Base Station Scene Recognition Method Based on Improved SVM Model

Abstract

This application discloses a base station scene recognition method based on an improved SVM model, relating to the field of artificial intelligence technology. The method includes: acquiring DEM data of a target site and extracting elevation features from the DEM data according to the base station coverage area of ​​the target site; generating a surrounding terrain elevation image of the target site based on the elevation features; and identifying the base station scene type corresponding to the target site based on the surrounding terrain elevation image. This application can improve the intelligence level and accuracy of base station scene recognition.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a base station scene recognition method, apparatus, device, computer-readable storage medium, and computer program product based on an improved SVM model. Background Technology

[0002] With the development of mobile communication networks and the large-scale deployment of 5G networks, base station site selection and planning, as an essential part of network construction, deserve attention.

[0003] Currently, base station scene identification is usually based on rule-based classification methods using geographic information system data, thereby enabling base station site selection and planning.

[0004] However, terrain data is usually quite complex, and the threshold judgment rules for different terrain categories in rule-based classification methods are usually quite simple, which leads to low accuracy in base station scene recognition. Summary of the Invention

[0005] The main purpose of this application is to provide a base station scene recognition method based on an improved SVM model, which aims to improve the accuracy of base station scene recognition.

[0006] To achieve the above objectives, this application proposes a base station scene recognition method based on an improved SVM model, the method comprising: Obtain DEM data (Digital Elevation Model, which can be used as digital representation of terrain surface morphology to characterize the undulation of terrain) for the target site. Based on the base station coverage of the target site, extract elevation features from the DEM data; Based on the elevation features, generate an elevation image of the surrounding terrain of the target station; Based on the surrounding terrain elevation image, the base station scene type corresponding to the target site is identified.

[0007] Furthermore, to achieve the above objectives, this application also proposes a base station scene recognition device based on an improved SVM model, the device comprising: The feature extraction module is used to acquire the DEM data of the target site and extract elevation features from the DEM data according to the base station coverage of the target site. An image generation module is used to generate an elevation image of the surrounding terrain of the target station based on the elevation features; The scene recognition module is used to identify the base station scene type corresponding to the target site based on the surrounding terrain elevation image.

[0008] Furthermore, to achieve the above objectives, this application also proposes a base station scene recognition device based on an improved SVM model. The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the base station scene recognition method based on the improved SVM model described above.

[0009] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the base station scene recognition method based on the SVM improved model described above.

[0010] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the base station scene recognition method based on the SVM improved model as described above.

[0011] The technical solution proposed in this application has at least the following technical effects: In this application, DEM data of the target site can be obtained, and then elevation features can be extracted from the DEM data based on the base station coverage area of ​​the target site. This allows for the extraction of standard-sized elevation features from complex DEM data. Since the elevation features are extracted from the DEM data based on the base station coverage area of ​​the target site, they also match the base station coverage area of ​​the target site. The elevation features accurately characterize the surrounding terrain conditions of the target site. Based on these elevation features, a surrounding terrain elevation image of the target site is generated, which can fully, accurately, and intuitively characterize the surrounding terrain conditions. Therefore, based on the surrounding terrain elevation image, the base station scene type corresponding to the target site can be identified, ensuring the accuracy of base station scene identification. Furthermore, since the surrounding terrain elevation image is flexibly generated based on DEM data and base station coverage area, the process of identifying the base station scene type corresponding to the target site in this application can adapt to different types of site terrain. Compared to base station scene discrimination based on a single threshold judgment rule, the base station scene identification in this application has a wider range of adaptability and is more accurate, thus further ensuring the accuracy of base station scene identification and solving the technical problem of low accuracy in base station scene identification. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A flowchart illustrating the base station scene recognition method based on the SVM improved model provided in this application; Figure 2 This is a schematic diagram of the matching and verification process of a base station scene recognition method based on an improved SVM model in one embodiment of this application. Figure 3 This is a schematic diagram of the engineering verification process of a base station scene recognition method based on an improved SVM model in one embodiment of this application; Figure 4 This is a schematic diagram of a base station scene recognition device module based on an improved SVM model in one embodiment of this application; Figure 5 This is a schematic diagram of the hardware operating environment involved in the base station scene recognition method based on the SVM improved model in the embodiments of this application.

[0015] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0018] It should be noted that the executing entity of this embodiment may include, but is not limited to, a server, a computer, or any terminal device capable of performing the above functions. This embodiment does not specifically limit this.

[0019] With the development of mobile communication networks and the large-scale deployment of 5G networks, base station site selection and planning, as an essential part of network construction, deserve attention.

[0020] Currently, base station scene identification is usually based on rule-based classification methods using geographic information system data, thereby enabling base station site selection and planning.

[0021] However, terrain data is usually quite complex, and the threshold judgment rules for different terrain categories in rule-based classification methods are usually quite simple, which leads to low accuracy in base station scene recognition.

[0022] Based on this, the embodiments of this application provide a base station scene recognition method based on an improved SVM model, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the base station scene recognition method proposed in this embodiment. The method includes steps 101 to 104: Step 101: Obtain the DEM data of the target site.

[0023] As an example, step 101 includes: Obtain the coordinate information of the target station, which may include longitude and latitude. Based on the target station's coordinate information, read the DEM data of the surrounding area from the DEM database. It should be noted that the DEM data can be used to characterize the topographic elevation information of the area surrounding the target station, and this DEM data can serve as the data basis for subsequent elevation feature extraction and topographic image generation.

[0024] Step 102: Extract elevation features from DEM data based on the base station coverage of the target site; Among them, elevation features can be used to describe the numerical characteristics of the terrain undulations around the target site, and to characterize the spatial form of the geographical environment in which the target site is located, so that subsequent steps can generate terrain images and identify base station scene types based on elevation features.

[0025] As an example, step 102 includes: Determine the site coordinate system corresponding to the target site; determine the corresponding sampling grid in the site coordinate system based on the base station coverage of the target site; generate an elevation matrix of the area surrounding the target site based on the sampling grid and DEM data; extract multiple valid elevation values ​​from the elevation matrix to obtain elevation features.

[0026] As an example, an elevation matrix of the area surrounding the target site is generated based on the sampling grid and DEM data, including: Each grid point in the sampling grid is mapped to the coordinate space of the DEM data, and the elevation value corresponding to each grid point is extracted from the DEM data. The extracted elevation values ​​are then filled into the positions of each grid point in the sampling grid to generate an elevation matrix of the area surrounding the target site.

[0027] The process of mapping each grid point in the sampling grid to the coordinate space of the DEM data includes: converting the station center-northeast celestial coordinates of each grid point in the sampling grid into geographic coordinates, with the Earth's center of mass as the origin and latitude, longitude and ellipsoidal height as the coordinate units; and locating the corresponding pixel position in the DEM data based on the converted geographic coordinates.

[0028] Extracting the elevation values ​​corresponding to each grid point from the DEM data can be achieved using bilinear interpolation. Specifically, the approximate elevation value of the target point is calculated by performing linear interpolation on the four nearest known pixels around the target point along both the X and Y axes. In other implementations, nearest neighbor interpolation or other deterministic interpolation methods can be used instead of bilinear interpolation.

[0029] Step 103: Generate an elevation image of the surrounding terrain of the target station based on the elevation features; As an example, step 103 includes: the elevation features include multiple valid elevation values; the valid elevation value with the smallest value is determined from the elevation features as the benchmark value; a relative elevation feature is generated based on the difference between the benchmark value and each valid elevation value, the relative elevation feature is used to eliminate the difference in absolute elevation benchmark of the station, so that the feature mainly describes the topographic relief around the station; the relative elevation feature is mapped to a grayscale image, and the grayscale image is used as the elevation image of the surrounding topography of the target station.

[0030] It should be noted that the surrounding terrain elevation image can be used to visualize image data representing the terrain undulation around the target site. Each pixel value in the image data corresponds to the terrain elevation information at the corresponding location in the sampling grid around the target site. The surrounding terrain elevation image is used to convert the elevation features extracted in step 102 into a visualized image format for subsequent image processing and base station scene classification and recognition based on image features.

[0031] As an example, mapping relative elevation features to a grayscale image and using the grayscale image as a surrounding terrain elevation image of the target site can include the following sub-steps: Gray-scale mapping is performed on the relative elevation difference to generate a surrounding terrain elevation image. This surrounding terrain elevation image is an 8-bit grayscale image, employing an 8-bit truncation quantization rule: for valid locations, let g_ij = min(max(round(Δh_ij),0),255), which maps the relative elevation difference to a grayscale value range of 0 to 255, obtaining the grayscale value for each valid location. For invalid locations, the validity mask generated in step 102 is used for marking (e.g., marking as a specific grayscale value or leaving it blank), and they are not included in the above mapping calculation. For flat areas with minimal topographic relief, the relative elevation difference approaches zero, corresponding to a grayscale value of 0, which appears as a black or near-black area in the grayscale image; for mountainous areas with significant topographic relief, the relative elevation difference is larger, corresponding to a higher grayscale value, which appears as a bright area in the grayscale image.

[0032] In other implementations, a trained and validated low-quantile elevation can be used as a robust reference benchmark. The reference benchmark, quantization, scaling, and invalid value handling methods are all written into the feature configuration.

[0033] Step 104: Identify the base station scene type corresponding to the target site based on the surrounding terrain elevation image.

[0034] Specifically, the base station scenario type is used to characterize the category of the geographical environment in which the target site is located. The category is associated with the base station signal propagation characteristics, coverage capabilities, and engineering parameter configurations. The base station scenario type includes at least two types: mountain base stations and non-mountain base stations.

[0035] As an example, step 104 includes: concatenating image feature vectors from surrounding terrain elevation images to obtain a target feature vector; using the target feature vector as model input, performing base station scene recognition on the target site based on a preset SVM model to obtain a scene classification label for the target site, wherein the scene classification label is used to identify the base station scene type.

[0036] In the online recognition stage, the normalized feature vector is input into the trained SVM model. The model calculates the decision value of each preset scene category corresponding to the target site based on the normalized feature vector, and determines the scene classification result of the target site based on the magnitude of the decision value.

[0037] This embodiment, through the above-described method, can generate surrounding terrain elevation images based on the acquired station elevation features and automatically identify the scene type to which the station belongs, providing an accurate scene classification basis for the differentiated configuration of subsequent engineering parameters. The reliability information is used to effectively block low-reliability classification results, preventing unreliable classification results from entering the downstream engineering parameter verification stage, thus improving the reliability of the system's engineering applications.

[0038] As an example, before concatenating the image feature vectors from the surrounding terrain elevation images to obtain the target feature vector, the method also includes: The system acquires the task information to be reviewed for the target site, which represents the site identifier, base station geographical location, base station type, antenna height, and planned coverage distance. It reads the model feature configuration corresponding to the SVM model and performs a matching verification between the model feature configuration and the task information to be reviewed. If the matching verification is successful, it performs the step of concatenating the image feature vectors from the surrounding terrain elevation image to obtain the target feature vector. If the matching verification fails, it outputs an indication of model recognition anomalies.

[0039] In this embodiment, DEM data of the target site can be acquired, and then elevation features can be extracted from the DEM data based on the base station coverage of the target site. This allows for the extraction of standard-sized elevation features from complex DEM data. Since the elevation features are extracted from the DEM data based on the base station coverage of the target site, they also match the base station coverage of the target site. The elevation features accurately characterize the surrounding terrain of the target site. Based on these elevation features, a surrounding terrain elevation image of the target site is generated, which fully, accurately, and intuitively characterizes the surrounding terrain. Therefore, based on the surrounding terrain elevation image, the base station scene type corresponding to the target site can be identified, ensuring the accuracy of base station scene identification. Furthermore, since the surrounding terrain elevation image is flexibly generated based on DEM data and base station coverage, the process of identifying the base station scene type corresponding to the target site in this embodiment can adapt to different types of site terrain. Compared to base station scene discrimination based on a single threshold judgment rule, the base station scene identification in this application has a wider range of adaptability and is more accurate, thus further ensuring the accuracy of base station scene identification and solving the technical problem of low accuracy in base station scene identification.

[0040] In some embodiments, refer to Figure 2 Before concatenating the image feature vectors from the surrounding terrain elevation images to obtain the target feature vector, the process includes: Step 301: Obtain the task information to be reviewed for the target site. The task information to be reviewed is used to characterize the site identifier, base station geographical location, base station type, antenna height and planned coverage distance. Step 302: Read the model feature configuration corresponding to the SVM model and match and verify the model feature configuration with the task information to be reviewed; Step 303: If the matching verification is successful, perform the step of concatenating the image feature vectors in the surrounding terrain elevation images to obtain the target feature vector. Step 304: If the matching verification fails, output an indication message indicating that the model identification is abnormal.

[0041] As an example, the model feature configuration and the information of the task to be reviewed are matched and verified. The information of the task to be reviewed may include, but is not limited to, at least one of the following: site identifier, base station geographical location, base station type, antenna height and planned coverage distance.

[0042] The system reads the jointly published feature configurations from the model and verifies the site category, decision value, effective DEM ratio, and hashes of the model, DEM, and feature configurations. If the verification is successful, steps 101 through 103 are executed sequentially for SVM model training and parameter tuning. If the verification fails (i.e., any joint identifier does not match), the system stops the automatic verification process, outputs a rule missing anomaly, and marks the target site's overall status as requiring manual review.

[0043] As an example, the training steps of the SVM model include: acquiring sample sites of multiple known site types, and performing steps 101 to 103 as described above for each sample site to obtain the surrounding terrain elevation image corresponding to each sample site. In this embodiment, the base station scene recognition system acquires the coordinate information of the target site, including samples of site identifiers, longitude, latitude, and artificial terrain labels. The labels include two categories: mountain sites and non-mountain sites. Auditing is performed on duplicate identifiers, missing coordinates, and label conflicts; multiple site identifiers at the same physical location are grouped by coordinates during data segmentation to avoid entering different data subsets.

[0044] The surrounding terrain elevation image, i.e., the grayscale image, undergoes dimensionality reduction processing. The grayscale image is scaled to a preset size according to feature configuration and flattened in a fixed row and column order. The scaled grayscale image is flattened in row and column order to obtain a one-dimensional grayscale vector, and each element in the one-dimensional grayscale vector is normalized. The normalized feature vector is used as the feature input for training samples, and manually labeled station category tags are used as supervision signals to train an SVM binary classification model.

[0045] Specifically, the image is scaled to 100×100 pixels, resulting in a grayscale image containing 10,000 pixels. In this embodiment, the 10,000-dimensional vector is flattened in a fixed row and column order, and each grayscale value is normalized by dividing by 25, ensuring that each element value falls within the [0,1] interval, thus obtaining a normalized feature vector, forming the training and test sets. Data is segmented by site coordinates to ensure that data from the same site does not appear simultaneously in both the training and test sets. The preset size can be set by those skilled in the art according to actual needs; when the preset size is changed, a new feature configuration version should be created, and the model should be retrained and validated.

[0046] This embodiment compares linear kernels and radial basis kernels, uses 5-fold hierarchical cross-validation grouped by coordinates within the training data to select the penalty coefficient C, kernel parameter gamma, and class weights, uses macro-average F1 as the parameter selection index, and uses classification accuracy as the optimization objective to determine the optimal model parameters; after training, the model file, DEM file, feature configuration file, data list, and validation results are packaged together into a joint release record.

[0047] In some embodiments, refer to Figure 3 After step 103, that is, after identifying the base station scene type corresponding to the target site based on the surrounding terrain elevation image, the method further includes: Step 401: Using the base station scenario type as an index, obtain multiple base station verification parameters corresponding to the target site. The base station verification parameters include at least the mounting height verification parameter and the coverage distance verification parameter. Step 402: Perform height verification on the target site according to the height verification parameters; Step 403: Perform coverage distance verification on the target site based on the coverage distance verification parameters.

[0048] Among them, the antenna height verification parameters include the lower limit and the upper limit of antenna height, and the coverage distance verification parameters include the lower limit and the upper limit of coverage distance. The antenna height verification parameters and the coverage distance verification parameters correspond one-to-one with the base station scenario type.

[0049] In this embodiment, after identifying the base station scenario type of the target site using the SVM model, the base station scenario recognition system uses the base station scenario type as an index to query and obtain the corresponding mounting height verification parameters and coverage distance verification parameters from the rule base. If the mounting height verification parameters and coverage distance verification parameters are successfully obtained, it indicates that the rule base has configured corresponding verification rules for the base station scenario type, and the system then executes the mounting height verification and coverage distance verification steps. If the mounting height verification parameters and coverage distance verification parameters are not obtained, it indicates that the current system lacks the basis for automatic verification of the target site, and the system stops the automatic verification process, outputs a rule missing anomaly, and marks the overall status of the target site as awaiting manual review.

[0050] It should be noted that the compliance ranges for antenna height and coverage distance differ for different base station scenario types. For example, the lower limit for antenna height at mountainous sites is usually lower than that at non-mountainous sites, while the upper limit for coverage distance is usually higher. If the base station scenario type is not differentiated and a uniform verification parameter is used to verify all sites, mountainous sites may be mistakenly judged as unqualified due to excessively low antenna height, or non-mountainous sites may be mistakenly judged as unqualified due to excessively long coverage distance.

[0051] Therefore, in this embodiment, the identified base station scenario type is used as an index to obtain the verification parameter that uniquely corresponds to the scenario type. This ensures that the verification standard is accurately matched with the actual terrain scenario of the site, avoids the misjudgment problem caused by using a single fixed threshold for verification, avoids the inefficiency caused by manually selecting verification rules for each site, and significantly improves the accuracy and scenario adaptability of engineering parameter verification.

[0052] In some embodiments, the scene classification result includes credibility information. Based on the credibility information, engineering parameter verification is performed on the target site, including: obtaining verification rules and / or processing strategies corresponding to the target site based on the credibility information, wherein the coverage dimension of the verification rules is inversely proportional to the credibility, and the degree of human intervention in the processing strategy is inversely proportional to the credibility. The engineering parameters of the target site are then verified using the verification rules and / or processing strategies.

[0053] The aforementioned credibility information can be represented using numerical labels. For example, a numerical label of 0 to 1 can be used to represent credibility, where 0 indicates that the model is completely unreliable of the classification result, 1 indicates that the model is completely reliable of the classification result, and values ​​between 0 and 1 indicate partial reliability. A preset gating threshold can be, for example, 0.6. When the credibility is below 0.6, the model lacks sufficient confidence in the classification result, triggering a manual review branch; when the credibility is above 0.6, the model is allowed to enter the automatic rule branch.

[0054] Therefore, in this embodiment, appropriate verification rules and / or processing strategies can be selected based on the credibility information. The coverage dimension of the verification rules is directly proportional to the credibility; that is, the higher the credibility, the more lenient the verification and the fewer the coverage dimensions. The degree of manual intervention in the processing strategy is inversely proportional to the credibility; that is, the higher the credibility, the greater the probability of automatic approval. Thus, by using verification rules and / or processing strategies to verify the engineering parameters of the target site, the reliability and security of the verification can be considered while ensuring verification efficiency. Differentiated processing under different credibility levels can be achieved, avoiding the automatic entry of low-credibility identification results into the engineering parameter verification stage, which would lead to the adoption of erroneous conclusions. At the same time, it can also avoid the reduction of overall processing efficiency due to over-verification of high-credibility identification results.

[0055] The base station scene recognition device based on the improved SVM model provided in this application adopts the base station scene recognition method based on the improved SVM model in the above embodiments, which can solve the technical problem of low accuracy of base station scene recognition.

[0056] As an example, refer to Figure 4 , Figure 4 This is a schematic diagram of the module composition architecture of the base station scene recognition device based on the SVM improved model in this embodiment. The base station scene recognition device includes a feature extraction module 201, an image generation module 202, and a scene recognition module 203, wherein: The feature extraction module 201 is used to acquire the DEM data of the target site and extract elevation features from the DEM data according to the base station coverage of the target site; the image generation module 202 is used to generate the surrounding terrain elevation image of the target site according to the elevation features; the scene recognition module 203 is used to identify the base station scene type corresponding to the target site according to the surrounding terrain elevation image.

[0057] Optionally, the base station scene recognition device may further include an engineering verification module, which is used to obtain multiple base station verification parameters corresponding to the target site using the base station scene type as an index, wherein the base station verification parameters include at least a mounting height verification parameter and a coverage distance verification parameter; perform mounting height verification on the target site according to the mounting height verification parameter; and perform coverage distance verification on the target site according to the coverage distance verification parameter.

[0058] This application provides a base station scene recognition device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the base station scene recognition method in the above embodiment 1.

[0059] The following is for reference. Figure 5 This document illustrates a structural schematic diagram of a base station scene recognition device suitable for implementing embodiments of this application. The base station scene recognition device in these embodiments may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle-mounted terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The base station scene recognition device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0060] like Figure 5As shown, the base station scene recognition device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the base station scene recognition device based on the SVM improved model. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the base station scene recognition device to exchange data wirelessly or via wired communication with other devices. Although base station scene recognition devices with various systems are shown in the figures, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented alternatively.

[0061] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0062] The base station scene recognition device provided in this application employs the base station scene recognition method based on the improved SVM model in the above embodiments, which can solve the technical problem of low accuracy in base station scene recognition. Compared with the prior art, the beneficial effects of the base station scene recognition device based on the improved SVM model provided in this application are the same as those of the base station scene recognition method based on the improved SVM model provided in the above embodiments, and other technical features in the base station scene recognition device based on the improved SVM model are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0063] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0064] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0065] In addition, this application also provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to perform the steps of the base station scene recognition method in the above embodiments.

[0066] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0067] The aforementioned computer-readable storage medium may be included in a base station scene recognition device based on an improved SVM model; or it may exist independently and not be assembled into a base station scene recognition device based on an improved SVM model.

[0068] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the base station scene recognition device based on the SVM improved model, enable the base station scene recognition device based on the SVM improved model to perform base station scene recognition.

[0069] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0070] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0071] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0072] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned base station scene recognition method based on the improved SVM model, thereby solving the technical problem of low accuracy in base station scene recognition. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the base station scene recognition method based on the improved SVM model provided in the above embodiments, and will not be repeated here.

[0073] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the base station scene recognition method based on the improved SVM model described above.

[0074] The computer program product provided in this application can solve the technical problem of low accuracy in base station scene recognition. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the base station scene recognition method based on the SVM improved model provided in the above embodiments, and will not be repeated here.

[0075] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A base station scene recognition method based on an improved SVM model, characterized in that, The method includes the following steps: Obtain the DEM data of the target site; Based on the base station coverage of the target site, extract elevation features from the DEM data; Based on the elevation features, generate an elevation image of the surrounding terrain of the target station; Based on the surrounding terrain elevation image, the base station scene type corresponding to the target site is identified.

2. The method as described in claim 1, characterized in that, The step of extracting elevation features from the DEM data based on the base station coverage of the target site includes: Determine the station coordinate system corresponding to the target station; Based on the base station coverage area of ​​the target site, determine the corresponding sampling grid in the site coordinate system; Based on the sampling grid and the DEM data, an elevation matrix of the area surrounding the target site is generated; Multiple valid elevation values ​​are extracted from the elevation matrix to obtain elevation features.

3. The method as described in claim 1, characterized in that, The elevation features include multiple valid elevation values; generating a surrounding terrain elevation image of the target station based on the elevation features includes: The effective elevation value with the smallest value among the elevation features is determined as the benchmark value; A relative elevation feature is generated based on the difference between the benchmark value and each of the effective elevation values; The relative elevation feature is mapped to a grayscale image, and the grayscale image is used as the surrounding terrain elevation image of the target station.

4. The method as described in claim 1, characterized in that, The step of identifying the base station scene type corresponding to the target site based on the surrounding terrain elevation image includes: The image feature vectors in the surrounding terrain elevation images are concatenated to obtain the target feature vector; Using the target feature vector as model input, base station scene identification is performed on the target site based on a preset SVM model to obtain the scene classification label of the target site, wherein the scene classification label is used to identify the base station scene type.

5. The method as described in claim 4, characterized in that, Before concatenating the image feature vectors from the surrounding terrain elevation images to obtain the target feature vector, the method further includes: Obtain the pending task information of the target site, wherein the pending task information is used to characterize the site identifier, base station geographical location, base station type, antenna height and planned coverage distance of the site; Read the model feature configuration corresponding to the SVM model, and match and verify the model feature configuration with the task information to be reviewed; If the matching verification is successful, the step of concatenating the image feature vectors in the surrounding terrain elevation image to obtain the target feature vector is executed. If the matching verification fails, an indication message indicating that the model identification is abnormal will be output.

6. The method as described in claim 1, characterized in that, After identifying the base station scene type corresponding to the target site based on the surrounding terrain elevation image, the method further includes: Using the base station scenario type as an index, obtain multiple base station verification parameters corresponding to the target site, wherein the base station verification parameters include at least the mounting height verification parameter and the coverage distance verification parameter; The target site is subjected to hanging height verification based on the hanging height verification parameters; The coverage distance of the target site is verified based on the coverage distance verification parameters.

7. A base station scene recognition device based on an improved SVM model, characterized in that, The base station scene recognition device based on the SVM improved model includes: The feature extraction module is used to acquire the DEM data of the target site and extract elevation features from the DEM data according to the base station coverage of the target site. An image generation module is used to generate an elevation image of the surrounding terrain of the target station based on the elevation features; The scene recognition module is used to identify the base station scene type corresponding to the target site based on the surrounding terrain elevation image.

8. A base station scene recognition device based on an improved SVM model, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the base station scene recognition method based on the SVM improved model as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the base station scene recognition method based on the SVM improved model as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the base station scene recognition method based on the SVM improved model as described in any one of claims 1 to 6.

Images