A target recognition method and system based on dynamic clustering of interest points

Abstract

This application discloses a target recognition method and system based on dynamic clustering of points of interest. The method includes: in response to a target recognition request, obtaining the physical value and pixel value of the screen width of the target display device to be identified, calculating the screen resolution of the target display device, and calculating dynamic clustering parameters based on the screen resolution; obtaining a data set of points of interest to be clustered, scaling the points of interest data in the data set based on the dynamic clustering parameters to obtain cluster space coordinates; obtaining an array of points of interest constructed from the data of each point of interest corresponding to the cluster space coordinates, constructing a hash table based on the cluster space coordinates and the array of points of interest corresponding to the cluster space coordinates; performing multi-dimensional sorting on each array of points of interest, generating cluster centers of the data set of points of interest based on the results of the multi-dimensional sorting, and performing target recognition based on each generated cluster center. The method of this application improves the accuracy of target recognition.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a target recognition method and system based on dynamic clustering of points of interest. Background Art

[0002] With the rapid development of geographic information systems (GIS) and location-based services (LBS), point-of-interest (POI) data is finding increasingly widespread application in scenarios such as map navigation, virtual reality (VR), and augmented reality (AR). However, faced with massive amounts of POI data, traditional display methods can lead to information overload, preventing users from quickly accessing useful information and thus affecting immersive user experience and interactive effects.

[0003] In existing technologies, clustering algorithms are often used to reduce the amount of visual information. However, these algorithms rely on fixed grid sizes or fixed clustering parameters, and are unable to adjust clustering density based on device screen resolution or display area changes. This results in significant discrepancies in clustering results for the same physical area on different devices, leading to reduced object recognition accuracy.

[0004] It can be seen that how to improve the existing clustering method to improve the accuracy of target recognition has become a technical problem that needs to be solved urgently by those skilled in the art. Summary of the Invention

[0005] The present application provides a target recognition method and system based on dynamic clustering of points of interest to solve the technical problem of how to improve the existing clustering method to improve the accuracy of target recognition.

[0006] In order to solve the above technical problems, the embodiment of the present application provides a target recognition method based on dynamic clustering of interest points, including:

[0007] In response to the target recognition request, obtaining a data set of points of interest to be clustered, and obtaining a physical value of the screen width and a pixel value of the screen width of the target display device to be recognized;

[0008] Calculating a screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculating a dynamic clustering parameter according to the screen resolution;

[0009] Scaling each point of interest data in the point of interest data set based on the dynamic clustering parameter to obtain cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information and unique ID information;

[0010] Acquire an interest point array constructed by each of the interest point data corresponding to the cluster space coordinates, and construct a hash table according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates;

[0011] Perform multi-dimensional sorting on each of the interest point arrays, generate cluster centers of the interest point data set based on the results of the multi-dimensional sorting, and perform target recognition based on the generated cluster centers, wherein the multi-dimensional sorting is designed to perform sorting based on the level information, the coordinate information and the unique ID information of each of the interest point data in sequence.

[0012] As one preferred solution, the step of obtaining a dataset of points of interest to be clustered includes:

[0013] Acquire an initial point of interest dataset from multiple data sources, wherein the data sources include but are not limited to a map database, a user-defined point of interest list, and a data interface of a third-party geographic information service provider;

[0014] The initial interest point data set is cleaned and formatted to obtain an interest point data set to be clustered.

[0015] As one preferred solution, the calculating of the dynamic clustering parameters according to the screen resolution includes:

[0016]

[0017]

[0018] in, represents the dynamic clustering parameters, represents the power operation, Indicates rounding up. Indicates the screen resolution, Indicates the physical value of the screen width. Indicates the screen width in pixels.

[0019] As one preferred solution, the calculation frequency of the dynamic clustering parameters is adjusted based on the usage scenario of the target display device, specifically including:

[0020] When the screen resolution or display area of ​​the target display device changes, the dynamic clustering parameters are recalculated.

[0021] As one preferred solution, scaling each point of interest data in the point of interest data set based on the dynamic clustering parameter to obtain cluster space coordinates corresponding to the point of interest data includes:

[0022] Extract the coordinate information of each point of interest data, wherein the coordinate information includes abscissa and ordinate, expressed as ;

[0023] The horizontal coordinate and the vertical coordinate are scaled based on the dynamic clustering parameters to obtain the cluster space coordinates, which are expressed as:

[0024]

[0025]

[0026] in, represents the abscissa of the cluster space coordinates, The vertical coordinate of the cluster space coordinate is shown. Represents dynamic clustering parameters.

[0027] As one preferred solution, after obtaining the cluster space coordinates corresponding to the interest point data, the method further includes:

[0028] Performing a validity check on the cluster space coordinates, specifically including correcting the dynamic clustering parameters if the cluster space coordinates exceed a preset threshold range.

[0029] As one preferred solution, constructing a hash table according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates includes:

[0030] Constructing a hash table using the cluster space coordinates as keys and an array of interest points corresponding to the cluster space coordinates as values;

[0031] In which, during the construction process of the hash table, the load factor of the hash table is monitored in real time. If the load factor exceeds a preset load threshold, the hash table is expanded; the load factor is the ratio of the number of elements stored in the hash table to the capacity of the hash table.

[0032] As one preferred solution, performing multi-dimensional sorting on each of the interest point arrays and generating the cluster center of the interest point data set according to the result of the multi-dimensional sorting includes:

[0033] Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data;

[0034] Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, and update the first sorting result based on the distance information to obtain a second sorting result; wherein the cluster center is represented by , represents the cluster space coordinates, represents the dynamic clustering parameters;

[0035] Obtaining each second point of interest data having the same ranking in the second sorting result, and sorting the data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence;

[0036] The cluster center of the interest point data set is obtained according to the third interest point sequence.

[0037] Another embodiment of the present application provides an object recognition system based on dynamic clustering of interest points, including:

[0038] A first acquisition module is configured to acquire, in response to a target recognition request, a data set of points of interest to be clustered, and acquire a physical value of a screen width and a pixel value of a screen width of a target display device to be recognized;

[0039] a calculation module, configured to calculate a screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculate a dynamic clustering parameter according to the screen resolution;

[0040] a second acquisition module, configured to scale each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information, and unique ID information;

[0041] A construction module, configured to obtain an interest point array constructed by each of the interest point data corresponding to the cluster space coordinates, and construct a hash table according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates;

[0042] A sorting module is used to perform multi-dimensional sorting on each of the interest point arrays, generate cluster centers of the interest point data set based on the results of the multi-dimensional sorting, and perform target recognition based on each of the generated cluster centers, wherein the multi-dimensional sorting is designed to be sequentially sorted based on the level information, coordinate information and unique ID information of each interest point data.

[0043] As one preferred solution, the sorting module is further configured to:

[0044] Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data;

[0045] Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, and update the first sorting result based on the distance information to obtain a second sorting result; wherein the cluster center is represented by , represents the cluster space coordinates, represents the dynamic clustering parameters;

[0046] Obtaining each second point of interest data having the same ranking in the second sorting result, and sorting the data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence;

[0047] The cluster center of the interest point data set is obtained according to the third interest point sequence.

[0048] Compared with the prior art, the embodiments of the present application have the following advantages:

[0049] (1) This application dynamically adjusts clustering parameters by calculating the physical size and pixel width of the device screen in real time, thus achieving adaptive clustering density. This eliminates the need to develop separate adaptation logic for VR, AR, mobile phones, and other devices. One set of algorithms covers all scenarios, significantly reducing adaptation costs.

[0050] (2) This application sorts the clustered POIs according to the following rules: priority, Manhattan distance, and unique ID. In this way, in emergency scenarios (such as fires), high-priority targets (such as fire hydrants and hospitals) will be identified first, shortening the response time.

[0051] (3) This application scales the POI data based on dynamic clustering parameters to obtain cluster spatial coordinates and construct a hash table. This spatial hash storage method enables rapid aggregation and retrieval of POIs with nearly constant time complexity, significantly improving recognition efficiency and meeting the strict requirements of application scenarios such as real-time navigation and emergency response for rapid target recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flowchart of a target recognition method based on dynamic clustering of interest points in one embodiment of the present application;

[0053] Figure 2 Schematic diagram of a target recognition system based on dynamic clustering of interest points in one embodiment of the present application. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. The purpose of providing these embodiments is to make the disclosure of this application more thorough and comprehensive. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0055] In the description of this application, the terms "first," "second," "third," etc. are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first," "second," "third," etc. may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0056] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the two elements. The terms "vertical", "horizontal", "left", "right", "up", "down" and similar expressions used herein are for illustrative purposes only, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0057] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meanings as those commonly understood by those skilled in the art. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by those skilled in the art in specific circumstances.

[0058] An embodiment of the present application provides a target recognition method based on dynamic clustering of interest points. For details, see Figure 1 , Figure 1 The figure shows a flow chart of a target recognition method based on dynamic clustering of interest points in one embodiment of the present application, which includes steps S1-S5:

[0059] S1: In response to a target recognition request, obtain a data set of points of interest to be clustered, and obtain a physical value of the screen width and a pixel value of the screen width of a target display device to be recognized;

[0060] When a target recognition request is received, an event listening mechanism is first triggered. This mechanism can be implemented through a software-level callback function or message queue. For example, in a mobile application-based target recognition scenario, when a user clicks a specific button on the map interface to initiate a target recognition operation, the event handler within the application will capture this request and pass it to the module responsible for obtaining device screen information. In more complex distributed systems, the request may be sent from the client to the server over the network. After receiving the request, the request processing service on the server begins the process of obtaining device screen information.

[0061] Preferably, in one embodiment of the present application, obtaining a data set of points of interest to be clustered includes:

[0062] Obtaining an initial POI dataset from multiple data sources, where the data sources include but are not limited to a map database, a user-defined POI list, and a data interface of a third-party geographic information service provider;

[0063] The initial interest point dataset is cleaned and formatted uniformly to obtain the interest point dataset to be clustered.

[0064] In order to obtain comprehensive and accurate POI data, it is necessary to collect information from multiple different data sources, each of which has its own characteristics and advantages.

[0065] Map databases are one of the important sources of point of interest data, which usually contain a large amount of geographic information, such as commercial places, public facilities, transportation stations, etc. These data are professionally collected and organized, with high accuracy and completeness.

[0066] During use, users may define specific points of interest based on their needs and preferences. These points of interest have special meaning to the user and may be places they frequently visit or places of interest. The method of this application allows users to upload or input a list of their own points of interest and incorporate it into the initial points of interest dataset to meet personalized needs.

[0067] Third-party geographic information service providers possess specialized geographic data collection and processing capabilities, enabling them to provide diverse POI data. For example, some companies specializing in geographic information services collect and organize POI data through satellite remote sensing, field surveys, and other methods, and then make it available to other systems through data interfaces. By invoking these interfaces, POI data of varying types and sources can be obtained, enriching the initial dataset.

[0068] Initial POI datasets obtained from multiple data sources may have uneven data quality and inconsistent formats, necessitating data cleaning and formatting. The goal of data cleaning is to remove noise, duplicates, and errors, improving data quality. Specific steps include checking data integrity and removing POIs that are missing key information (such as coordinates and names); checking data accuracy and correcting incorrect coordinates, names, and other information; and removing duplicate POIs to avoid data redundancy. For example, by comparing the coordinates and names of POIs, POIs with identical coordinates and similar names can be treated as duplicates and removed.

[0069] POI data provided by different data sources may have different formats and structures. To facilitate subsequent processing and analysis, the data needs to be standardized. This includes standardizing the field names, data types, and encoding methods of the POI data. For example, coordinate information from different data sources can be uniformly converted to a pseudo-Mercator projection format, and POI names can be uniformly encoded using UTF-8.

[0070] S2: Calculate the screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculate the dynamic clustering parameters according to the screen resolution;

[0071] Preferably, in one embodiment of the present application, the dynamic clustering parameter is calculated according to the screen resolution, including:

[0072]

[0073]

[0074] in, represents the dynamic clustering parameters, represents the power operation, Indicates rounding up. Indicates the screen resolution, Indicates the physical value of the screen width. Indicates the screen width in pixels.

[0075] The screen resolution calculation formula correlates the screen's physical dimensions with the number of pixels, yielding the physical dimensions per pixel, or screen resolution. This value reflects the screen's level of detail and is crucial for subsequent point-of-interest clustering and object recognition. For example, on high-resolution devices, each pixel represents a smaller physical area, requiring more refined clustering to accurately represent points of interest. On low-resolution devices, however, pixels represent larger physical areas, allowing for more coarse clustering.

[0076] When performing calculations, ensure the accuracy and consistency of the obtained screen width physical value and screen width pixel value. If the obtained value is abnormal (such as a negative screen width physical value or a screen width pixel value of 0), error handling is required. Error information can be logged and the user can be notified of the device screen information acquisition anomaly, asking the user to check the device settings or try again. To improve calculation accuracy, use high-precision data types (such as floating-point numbers) when calculating screen resolution to avoid precision loss due to data type conversion.

[0077] After determining the screen resolution, the dynamic clustering parameter is further calculated. The dynamic clustering parameter calculation formula combines the screen resolution and screen width pixel values ​​to produce a dynamic clustering parameter that reflects the device's screen characteristics. This calculation method allows the clustering granularity to be dynamically adjusted based on the screen characteristics of different devices. For example, for high-resolution devices, the calculated dynamic clustering parameter value will be relatively small, resulting in finer clustering; whereas for low-resolution devices, the dynamic clustering parameter value will be larger, resulting in coarser clustering.

[0078] Preferably, in one embodiment of the present application, the calculation frequency of the dynamic clustering parameters is adjusted based on the usage scenario of the target display device, specifically including:

[0079] When the screen resolution or display area of ​​the target display device changes, the dynamic clustering parameters are recalculated.

[0080] To better adapt the dynamic clustering parameters to changes in the device screen, the calculation frequency needs to be adjusted based on the target display device's usage scenario. When the target display device's screen resolution or display area changes, the dynamic clustering parameters are immediately recalculated. This can be achieved by monitoring the device's screen change events. For example, on mobile devices, you can register a broadcast receiver for screen resolution changes; on desktop systems, you can receive screen change events through the screen change notification mechanism provided by the operating system. When a screen change event is detected, the system re-acquires the physical and pixel values ​​of the screen width and recalculates the screen resolution and dynamic clustering parameters according to the above steps to ensure that the clustering parameters promptly reflect the latest screen status.

[0081] S3: scaling each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain the cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information and unique ID information;

[0082] Preferably, in one embodiment of the present application, each point of interest data in the point of interest data set is scaled based on the dynamic clustering parameter to obtain the cluster space coordinates corresponding to the point of interest data, including:

[0083] Extract the coordinate information of each point of interest data, where the coordinate information includes the horizontal coordinate and the vertical coordinate, expressed as ;

[0084] Based on the dynamic clustering parameters, the horizontal and vertical coordinates are scaled to obtain the cluster space coordinates, which are expressed as:

[0085]

[0086]

[0087] in, represents the abscissa of the cluster space coordinates, The vertical coordinate of the cluster space coordinate is shown. Represents dynamic clustering parameters.

[0088] After obtaining the data set of interest points to be clustered, it is necessary to scale each interest point data based on the dynamic clustering parameters calculated previously to obtain the corresponding cluster space coordinates.

[0089] Each POI data contains coordinate information, rank information, and unique ID information. Coordinate information is usually expressed in the form of a pseudo-Mercator projection coordinate system, which is used to determine the location of the POI in geographic space. The rank information indicates the importance or priority of the POI. For example, hospitals and transportation hubs may have a higher rank. The unique ID information is used to uniquely identify each POI, facilitating subsequent query and management.

[0090] The purpose of scaling is to adjust the position of POIs in the cluster space based on dynamic clustering parameters, so that the clustering better matches the display characteristics of the device screen. Scaling is achieved by multiplying the POI coordinates by the dynamic clustering parameters. This method maps POIs from their original geographic space to the cluster space, ensuring more accurate clustering of POIs across different screen resolutions and display areas.

[0091] The resulting cluster space coordinates are used in subsequent clustering operations, allowing the system to group POIs into different clusters based on these coordinates. During the clustering process, cluster space coordinates can more accurately reflect the relative positional relationships between POIs, thereby improving clustering accuracy and efficiency. For example, in density-based clustering algorithms, cluster space coordinates can be used to calculate the distance between POIs and determine whether they belong to the same cluster.

[0092] In practical applications, due to factors such as the diversity of device screen characteristics, errors in data calculation, and the peculiarities of point-of-interest distribution, some cluster space coordinates may fall outside the reasonable range. Using these abnormal cluster space coordinates for subsequent clustering and object recognition operations can lead to confusion in clustering results and inaccurate object recognition, severely impacting overall system performance and user experience. Therefore, validating cluster space coordinates is essential.

[0093] Preferably, in one embodiment of the present application, after obtaining the cluster space coordinates corresponding to the interest point data, the method further includes:

[0094] The cluster space coordinates are checked for validity, including modifying the dynamic clustering parameters if the cluster space coordinates exceed a preset threshold range.

[0095] After obtaining the cluster space coordinates corresponding to the point of interest data, the system will perform a legitimacy check according to the preset threshold range.

[0096] First, check whether the cluster space horizontal coordinate is within the preset width threshold range. If it exceeds this range, it indicates that the cluster space coordinate of the interest point in the width direction may be abnormal. Similarly, check the cluster space vertical coordinate in the height direction. If it exceeds the range, it also indicates that the cluster space coordinate of the interest point in the height direction is abnormal.

[0097] S4: Obtain an interest point array constructed by the data of each interest point corresponding to the cluster space coordinates, and construct a hash table according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates;

[0098] Preferably, in one embodiment of the present application, a hash table is constructed according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates, including:

[0099] A hash table is constructed with cluster spatial coordinates as keys and the array of interest points corresponding to the cluster spatial coordinates as values;

[0100] During the construction of the hash table, the load factor of the hash table is monitored in real time. If the load factor exceeds a preset load threshold, the hash table is expanded. The load factor is the ratio of the number of elements stored in the hash table to the capacity of the hash table.

[0101] In object recognition methods based on dynamic POI clustering, a hash table is constructed to efficiently store and retrieve POI data. POI data with the same cluster spatial coordinates are first grouped into an array. For example, after dynamic parameter scaling, multiple POIs may be mapped to the same cluster coordinate (cx, cy). The coordinates, levels, IDs, and other information of these POIs are organized into an array to form a candidate dataset for that cluster unit. When creating the hash table, an initial capacity (e.g., 16 buckets) is preset, with each bucket storing key-value pairs. The key is the cluster spatial coordinate (cx, cy), and the value is the corresponding POI array. Initially, the hash table is empty, waiting to be populated with data.

[0102] Traverse all cluster coordinates and their corresponding interest point arrays, mapping each key-value pair to a bucket using a hash function. For example, use a hash algorithm with (cx*31+cy)% capacity to convert the coordinates into bucket indices to ensure that different coordinates are evenly distributed.

[0103] Load factor = number of stored elements / total capacity of the hash table. For example, a hash table with a capacity of 100 storing 75 elements has a load factor of 0.75. This metric reflects the storage density of the hash table. The system automatically calculates the current load factor each time a new key-value pair is inserted. When the load factor exceeds a preset threshold (typically 0.75), a capacity expansion is triggered to prevent increased hash conflicts and performance degradation. Capacity expansion doubles the hash table capacity (for example, from 16 to 32) and creates a new bucket array. The hash function recalculates the hash values ​​for all existing keys at the new capacity to ensure even data distribution.

[0104] In one embodiment of the present application, the steps of data expansion include: traversing each storage bucket of the original hash table; for each key-value pair, recalculating the hash value using the new capacity and determining the new storage bucket location; inserting the key-value pair into the new storage bucket and handling possible hash conflicts (through a linked list or linear probing); updating the hash table capacity and element count to complete the migration.

[0105] S5: Perform multi-dimensional sorting on each POI array, generate cluster centers of the POI dataset based on the results of the multi-dimensional sorting, and perform target recognition based on the generated cluster centers, wherein the multi-dimensional sorting is designed to sort based on the level information, coordinate information, and unique ID information of each POI data in turn.

[0106] Preferably, in one embodiment of the present application, performing multi-dimensional sorting on each interest point array and generating the cluster center of the interest point data set according to the result of the multi-dimensional sorting includes:

[0107] Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data;

[0108] Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, update the first sorting result based on the distance information, and obtain the second sorting result; wherein the cluster center is represented as , represents the cluster space coordinates, represents the dynamic clustering parameters;

[0109] Obtaining the second points of interest data of the same order in the second sorting result, and sorting the second points of interest data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence;

[0110] The cluster center of the interest point data set is obtained according to the third interest point sequence.

[0111] The level information in the POI data reflects the importance or priority of the POI. For example, in the POI data of a city map, some important public facilities may have a higher level, while ordinary stores may have a lower level.

[0112] In order to sort the POI array according to its rank information, this application uses a quick sort algorithm. Quick sort is an efficient sorting algorithm with an average time complexity of O(nlogn). Its basic idea is to select a base element, divide the array into two parts, so that all elements in the left part are less than or equal to the base element, and all elements in the right part are greater than the base element, and then recursively sort the left and right parts. After quick sorting, this application obtains a first sorting result arranged according to the rank information from high to low (or from low to high, depending on the specific needs). In this way, in subsequent processing, POIs with higher ranks will be given priority.

[0113] In the first sorting result, there may be multiple points of interest with the same level information. For these points of interest with the same level (i.e., the first point of interest data), further processing is required. First, the representation of the cluster center needs to be clarified. Here, the cluster center is represented as Then, the distance between each first POI and the cluster center is calculated using the common Euclidean distance formula. Based on the calculated distance information, the first sorting results are reordered so that POIs with the same rank are placed first, resulting in a second sorting result. This is done to prioritize POIs closer to the cluster center when the rank is the same, as they are likely to be more closely associated with the cluster center.

[0114] In the second sorting result, there may be POIs (i.e., second POI data) with the same distance from the cluster center and the same rank. To further distinguish these POIs, we use unique ID information for sorting. The unique ID uniquely identifies each POI, and sorting it ensures that the final sorting result is unique. Sorting these second POI data in ascending order of unique ID (or vice versa) yields the third POI sequence.

[0115] After obtaining the third interest point sequence, the cluster centers of the interest point data set can be generated according to the sequence.

[0116] In one embodiment of the present application, the first few points of interest in the sorted sequence of points of interest are selected (the specific number can be determined according to actual needs), and then the average value of the coordinates of these points of interest is calculated as the coordinates of the cluster center.

[0117] After generating the cluster centers, we can then perform target recognition based on these cluster centers. There are many specific methods for target recognition, such as region division and feature matching.

[0118] Through the above multi-dimensional sorting and cluster center generation process, points of interest can be clustered and targets can be identified more accurately, thereby improving the accuracy and efficiency of recognition.

[0119] Another embodiment of the present application provides a target recognition system based on dynamic clustering of interest points. For details, see Figure 2 , Figure 2 FIG. 1 is a schematic diagram of an object recognition system based on dynamic clustering of interest points in one embodiment of the present application, which includes:

[0120] A first acquisition module 11 is configured to acquire, in response to a target recognition request, a dataset of points of interest to be clustered, and acquire a physical value of a screen width and a pixel value of a screen width of a target display device to be recognized;

[0121] A calculation module 12 is used to calculate the screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculate the dynamic clustering parameter according to the screen resolution;

[0122] A second acquisition module 13 is configured to scale each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information, and unique ID information;

[0123] A construction module 14 is configured to obtain an interest point array constructed by data of each interest point corresponding to the cluster space coordinates, and to construct a hash according to the cluster space coordinates and the interest point array corresponding to the cluster space coordinates;

[0124] The sorting module 15 is used to perform multi-dimensional sorting on each POI array, generate cluster centers of the POI data set based on the results of the multi-dimensional sorting, and perform target recognition based on the generated cluster centers, wherein the multi-dimensional sorting is designed to sort based on the level information, coordinate information and unique ID information of each POI data in turn.

[0125] Preferably, in one embodiment of the present application, the sorting module is further configured to:

[0126] Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data;

[0127] Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, and update the first sorting result based on the distance information to obtain a second sorting result; wherein the cluster center is represented by , represents the cluster space coordinates, represents the dynamic clustering parameters;

[0128] Obtaining each second point of interest data having the same ranking in the second sorting result, and sorting the data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence;

[0129] The cluster center of the interest point data set is obtained according to the third interest point sequence.

[0130] Compared with the prior art, the embodiments of the present application have the following advantages:

[0131] (1) This application dynamically adjusts clustering parameters by calculating the physical size and pixel width of the device screen in real time, thus achieving adaptive clustering density. This eliminates the need to develop separate adaptation logic for VR, AR, mobile phones, and other devices. One set of algorithms covers all scenarios, significantly reducing adaptation costs.

[0132] (2) This application sorts the clustered POIs according to the following rules: priority, Manhattan distance, and unique ID. In this way, in emergency scenarios (such as fires), high-priority targets (such as fire hydrants and hospitals) will be identified first, shortening the response time.

[0133] (3) This application scales the POI data based on dynamic clustering parameters to obtain cluster spatial coordinates and construct a hash table. This spatial hash storage method enables rapid aggregation and retrieval of POIs with nearly constant time complexity, significantly improving recognition efficiency and meeting the strict requirements of application scenarios such as real-time navigation and emergency response for rapid target recognition.

[0134] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A target recognition method based on dynamic clustering of interest points, characterized in that: include: In response to the target recognition request, obtaining a data set of points of interest to be clustered, and obtaining a physical value of the screen width and a pixel value of the screen width of the target display device to be recognized; Calculating the screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculating the dynamic clustering parameter according to the screen resolution; wherein calculating the dynamic clustering parameter according to the screen resolution includes: in, represents the dynamic clustering parameters, represents the power operation, Indicates rounding up. Indicates the screen resolution, Indicates the physical value of the screen width. Indicates the screen width in pixels; Scaling each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information, and unique ID information; wherein scaling each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data includes: Extract the coordinate information of each point of interest data, wherein the coordinate information includes abscissa and ordinate, expressed as ; The horizontal coordinate and the vertical coordinate are scaled based on the dynamic clustering parameters to obtain the cluster space coordinates, which are expressed as: in, represents the abscissa of the cluster space coordinates, The vertical coordinate of the cluster space coordinate is shown. represents the dynamic clustering parameters; Obtaining an array of points of interest constructed from the data of each point of interest corresponding to the cluster spatial coordinates, and constructing a hash table based on the cluster spatial coordinates and the array of points of interest corresponding to the cluster spatial coordinates; wherein constructing the hash table based on the cluster spatial coordinates and the array of points of interest corresponding to the cluster spatial coordinates includes: Constructing a hash table using the cluster space coordinates as keys and an array of interest points corresponding to the cluster space coordinates as values; During the construction of the hash table, the load factor of the hash table is monitored in real time. If the load factor exceeds a preset load threshold, the hash table is expanded. The load factor is the ratio of the number of elements stored in the hash table to the capacity of the hash table. Perform multi-dimensional sorting on each of the interest point arrays, generate cluster centers of the interest point data set based on the results of the multi-dimensional sorting, and perform target recognition based on the generated cluster centers, wherein the multi-dimensional sorting is designed to perform sorting based on the level information, the coordinate information and the unique ID information of each of the interest point data in sequence.

2. The target recognition method based on dynamic clustering of interest points according to claim 1, characterized in that: The step of obtaining a data set of points of interest to be clustered includes: Acquire an initial point of interest dataset from multiple data sources, wherein the data sources include but are not limited to a map database, a user-defined point of interest list, and a data interface of a third-party geographic information service provider; The initial interest point data set is cleaned and formatted to obtain an interest point data set to be clustered.

3. The target recognition method based on dynamic clustering of interest points according to claim 1, characterized in that: The calculation frequency of the dynamic clustering parameters is adjusted based on the usage scenario of the target display device, specifically including: When the screen resolution or display area of ​​the target display device changes, the dynamic clustering parameters are recalculated.

4. The target recognition method based on dynamic clustering of interest points according to claim 1, characterized in that: After obtaining the cluster space coordinates corresponding to the interest point data, the method further includes: Performing a validity check on the cluster space coordinates, specifically including correcting the dynamic clustering parameters if the cluster space coordinates exceed a preset threshold range.

5. The target recognition method based on dynamic clustering of interest points according to claim 1, characterized in that: The performing multi-dimensional sorting on each of the interest point arrays and generating the cluster center of the interest point data set according to the result of the multi-dimensional sorting includes: Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data; Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, and update the first sorting result based on the distance information to obtain a second sorting result; wherein the cluster center is represented by , represents the cluster space coordinates, represents the dynamic clustering parameters; Obtaining each second point of interest data having the same ranking in the second sorting result, and sorting the data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence; The cluster center of the interest point data set is obtained according to the third interest point sequence.

6. A target recognition system based on dynamic clustering of interest points, characterized in that: include: A first acquisition module is configured to acquire, in response to a target recognition request, a data set of points of interest to be clustered, and acquire a physical value of a screen width and a pixel value of a screen width of a target display device to be recognized; A calculation module is configured to calculate the screen resolution of the target display device based on the physical value of the screen width and the pixel value of the screen width, and calculate dynamic clustering parameters according to the screen resolution; wherein calculating the dynamic clustering parameters according to the screen resolution includes: in, represents the dynamic clustering parameters, represents the power operation, Indicates rounding up. Indicates the screen resolution, Indicates the physical value of the screen width. Indicates the screen width in pixels; A second acquisition module is configured to scale each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data, wherein each point of interest data includes coordinate information, level information, and unique ID information; wherein scaling each point of interest data in the point of interest data set based on the dynamic clustering parameters to obtain cluster space coordinates corresponding to the point of interest data includes: Extract the coordinate information of each point of interest data, wherein the coordinate information includes abscissa and ordinate, expressed as ; The horizontal coordinate and the vertical coordinate are scaled based on the dynamic clustering parameters to obtain the cluster space coordinates, which are expressed as: in, represents the abscissa of the cluster space coordinates, The vertical coordinate of the cluster space coordinate is shown. represents the dynamic clustering parameters; A construction module is configured to obtain an array of interest points constructed from the data of each interest point corresponding to the cluster spatial coordinates, and to construct a hash table based on the cluster spatial coordinates and the array of interest points corresponding to the cluster spatial coordinates; wherein the construction of the hash table based on the cluster spatial coordinates and the array of interest points corresponding to the cluster spatial coordinates includes: Constructing a hash table using the cluster space coordinates as keys and an array of interest points corresponding to the cluster space coordinates as values; During the construction of the hash table, the load factor of the hash table is monitored in real time. If the load factor exceeds a preset load threshold, the hash table is expanded. The load factor is the ratio of the number of elements stored in the hash table to the capacity of the hash table. A sorting module is used to perform multi-dimensional sorting on each of the interest point arrays, generate cluster centers of the interest point data set based on the results of the multi-dimensional sorting, and perform target recognition based on the generated cluster centers, wherein the multi-dimensional sorting is designed to sort based on the level information, coordinate information and unique ID information of each interest point data in sequence.

7. The target recognition system based on dynamic clustering of interest points according to claim 6, characterized in that: The sorting module is further used to: Sorting the level information of each point of interest data in the point of interest array to obtain a first sorting result; wherein the level information reflects the importance or priority of each point of interest data; Obtain the first points of interest data with the same ranking in the first sorting result, calculate the distance information between each first point of interest data and the cluster center, and update the first sorting result based on the distance information to obtain a second sorting result; wherein the cluster center is represented by , represents the cluster space coordinates, represents the dynamic clustering parameters; Obtaining each second point of interest data having the same ranking in the second sorting result, and sorting the data according to the unique ID information of each second point of interest data to obtain a third point of interest sequence; The cluster center of the interest point data set is obtained according to the third interest point sequence.

Images