.

[0169] In one implementation, considering that POI name information is usually a series of consecutive sentences or located in the same field, and the character size in the same text field or sentence is the same, but the text size between different sentences or fields may be different, and the POI name is often the largest font or the middle line of text, therefore, by identifying the aggregation of the remaining text information in the panoramic image, some invalid POI name information can be filtered out, and then valid POI name information can be extracted from it.

[0170] Specifically, the clustering of the remaining text information in the panoramic image can be identified to determine whether the remaining text information forms a cluster or multiple clusters in the panoramic image. If the remaining text information appears as a cluster in the panoramic image, it can be directly identified as target text information containing POI names, and the name information in the target text information can be identified according to the structural characteristics of POI names. If the remaining text information appears as multiple clusters in the panoramic image, the target cluster most likely to contain POI name information can be identified from the multiple clusters according to the presentation characteristics of POI name information in the panoramic image, such as merchant names usually occupying the largest area in billboards or being located in the most central position. Finally, the POI name information can be identified from the text information at the target cluster according to the structural characteristics of POI names.

[0171] In this way, the POI name information can be identified more accurately from the panoramic image through this implementation method.

[0172] Optionally, when the remaining text information is presented as multiple clusters in the panoramic image, determining the target cluster among the multiple clusters includes:

[0173] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0174] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0175] In this embodiment, when it is determined that the remaining text information appears as multiple clusters in the panoramic image, the target cluster can be further determined based on the size of the multiple clusters. Specifically, if the size of each cluster is inconsistent, considering that the POI name information usually occupies the largest block, the largest cluster can be determined as the target cluster. If the size of each cluster is consistent, considering that the POI name information is usually located in the middle, the cluster located at the center of the multiple clusters can be determined as the target cluster. More specifically, the panoramic image of the target POI can be segmented using a nine-square grid method, and then the horizontal text cluster closest to the center grid cell can be selected as the target cluster, while the remaining clusters are filtered out.

[0176] In this way, the location of POI name information in the image can be determined more accurately, thereby helping to accurately identify POI name information.

[0177] Optionally, identifying the name information in the target text information according to the POI name structure features includes:

[0178] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0179] The identified name elements are combined in a preset order to obtain name information.

[0180] In one implementation, a POI name element thesaurus can be pre-established. For example, it can be built based on the minimum constituent elements of a POI name, such as a list of category terms, a place name database, or a brand thesaurus. The minimum constituent elements can be local place names, foreign place names, name prefixes, brand names, name category terms, name suffixes, etc. As shown in Table 1 below, the established POI name element thesaurus can include the following name elements:

[0181] Table 1 Minimum Components of the POI Name Field

[0182]

[0183] In this way, based on a pre-established POI name element lexicon, name elements in the target text information can be identified, such as the pure brand name and name category words. The identified name elements can then be combined in a preset order to obtain the name information. The preset order can be the order of local place name, non-local place name, name prefix, pure brand name, name category word, and name suffix. For example, after identifying that the target text information includes the local place name "Changsha", the brand name "field" and the category word "cake", the target POI name information can be obtained as "Changsha Field Cake" by combining the place name, brand name, and category word in that order.

[0184] It should be noted that when combining name elements, machine learning can be used to continuously accumulate the combination of each smallest constituent element, thereby resolving the accurate POI name.

[0185] In this way, the POI name information can be accurately and quickly identified through this implementation method.

[0186] Optionally, after step 104, the method further includes:

[0187] The address information of the target POI is determined based on the center point coordinates of the target POI.

[0188] In one implementation, the geographical location information of the target POI on the map can also be determined based on the center point coordinates of the target POI, thereby obtaining the address information of the target POI, such as determining the province and city to which the target POI belongs. This allows for the acquisition of more comprehensive POI information.

[0189] Optionally, determining the address information of the target POI based on the center point coordinates of the target POI includes:

[0190] Based on the center point coordinates of the target POI and the pre-constructed administrative region data, the administrative location information of the target POI is determined, wherein the administrative location information includes at least one of the following: province, city, district / county, and township / street information.

[0191] In order to obtain accurate administrative location information of the target POI at all levels, administrative region data at all levels can be pre-constructed, such as province outline data, city outline data, district / county outline data, township / street outline data, etc. Then, based on the province, city, district / county, township / street, etc., where the center point coordinates of the target POI fall, the administrative location information of the target POI can be determined. For example, the province corresponding to the province outline containing the center point coordinates of the target POI is determined as the province to which the target POI belongs; the city corresponding to the city outline containing the center point coordinates of the target POI is determined as the city to which the target POI belongs; the district / county corresponding to the district / county outline containing the center point coordinates of the target POI is determined as the district / county to which the target POI belongs; and the township / street corresponding to the township / street outline containing the center point coordinates of the target POI is determined as the township / street to which the target POI belongs.

[0192] In this way, the province, city, district / county, and township / street information of the target POI can be obtained through this implementation method, making the target POI information more detailed.

[0193] Optionally, after determining the administrative location information of the target POI, the method further includes:

[0194] Determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0195] Determine the road point on each of the N roads that is closest to the coordinates of the center point, thus obtaining N road points;

[0196] The target road containing the road point that is closest to the coordinates of the center point among the N road points is determined as the road information where the target POI is located.

[0197] In one implementation, a circle can be defined with the center coordinates of the target POI as the center and a preset distance as the radius, such as 2 kilometers. This allows the determination of road information within the circle. Specifically, it can be based on pre-acquired road point sequence data to determine the road point sequence within the circle and the roads corresponding to these road point sequences. A road may have one or more road points. The circle is the preset range, thus determining N roads within the preset range and the road point information on the N roads.

[0198] Then, for each road, the distance between each road point and the coordinates of the center point can be calculated separately, thereby determining the road point on each road that is closest to the coordinates of the center point. In this way, N such road points can be determined for N roads.

[0199] Next, based on the distance between each of the above N road points and the coordinates of the center point, the road point with the shortest distance can be determined as the target road point, and the road where the target road point is located can be determined as the road information where the target POI is located.

[0200] In this way, the road name information of the target POI can be further determined through this implementation method, making the POI information more complete.

[0201] Optionally, after determining the target road containing the road point among the N road points that is closest to the coordinates of the center point as the road information where the target POI is located, the method further includes:

[0202] Obtain information on M house numbers containing the target road, where M is an integer greater than or equal to 1;

[0203] The doorplate information that is closest to the center point coordinates of the target POI among the M doorplate information is determined as the doorplate information of the target POI.

[0204] In this embodiment, the address information of the target POI can be further determined. Specifically, all road address information containing the target road can be obtained from a pre-established road database, resulting in M ​​address information entries. Then, the distance between the coordinates corresponding to each address information entry and the coordinates of the center point can be determined, such as k1, k2, ..., k j , ..., k m Therefore, the address information with the shortest distance to the center point coordinates can be determined as the target address information. Let k be an example. j If the shortest distance is found, then k can be determined. j The corresponding address information is the target address information, which can be identified as the address information of the target POI.

[0205] In this way, the specific road address information of the target POI can be further determined through this implementation method, making the POI information more complete.

[0206] Optionally, determining the target address information among the M address information that is closest to the center point coordinates of the target POI as the address information of the target POI includes:

[0207] Determine a first distance between a first road point and the center point coordinates of the target POI, and a second distance between a second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0208] If the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance, then the target address information is determined to be the address information of the target POI.

[0209] In one implementation, in order to obtain more accurate road address information for the target POI, the target address information can be verified as the road address information of the target POI based on the two road points that are closest to and second closest to the center point among the N road points.

[0210] Specifically, after determining the N road points, the first road point closest to the center point coordinates and the second road point second closest to the center point coordinates can be further determined. The distance between the first road point and the center point coordinates is denoted as the first distance l1, and the distance between the second road point and the center point coordinates is denoted as the second distance l2, where l2... <l1。

[0211] After determining the target address information among the M address information that is closest to the center point coordinates of the target POI, we can first determine the distance between the target address information and the center point coordinates, let's assume it's k. j The magnitude of the distance l2 from the second distance, if k j If k ≤ l2, then the target address information can be determined and recorded as the address information of the target POI. j If the value is greater than 12, it can be considered that the optimal road address information cannot be obtained at present, and therefore the target address information should not be determined as the address information of the target POI.

[0212] In this way, the road address information of the target POI can be accurately determined, avoiding errors in address information determination.

[0213] To facilitate a better understanding of the embodiments of the present invention, the following description is provided in conjunction with... Figure 3 The specific implementation methods of the embodiments of the present invention are illustrated by examples:

[0214] Against the backdrop of the development of high-precision maps for autonomous driving, data collection vehicles are equipped with instruments such as 3D laser scanners, high-precision cameras, and BeiDou positioning terminals, which enable them to generate high-precision maps and POI information.

[0215] Step 31: Select the merchants to be collected based on the laser scanning dot matrix.

[0216] An autonomous map-collecting vehicle is used to acquire 3D point cloud data and panoramic images of the map. First, the point cloud's location is determined based on the reflection intensity recorded in the 3D point cloud. Then, based on the specific time of data acquisition, panoramic images from a high-precision camera and latitude and longitude coordinates calculated from a BeiDou terminal are obtained. This point cloud is recorded as the point cloud of merchant A, and merchant A is the target POI.

[0217] Step 32: Obtain the point cloud information of each point in Merchant A's dot matrix.

[0218] Obtain the relative coordinates (X, Y, X) of each point cloud in point matrix A. i Y i Z i ) and three Euler angles, namely heading (φ), pitch (θ) and roll (γ).

[0219] Step 33: Convert the coordinates of each point cloud in Merchant A's dot matrix to WGS84 coordinates.

[0220] The rotation matrix R, converted to the WGS84 coordinate system, can be calculated using the Euler angles obtained in step 32 and the aforementioned formula 1.

[0221] Then input the point cloud coordinates (X) i Y i Z i Given the rotation matrix R and the translation matrix T, the real-world WGS84 coordinates (X, Y, Z) can be obtained according to Formula 2 above. i ′, Y i ′, Z i ′).

[0222] Step 34: Generate the center point coordinates and outline information of merchant A.

[0223] like Figure 2 As shown, the point coordinates obtained in step 33 are placed into a planar dynamic nine-square grid, assuming the black squares represent the area where merchant A is located. If some squares in the nine-square grid only fall into part of the point cloud, these squares are further divided into nine smaller squares; the outline formed by the squares containing the point cloud is the outline of merchant A, as shown. Figure 2 As shown, the black grid represents the shape of merchant A defined by the dynamic nine-square grid, and the thick black line represents the outline of merchant A; the coordinates of the point cloud on the outline are the outline coordinates of merchant A, and the coordinates of the center point of the dynamic nine-square grid are the center point coordinates of merchant A (X). i ", Y i ", Z i ″).

[0224] Step 35: Obtain the POI name and phone number information based on the panoramic image of Merchant A.

[0225] Using a dimensionality reduction-based scene text recognition method, the name and phone number information of POIs are extracted from panoramic images.

[0226] Dimensionality reduction includes the following dimensions:

[0227] Step 5.1 Filter out meaningless information from the image.

[0228] To obtain valid name information for Points of Interest (POIs) in an image, meaningless information needs to be removed from the identified text. For example, other minor POI information in the image can be filtered out based on text clarity, i.e., text color difference.

[0229] Step 5.2 Identify and strip telephone information by the number of digits.

[0230] Due to the unique nature of POI (Point of Interest) phone numbers, phone information in images can be identified based on the number of digits. Phone information is categorized into: landline numbers, mobile phone numbers, and toll-free numbers. For landline numbers, the area code and number of digits can be used to identify the phone number based on the user's location. For mobile phone numbers, the number of digits and the first three digits can be used to identify the phone number. For toll-free numbers, the number of digits and the first digit can be used to identify the phone number.

[0231] Based on the above logic, identify the phone information from step 5.2 and complete the stripping process. If no phone information is found, proceed to step 5.3.

[0232] Step 5.3 Filter out invalid POI name information.

[0233] Analyze the remaining characters in the image and examine their clustering. If all remaining characters are clustered and have the same size, proceed to step 5.4. If the remaining characters are in multiple clusters, determine if the characters in each cluster have the same size. If they do not, retain only the cluster with the largest character size and proceed to step 5.4. If they do, use a 3x3 grid to divide the image, select the horizontal cluster closest to the center cell of the 3x3 grid, filter out the remaining clusters, and proceed to step 5.4.

[0234] Step 5.4 Identify the name information of the POI.

[0235] Based on the list of classification terms, the database of place names, and the database of brand terms, the POI name is identified from the remaining text. Since the POI name has certain characteristics, namely, the name is composed of the smallest elements of "local place name", "out-of-town place name", "name prefix", "pure brand name", "name classification term" and "name suffix", see Table 1 above for details.

[0236] By using machine learning, the system continuously accumulates combinations of the smallest elements and resolves the accurate POI name.

[0237] Step 36: Obtain the address information of the POI.

[0238] Specifically, the steps include the following:

[0239] Step 6.1 Obtain the merchant's province, city, district / county, and township / street information.

[0240] Based on the center point coordinates calculated in step 34, the province, city, district / county, and township / street information of merchant A are calculated using the province outline data, city outline data, district / county outline data, and township / street outline data, respectively.

[0241] Step 6.2 Obtain the name of the road where the merchant is located.

[0242] Draw a circle with a radius of 2 kilometers, centered on the center point coordinates of merchant A. Overlay the road point sequence data onto the drawn circle to identify all road point sequences contained within the circle. Connect the center point of merchant A to all road point sequences for each road, calculating the length of each line segment. Find and record the road point closest to the merchant on each road, along with its corresponding line segment length. Compare the shortest line segment lengths for each road, retaining the shortest segment l1 and the second shortest segment l2. Return the road name RD1 to which l1 belongs, as the road information for the location of merchant A.

[0243] Step 6.3 Obtain the street address where the merchant is located.

[0244] Based on the road name RD1 obtained in step 6.2, retrieve all road addresses containing RD1 and their corresponding coordinates from the road database. Connect the center point of merchant A to all road address points containing RD1, and calculate the lengths k1, k2, ..., k of each line segment. j , ..., k m When k j It is the shortest line segment, and k j When ≤l2, record k. j The address number in the corresponding road address point is the road address of the merchant's location, and proceed to step 6.4; when k j It is the shortest line segment, and k j When the value is >l2, this merchant cannot obtain the optimal road address information and proceeds to step 6.4.

[0245] Step 6.4 Generate merchant address information

[0246] The results of steps 6.1, 6.2, and 6.3 are combined to generate the address information of merchant A.

[0247] Step 37: Generate a complete POI record.

[0248] The results from steps 34, 5.2, 5.4 and 6.4 are combined to form the name, address, telephone number, center point coordinates and outline information of merchant A.

[0249] The above exemplary embodiments can achieve at least one of the following beneficial effects:

[0250] By innovating with 3D laser scanning technology and dynamic nine-square grid technology, the outline and center point coordinate information of POI are generated.

[0251] By using a scene text recognition method based on dimensionality reduction, the name and phone number of the Point of Interest (POI) can be extracted from the image.

[0252] By innovating the address generation algorithm based on POI coordinates, we can generate POI address information that best reflects the actual situation.

[0253] The data processing method of this invention involves: acquiring point cloud data of a target point of interest (POI); converting the coordinates of each point cloud in the point cloud data into geocentric coordinates; determining the planar region mapped by the geocentric coordinates of the point cloud data; determining the contour information of the target POI based on the positional distribution of the point cloud data within the planar region; and determining the center point coordinates of the planar region as the center point coordinates of the target POI. In this way, the center point coordinates and contour information of the target POI can be obtained without manual intervention.

[0254] This invention also provides a data processing apparatus. See [link to previous document]. Figure 4 , Figure 4 This is a structural diagram of the data processing apparatus provided in an embodiment of the present invention. Since the principle by which the data processing apparatus solves the problem is similar to the data processing method in this embodiment, the implementation of this data processing apparatus can be referred to the implementation of the method, and repeated details will not be described again.

[0255] like Figure 4 As shown, the data processing device 400 includes:

[0256] The first acquisition module 401 is used to acquire the dot matrix data of the target point of interest (POI);

[0257] The conversion module 402 is used to convert the coordinates of each point cloud in the point matrix data into geocentric coordinates.

[0258] The first determining module 403 is used to determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data;

[0259] The second determining module 404 is used to determine the contour information of the target POI based on the positional distribution of the dot matrix data in the planar region, and to determine the center point coordinates of the planar region as the center point coordinates of the target POI.

[0260] Optional, optional, the first determining module 403 includes:

[0261] A mapping unit is used to map the geocentric coordinates of the dot matrix data onto a planar nine-square grid.

[0262] A partitioning unit is used to divide the first grid in the planar nine-grid into nine smaller grids that only map a portion of the point cloud from the point matrix data.

[0263] The second determining module 404 includes:

[0264] The first determining unit is used to determine the contour information of the target POI based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids;

[0265] The second determining unit is used to determine the center point coordinates of the planar nine-square grid as the center point coordinates of the target POI.

[0266] Optionally, the data processing device 400 also includes:

[0267] The second acquisition module is used to acquire a panoramic image of the target POI;

[0268] The recognition module is used to perform text recognition on the panoramic image to obtain at least one of the name and telephone information of the target POI.

[0269] Optionally, the identification module includes:

[0270] The first recognition submodule is used to perform text recognition on the panoramic image to obtain text information in the panoramic image;

[0271] The filtering submodule is used to filter the text information.

[0272] The second recognition submodule is used to identify telephone information in the filtered text information according to the characteristics of telephone data structure.

[0273] The third identification submodule is used to identify the name information in the remaining text information according to the POI name structure characteristics, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0274] Optionally, the third identification submodule includes:

[0275] The first recognition unit is used to identify the aggregation of the remaining text information in the panoramic image;

[0276] The third determining unit is used to determine the remaining text information as target text information when the remaining text information is presented as a cluster in the panoramic image;

[0277] The fourth determining unit is used to determine the target cluster among the multiple clusters when the remaining text information is presented as multiple clusters in the panoramic image, and to determine the text information at the target cluster as the target text information;

[0278] The second identification unit is used to identify the name information in the target text information according to the POI name structure features.

[0279] Optionally, the fourth determining unit is used for:

[0280] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0281] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0282] Optionally, the second identification unit is used for:

[0283] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0284] The identified name elements are combined in a preset order to obtain name information.

[0285] Optionally, the data processing device 400 also includes:

[0286] The third determining module is used to determine the address information of the target POI based on the center point coordinates of the target POI.

[0287] Optionally, the third determining module is used to determine the administrative location information of the target POI based on the center point coordinates of the target POI and pre-constructed administrative region data.

[0288] Optionally, the data processing device 400 also includes:

[0289] The fourth determining module is used to determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0290] The fifth determining module is used to determine the road point on each of the N roads that is closest to the coordinates of the center point, thereby obtaining N road points;

[0291] The sixth determining module is used to determine the target road where the road point closest to the center point's coordinates is located among the N road points as the road information where the target POI is located.

[0292] Optionally, the data processing device 400 also includes:

[0293] The third acquisition module is used to acquire M house number information containing the target road, where M is an integer greater than or equal to 1;

[0294] The seventh determining module is used to determine the target door information that is closest to the center point coordinates of the target POI among the M door information information as the door information of the target POI.

[0295] Optionally, the seventh determining module includes:

[0296] The fifth determining unit is used to determine a first distance between the first road point and the center point coordinates of the target POI, and a second distance between the second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0297] The sixth determining unit is used to determine the target address information as the address information of the target POI when the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance.

[0298] The data processing apparatus provided in this embodiment of the invention can execute the above-described method embodiments, and its implementation principle and technical effect are similar, so it will not be described again here.

[0299] The data processing apparatus 400 of this embodiment acquires point cloud data of a target point of interest (POI); converts the coordinates of each point cloud in the point cloud data into geocentric coordinates; determines the planar region mapped by the geocentric coordinates of the point cloud data; determines the contour information of the target POI based on the positional distribution of the point cloud data in the planar region, and determines the center point coordinates of the planar region as the center point coordinates of the target POI. In this way, the center point coordinates and contour information of the target POI can be obtained without manual intervention.

[0300] This invention also provides a data processing device. Since the principle by which the data processing device solves the problem is similar to the data processing method in this invention, the implementation of this data processing device can be found in the implementation of the method, and repeated details will not be described again. Figure 5 As shown, the data processing device of this embodiment includes: a processor 500, configured to read a program from a memory 520 and execute the following processes:

[0301] Obtain the raster data of the target point of interest (POI);

[0302] Convert the coordinates of each point cloud in the matrix data to geocentric coordinates.

[0303] Determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data;

[0304] Based on the positional distribution of the dot matrix data in the planar region, the contour information of the target POI is determined, and the center point coordinates of the planar region are determined as the center point coordinates of the target POI.

[0305] Transceiver 510 is used to receive and send data under the control of processor 500.

[0306] Among them, Figure 5 In this context, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 500) and memory (memory 520). The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 510 may be multiple elements, including transmitters and transceivers, providing a unit for communicating with various other devices over a transmission medium. The processor 500 is responsible for managing the bus architecture and general processing, and the memory 520 may store data used by the processor 500 during operation.

[0307] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0308] Map the geocentric coordinates of the dot matrix data onto a nine-square grid in a plane.

[0309] For the first grid in the planar nine-grid that only maps a portion of the point cloud in the point matrix data, the first grid is divided into nine smaller grids.

[0310] Based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids, the contour information of the target POI is determined;

[0311] The coordinates of the center point of the planar nine-square grid are determined as the coordinates of the center point of the target POI.

[0312] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0313] Obtain a panoramic image of the target POI;

[0314] Text recognition is performed on the panoramic image to obtain at least one of the target POI's name and telephone information.

[0315] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0316] Text recognition is performed on the panoramic image to obtain the text information in the panoramic image;

[0317] The text information is filtered.

[0318] Based on the structural characteristics of telephone data, identify telephone information in filtered text information;

[0319] Based on the POI name structure characteristics, identify the name information in the remaining text information, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0320] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0321] Identify the clustering of the remaining text information in the panoramic image;

[0322] If the remaining text information appears as a cluster in the panoramic image, then the remaining text information is determined to be the target text information.

[0323] When the remaining text information is presented as multiple clusters in the panoramic image, the target cluster among the multiple clusters is determined, and the text information at the target cluster is determined as the target text information;

[0324] Based on the POI name structure features, identify the name information in the target text information.

[0325] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0326] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0327] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0328] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0329] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0330] The identified name elements are combined in a preset order to obtain name information.

[0331] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0332] The address information of the target POI is determined based on the center point coordinates of the target POI.

[0333] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0334] Based on the center point coordinates of the target POI and the pre-constructed administrative region data, the administrative location information of the target POI is determined.

[0335] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0336] Determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0337] Determine the road point on each of the N roads that is closest to the coordinates of the center point, thus obtaining N road points;

[0338] The target road containing the road point that is closest to the coordinates of the center point among the N road points is determined as the road information where the target POI is located.

[0339] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0340] Obtain information on M house numbers containing the target road, where M is an integer greater than or equal to 1;

[0341] The doorplate information that is closest to the center point coordinates of the target POI among the M doorplate information is determined as the doorplate information of the target POI.

[0342] Optionally, the processor 500 is also used to read the program from the memory 520 and perform the following steps:

[0343] Determine a first distance between a first road point and the center point coordinates of the target POI, and a second distance between a second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0344] If the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance, then the target address information is determined to be the address information of the target POI.

[0345] The data processing device provided in this embodiment of the invention can execute the above-described method embodiments, and its implementation principle and technical effect are similar, so it will not be described again here.

[0346] Furthermore, the computer-readable storage medium of this embodiment of the invention is used to store a computer program, which can be executed by a processor to implement the following steps:

[0347] Obtain the raster data of the target point of interest (POI);

[0348] Convert the coordinates of each point cloud in the matrix data to geocentric coordinates.

[0349] Determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data;

[0350] Based on the positional distribution of the dot matrix data in the planar region, the contour information of the target POI is determined, and the center point coordinates of the planar region are determined as the center point coordinates of the target POI.

[0351] Optionally, determining the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data includes:

[0352] Map the geocentric coordinates of the dot matrix data onto a nine-square grid in a plane.

[0353] For the first grid in the planar nine-grid that only maps a portion of the point cloud in the point matrix data, the first grid is divided into nine smaller grids.

[0354] The step of determining the contour information of the target POI based on the distribution of the dot matrix data in the planar region, and determining the center point coordinates of the planar region as the center point coordinates of the target POI, includes:

[0355] Based on the distribution of the dot matrix data in the planar nine-square grid and the nine zigzag grids, the contour information of the target POI is determined;

[0356] The coordinates of the center point of the planar nine-square grid are determined as the coordinates of the center point of the target POI.

[0357] Optionally, the method further includes:

[0358] Obtain a panoramic image of the target POI;

[0359] Text recognition is performed on the panoramic image to obtain at least one of the target POI's name and telephone information.

[0360] Optionally, the text recognition of the panoramic image includes:

[0361] Text recognition is performed on the panoramic image to obtain the text information in the panoramic image;

[0362] The text information is filtered.

[0363] Based on the structural characteristics of telephone data, identify telephone information in filtered text information;

[0364] Based on the POI name structure characteristics, identify the name information in the remaining text information, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

[0365] Optionally, identifying the name information in the remaining text information according to the POI name structure features includes:

[0366] Identify the clustering of the remaining text information in the panoramic image;

[0367] If the remaining text information appears as a cluster in the panoramic image, then the remaining text information is determined to be the target text information.

[0368] When the remaining text information is presented as multiple clusters in the panoramic image, the target cluster among the multiple clusters is determined, and the text information at the target cluster is determined as the target text information;

[0369] Based on the POI name structure features, identify the name information in the target text information.

[0370] Optionally, when the remaining text information is presented as multiple clusters in the panoramic image, determining the target cluster among the multiple clusters includes:

[0371] If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster;

[0372] If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

[0373] Optionally, identifying the name information in the target text information according to the POI name structure features includes:

[0374] Based on a pre-established POI name element lexicon, identify the name elements in the target text information;

[0375] The identified name elements are combined in a preset order to obtain name information.

[0376] Optionally, after determining the center point coordinates of the planar region as the center point coordinates of the target POI, the method further includes:

[0377] The address information of the target POI is determined based on the center point coordinates of the target POI.

[0378] Optionally, determining the address information of the target POI based on the center point coordinates of the target POI includes:

[0379] Based on the center point coordinates of the target POI and the pre-constructed administrative region data, the administrative location information of the target POI is determined.

[0380] Optionally, after determining the administrative location information of the target POI, the method further includes:

[0381] Determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1;

[0382] Determine the road point on each of the N roads that is closest to the coordinates of the center point, thus obtaining N road points;

[0383] The target road containing the road point that is closest to the coordinates of the center point among the N road points is determined as the road information where the target POI is located.

[0384] Optionally, after determining the target road containing the road point among the N road points that is closest to the coordinates of the center point as the road information where the target POI is located, the method further includes:

[0385] Obtain information on M house numbers containing the target road, where M is an integer greater than or equal to 1;

[0386] The doorplate information that is closest to the center point coordinates of the target POI among the M doorplate information is determined as the doorplate information of the target POI.

[0387] Optionally, determining the target address information among the M address information that is closest to the center point coordinates of the target POI as the address information of the target POI includes:

[0388] Determine a first distance between a first road point and the center point coordinates of the target POI, and a second distance between a second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance;

[0389] If the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance, then the target address information is determined to be the address information of the target POI.

[0390] In the several embodiments provided in this application, it should be understood that the disclosed methods and apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0391] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0392] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute some steps of the transmission and reception methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0393] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A data processing method, characterized in that, include: Obtain the dot matrix data of the target POI; Convert the coordinates of each point cloud in the matrix data to geocentric coordinates. Determine the planar region mapped by the geocentric coordinate system coordinates of the point data; Based on the positional distribution of the dot matrix data in the planar region, the contour information of the target POI is determined, and the center point coordinates of the planar region are determined as the center point coordinates of the target POI. Determining the planar region mapped by the geocentric coordinate system coordinates of the matrix data includes: Map the geocentric coordinates of the dot matrix data onto a nine-square grid in a plane. For the first grid in the planar nine-grid that only maps a portion of the point cloud in the point matrix data, the first grid is further divided into nine smaller grids until each grid or each smaller grid contains point cloud data, or contains no point cloud data at all. The step of determining the contour information of the target POI based on the distribution of the dot matrix data in the planar region, and determining the center point coordinates of the planar region as the center point coordinates of the target POI, includes: The outline of the target POI is determined by the edge grid or edge-gender grid that contains point clouds in the planar nine-square grid and the gender grid. The coordinates of the center point of the planar nine-square grid are determined as the coordinates of the center point of the target POI.

2. The method according to claim 1, characterized in that, The method further includes: Obtain a panoramic image of the target POI; Text recognition is performed on the panoramic image to obtain at least one of the target POI's name and telephone information.

3. The method according to claim 2, characterized in that, The text recognition of the panoramic image includes: Text recognition is performed on the panoramic image to obtain the text information in the panoramic image; The text information is filtered. Based on the structural characteristics of telephone data, identify telephone information in filtered text information; Based on the POI name structure characteristics, identify the name information in the remaining text information, wherein the remaining text information is the information in the filtered text information excluding the telephone information.

4. The method according to claim 3, characterized in that, The step of identifying name information in the remaining text information according to the POI name structure features includes: Identify the clustering of the remaining text information in the panoramic image; If the remaining text information appears as a cluster in the panoramic image, then the remaining text information is determined to be the target text information. When the remaining text information is presented as multiple clusters in the panoramic image, the target cluster among the multiple clusters is determined, and the text information at the target cluster is determined as the target text information; Based on the POI name structure features, identify the name information in the target text information.

5. The method according to claim 4, characterized in that, When the remaining text information appears as multiple clusters in the panoramic image, determining the target cluster among the multiple clusters includes: If the remaining text information is presented as multiple clusters in the panoramic image, and the size of each cluster is different, the largest cluster among the multiple clusters is determined as the target cluster; If the remaining text information is presented as multiple clusters in the panoramic image, and each cluster is of the same size, the central cluster among the multiple clusters is determined as the target cluster.

6. The method according to claim 4, characterized in that, The step of identifying the name information in the target text information according to the POI name structure features includes: Based on a pre-established POI name element lexicon, identify the name elements in the target text information; The identified name elements are combined in a preset order to obtain name information.

7. The method according to claim 1, characterized in that, After determining the center point coordinates of the planar region as the center point coordinates of the target POI, the method further includes: The address information of the target POI is determined based on the center point coordinates of the target POI.

8. The method according to claim 7, characterized in that, Determining the address information of the target POI based on its center point coordinates includes: Based on the center point coordinates of the target POI and the pre-constructed administrative region data, the administrative location information of the target POI is determined.

9. The method according to claim 8, characterized in that, After determining the administrative location information of the target POI, the method further includes: Determine N roads within a preset range of the center point coordinates of the target POI and the road point information on the N roads, where N is an integer greater than or equal to 1; Determine the road point on each of the N roads that is closest to the coordinates of the center point, thus obtaining N road points; The target road containing the road point that is closest to the coordinates of the center point among the N road points is determined as the road information where the target POI is located.

10. The method according to claim 9, characterized in that, After determining the target road containing the road point among the N road points that is closest to the coordinates of the center point as the road information of the target POI, the method further includes: Obtain information on M house numbers containing the target road, where M is an integer greater than or equal to 1; The doorplate information that is closest to the center point coordinates of the target POI among the M doorplate information is determined as the doorplate information of the target POI.

11. The method according to claim 10, characterized in that, The step of determining the target address information of the target POI as the address information of the target POI from the M address information information that is closest to the center point coordinates of the target POI includes: Determine a first distance between a first road point and the center point coordinates of the target POI, and a second distance between a second road point and the center point coordinates, wherein the first road point is the road point among the N road points that is closest to the center point coordinates, and the second road point is the road point among the N road points whose distance to the center point coordinates is second only to the first distance; If the distance between the target address information closest to the center point coordinates of the target POI among the M address information is less than or equal to the second distance, then the target address information is determined to be the address information of the target POI.

12. A data processing apparatus, characterized in that, include: The first acquisition module is used to acquire the dot matrix data of the target POI; The conversion module is used to convert the coordinates of each point cloud in the point cloud data into geocentric coordinates. The first determining module is used to determine the planar region mapped by the geocentric coordinate system coordinates of the dot matrix data; The second determining module is used to determine the contour information of the target POI based on the positional distribution of the dot matrix data in the planar area, and to determine the center point coordinates of the planar area as the center point coordinates of the target POI. The first determining module includes: The mapping unit is used to map the geocentric coordinates of the dot matrix data onto a planar nine-square grid. The division unit is used to further divide the first grid in the planar nine-grid into nine smaller grids, until each grid or the smaller grid contains all point cloud data, or none of them contain any point cloud data. The second determining module includes: The first determining unit is used to determine the outline of the target POI as the outline of the edge grid or edge-gender grid composed of point clouds in the planar nine-square grid and the gender grid. The second determining unit is used to determine the center point coordinates of the planar nine-square grid as the center point coordinates of the target POI.

13. A data processing device, comprising: A transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor; characterized in that the processor is configured to read the program in the memory to implement the steps of the data processing method as described in any one of claims 1 to 11.

14. A computer-readable storage medium for storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the data processing method as described in any one of claims 1 to 11.

Images