A semantic segmentation-based WebGIS map annotation automatic avoidance method and device
By adopting a WebGIS map annotation automatic avoidance method based on semantic segmentation, the problem of overlapping and occlusion between annotations and map features is solved, achieving efficient annotation display and information transmission, which is applicable to scenarios such as urban planning, traffic management and emergency dispatch.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 武汉智博创享科技股份有限公司
- Filing Date
- 2026-03-18
- Publication Date
- 2026-06-19
Smart Images
Figure CN122244228A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of map rendering technology, specifically referring to a method and device for automatic obstacle avoidance in WebGIS map annotations based on semantic segmentation. Background Technology
[0002] WebGIS (Web Geographic Information System) serves as a crucial medium for displaying and interacting with geographic information, and is widely used in various fields such as urban planning, traffic management, and emergency dispatch. Map annotations, as a key element in WebGIS for conveying geographic information, can intuitively label core information such as place names, road names, and water system names on maps. Their display quality directly impacts the efficiency and user experience of acquiring geographic information.
[0003] Existing WebGIS map annotation display technologies commonly suffer from inaccurate annotation avoidance and low levels of intelligence. Traditional annotation avoidance methods often employ rule-based static avoidance strategies, such as fixing annotation offset distances and hiding overlapping annotations in a preset order, without fully considering the differences in map feature types and spatial distribution characteristics. When the map zoom level changes or map features are densely distributed, annotations easily overlap with map features, and annotations may obscure each other, making it difficult for users to clearly identify annotation information and thus affecting the effective transmission of geographic information.
[0004] Traditional annotation avoidance methods are not flexible enough to adapt to the annotation display needs of different application scenarios and cannot meet users' demands for improved map annotation display accuracy and intelligence. Therefore, a WebGIS map annotation automatic avoidance method and device based on semantic segmentation is proposed. Summary of the Invention
[0005] To address the above issues, this paper proposes an automatic obstacle avoidance method and device for WebGIS map annotations based on semantic segmentation to overcome the shortcomings of existing technologies, such as inaccurate obstacle avoidance, low level of intelligence, and inability to adapt to different map element types and application scenarios.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method and apparatus for automatic obstacle avoidance in WebGIS map annotation based on semantic segmentation, comprising: The system includes a map data acquisition module, a semantic segmentation module, an annotation information extraction module, an obstacle avoidance decision-making module, and an annotation rendering output module; among which, The map data acquisition module is used to acquire map data to be processed from the WebGIS platform. The map data to be processed includes map image data and initial annotation data. The semantic segmentation module is connected to the map data acquisition module and is used to perform semantic segmentation processing on the map image data to obtain map element semantic segmentation results. The map element semantic segmentation results include spatial location and range information of different types of map elements. The annotation information extraction module is connected to the map data acquisition module and is used to extract annotation core information from the initial annotation data. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. The avoidance decision module is connected to the semantic segmentation module and the annotation information extraction module, and includes a conflict detection unit and an avoidance strategy generation unit. The conflict detection unit is used to determine whether there is a spatial conflict between the initial annotation and the map element, and between the initial annotation itself, based on the semantic segmentation results of the map element and the core information of the annotation. The avoidance strategy generation unit is used to generate a corresponding annotation avoidance adjustment scheme for annotations with spatial conflicts, by combining the priority of the map element and the priority of the annotation. The annotation rendering output module is connected to the avoidance decision module. It is used to adjust the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and then output the adjusted annotation to the WebGIS platform after fusion rendering with the map image data.
[0007] As a further description of the above technical solution: Includes the following steps: S1. Obtain the map data to be processed from the WebGIS platform through the map data acquisition module. The map data to be processed includes map image data and initial annotation data. S2. The map image data is semantically segmented using the semantic segmentation module to obtain map element semantic segmentation results containing spatial location and range information of different types of map elements. S3. Extract annotation core information from the initial annotation data through the annotation information extraction module. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. S4. Through the conflict detection unit of the avoidance decision module, combined with the semantic segmentation results of the map elements and the core information of the annotation, determine whether there is a spatial conflict between the initial annotation and the map elements, and between the initial annotation; if there is a spatial conflict, through the avoidance strategy generation unit of the avoidance decision module, combined with the priority of the map elements and the priority of the annotation, generate an annotation avoidance adjustment scheme. S5. The annotation rendering output module adjusts the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and after the adjusted annotation is fused and rendered with the map image data, it is output to the WebGIS platform.
[0008] As a further description of the above technical solution: The specific process of semantic segmentation in step S2 is as follows: input map image data into a preset semantic segmentation model, identify different types of elements such as roads, buildings, water systems, and green spaces in the map through the semantic segmentation model, output the mask image and spatial coordinate range corresponding to each element, and form the semantic segmentation result of map elements.
[0009] As a further description of the above technical solution: The specific process for determining whether there is a spatial conflict between the initial annotation and map elements, and between the initial annotation, in step S4 is as follows: based on the spatial coordinate range of map elements and the initial position coordinates and annotation size parameters of the annotation, calculate the overlapping area between the annotation area and each map element area, as well as the overlapping area between different annotation areas. If the overlapping area is greater than a preset threshold, a spatial conflict is determined to exist.
[0010] As a further description of the above technical solution: The priorities of map elements and annotations mentioned in step S4 are either preset priorities or dynamically adjusted priorities based on the application scenario of the WebGIS platform. The priority of map elements, from highest to lowest, is roads, buildings, water systems, and green spaces; the priority of annotations is set synchronously according to the priority of the map elements corresponding to the annotations.
[0011] As a further description of the above technical solution: The annotation avoidance adjustment scheme mentioned in step S4 includes a position adjustment scheme and a size adjustment scheme; When a label conflicts with a high-priority map feature, a position adjustment scheme is adopted to shift the label to a non-conflicting area around the conflicting feature; when labels conflict with each other and have the same priority, a size adjustment scheme that reduces the size proportionally or a position adjustment scheme that shifts alternately is adopted.
[0012] As a further description of the above technical solution: The semantic segmentation module uses a deep learning-based semantic segmentation model, which is trained on a map image sample dataset containing map image samples labeled with feature types.
[0013] As a further description of the above technical solution: It also includes a priority configuration module, which is connected to the avoidance decision module. The priority configuration module is used to allow users to preset or dynamically adjust the priority of map elements and annotations, and transmit the configured priority information to the avoidance decision module.
[0014] This invention adopts the following technical solution: an automatic obstacle avoidance device for WebGIS map annotation based on semantic segmentation, comprising: The components include map data acquisition, semantic segmentation, annotation information extraction, obstacle avoidance decision-making, and annotation rendering output. The map data acquisition component is used to acquire map data to be processed from the WebGIS platform. The map data to be processed includes map image data and initial annotation data. The semantic segmentation component is connected to the map data acquisition component and is used to perform semantic segmentation processing on the map image data to obtain map element semantic segmentation results. The map element semantic segmentation results include spatial location and range information of different types of map elements. The annotation information extraction component is connected to the map data acquisition component and is used to extract annotation core information from the initial annotation data. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. The avoidance decision component is connected to the semantic segmentation component and the annotation information extraction component, respectively. It is used to determine whether there is a spatial conflict between the initial annotation and the map element, and between the initial annotation, based on the semantic segmentation result of the map element and the core information of the annotation. For annotations with spatial conflicts, it generates a corresponding annotation avoidance adjustment scheme by combining the priority of the map element and the priority of the annotation. The annotation rendering output component is connected to the avoidance decision component and is used to adjust the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and then output the adjusted annotation to the WebGIS platform after fusion rendering with the map image data. The device also includes a priority configuration component, which is connected to the avoidance decision component. The priority configuration component is used to allow users to preset or dynamically adjust the priority of map elements and annotations, and to transmit the configured priority information to the avoidance decision component.
[0015] The beneficial effects achieved by the present invention using the above structure are as follows: (1) In this invention, by introducing a semantic segmentation module based on deep learning, different types of elements such as roads, buildings, water systems, and green spaces in map image data can be accurately identified, and semantic segmentation results containing spatial location and range information can be output. This completely solves the pain points of traditional methods that cannot distinguish map element types and have a single avoidance logic. Combined with the conflict detection unit based on the quantitative judgment of spatial coordinate range and overlapping area, the conflict scenarios between annotations and map elements and between annotations can be accurately located. Combined with the differentiated avoidance strategy—position offset for high-priority map elements and size reduction or alternating offset for annotations of the same priority—the core map elements are not obscured, and the annotation information is preserved to the maximum extent, which greatly improves the display clarity and information transmission efficiency of map annotations.
[0016] (2) In this invention, the level of intelligence and flexibility of the technical solution are enhanced. This invention constructs a complete technical chain through modular design. From map data acquisition, semantic segmentation, annotation information extraction to avoidance decision-making and rendering output, each module works together to achieve fully automated processing. Conflict detection and avoidance adjustment can be completed without manual intervention, significantly reducing operating costs. At the same time, the added priority configuration module supports users to preset or dynamically adjust the priority of map elements and annotations. The semantic segmentation model can adapt to map data with different scaling ratios and different element densities. The preset threshold can also be flexibly adjusted according to the display accuracy requirements, making it widely adaptable to various WebGIS application scenarios such as urban planning, traffic management, and emergency dispatch, solving the problem of poor adaptability of traditional static avoidance methods. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the device for an automatic obstacle avoidance method for WebGIS map annotation based on semantic segmentation proposed in this invention. Detailed Implementation
[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0019] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.
[0020] To address the problems of inaccurate obstacle avoidance, low intelligence, and inability to adapt to different application scenarios in existing WebGIS map annotations, this application proposes a method and apparatus for automatic obstacle avoidance in WebGIS map annotations based on semantic segmentation. Figure 1 As shown, an automatic obstacle avoidance device for WebGIS map annotations based on semantic segmentation includes a map data acquisition module, a semantic segmentation module, an annotation information extraction module, an obstacle avoidance decision module, and an annotation rendering output module. The map data acquisition module acquires map data to be processed from the WebGIS platform, including map image data and initial annotation data. The semantic segmentation module, connected to the map data acquisition module, performs semantic segmentation on the map image data to obtain semantic segmentation results for map elements. These results include spatial location and extent information for different types of map elements. The annotation information extraction module, also connected to the map data acquisition module, extracts core annotation information from the initial annotation data, including annotation text. The system comprises the following modules: content, annotation size parameters, and initial annotation coordinates; the avoidance decision module, connected to both the semantic segmentation module and the annotation information extraction module, is the core module, including a conflict detection unit and an avoidance strategy generation unit; the conflict detection unit determines whether there is a spatial conflict between the initial annotation and map elements, and between the initial annotation and the annotation itself, based on the semantic segmentation results of map elements and the core information of the annotation; the avoidance strategy generation unit generates corresponding annotation avoidance adjustment schemes for annotations with spatial conflicts, combining the priorities of map elements and annotations; and the annotation rendering output module, connected to the avoidance decision module, adjusts the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and then merges and renders the adjusted annotation with map image data before outputting it to the WebGIS platform.
[0021] An automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation includes the following steps: S1. Obtaining map data to be processed from the WebGIS platform through a map data acquisition module. The map data to be processed includes map image data and initial annotation data. S2. Performing semantic segmentation on the map image data through a semantic segmentation module to obtain map element semantic segmentation results containing spatial location and extent information of different types of map elements. S3. Extracting core annotation information from the initial annotation data through an annotation information extraction module. The core annotation information includes annotation text content, annotation size parameters, and initial position coordinates of the annotation. S4. Determining whether there is a spatial conflict between the initial annotation and map elements, and between the initial annotation and the map elements, through the conflict detection unit of the obstacle avoidance decision module, combined with the map element semantic segmentation results and the core annotation information. If a spatial conflict exists, generating an annotation obstacle avoidance adjustment scheme through the obstacle avoidance strategy generation unit of the obstacle avoidance decision module, combined with map element priority and annotation priority. S5. Adjusting the position or size of the initial annotation according to the annotation obstacle avoidance adjustment scheme through an annotation rendering output module, fusing and rendering the adjusted annotation with the map image data, and outputting it to the WebGIS platform.
[0022] In this way, the semantic segmentation module can perform accurate semantic segmentation on map image data, accurately identify the spatial location and range of different types of map elements, provide accurate data support for the subsequent formulation of differentiated avoidance strategies, and effectively solve the problem of inaccurate avoidance caused by the inability of traditional methods to distinguish map element types.
[0023] Furthermore, the avoidance decision module generates differentiated avoidance adjustment schemes by combining the priority of map elements and annotations. It adopts position adjustment or size adjustment methods for different conflict scenarios, which not only ensures the clear display of high-priority map elements and annotations, but also minimizes annotation occlusion, thereby improving the display effect of map annotations and the efficiency of information transmission.
[0024] The automatic obstacle avoidance method and apparatus for WebGIS map annotation based on semantic segmentation in the embodiments of this application can be applied to scenarios such as urban planning WebGIS systems, traffic management WebGIS platforms, and emergency dispatch geographic information systems. Specifically, it is not limited to WebGIS applications in the fields of urban public services, land and resources management, and environmental monitoring.
[0025] See Figure 1 ; The system comprises the following modules: a map data acquisition module, which acquires map data to be processed from the WebGIS platform, including map image data and initial annotation data; a semantic segmentation module, connected to the map data acquisition module, which performs semantic segmentation on the map image data; an annotation information extraction module, also connected to the map data acquisition module, which extracts core annotation information; an avoidance decision module, which is the core module for conflict detection and avoidance scheme generation, including at least a conflict detection unit and an avoidance strategy generation unit; an annotation rendering output module, connected to the avoidance decision module, which adjusts and renders annotations; and a priority configuration module, also connected to the avoidance decision module, which configures priority information.
[0026] The avoidance decision module refers to the module that determines conflicts and generates avoidance strategies based on the semantic information and annotation information of map elements. It includes at least a conflict detection unit and an avoidance strategy generation unit. The conflict detection unit is used to determine whether there is a spatial conflict between the annotation and the map element, or between the annotations themselves; the avoidance strategy generation unit is used to generate avoidance adjustment schemes for conflicting annotations.
[0027] In one embodiment, the semantic segmentation module uses a deep learning-based semantic segmentation model, which is trained on a map image sample dataset containing map image samples labeled with feature types.
[0028] In one embodiment, the device further includes a priority configuration module, which is connected to the avoidance decision module and is used to allow users to preset or dynamically adjust the priority of map elements and annotations, and to transmit the configured priority information to the avoidance decision module.
[0029] In one embodiment, the annotation avoidance adjustment scheme includes a location adjustment scheme and a size adjustment scheme. When an annotation conflicts with a high-priority map feature, the location adjustment scheme is used to shift the annotation to a non-conflicting area around the conflicting feature. When annotations conflict with each other and have the same priority, a size adjustment scheme that reduces the size proportionally or a location adjustment scheme that shifts alternately is used.
[0030] To better understand the working process of the automatic obstacle avoidance method and apparatus for WebGIS map annotation based on semantic segmentation according to the embodiments of this application, a specific embodiment is described below: First, the map data to be processed is acquired from the urban traffic management WebGIS platform through the map data acquisition module. This data includes map image data of a certain area, containing elements such as roads, buildings, water systems, and green spaces, as well as initial annotation data, including annotation information such as road names, community names, and river names. Next, the map image data is input into a deep learning-based semantic segmentation model. The semantic segmentation model identifies elements such as roads, buildings, water systems, and green spaces in the map, and outputs the mask images and spatial coordinate ranges of each element, forming the semantic segmentation result of map elements. Then, the annotation information extraction module extracts the annotation text content, annotation size parameters, and initial location coordinates (such as latitude and longitude coordinates) from the initial annotation data. Finally, the conflict detection unit, based on the spatial coordinate range of map elements, and... The annotation module records core information, calculates the overlap area between the annotation area and each map feature area, and the overlap area between different annotation areas. If the overlap area is greater than a preset threshold, a spatial conflict is determined. If a spatial conflict exists, the avoidance strategy generation unit generates an avoidance adjustment scheme based on preset priorities (map feature priority: roads > buildings > water systems > green spaces; annotation priority is synchronized with the corresponding feature priority). For example, if the annotation "Chaoyang Road" conflicts with a building feature, and the building feature has a higher priority than the annotation, a position adjustment scheme is adopted to shift the "Chaoyang Road" annotation 15px to the non-conflicting area around the building. Finally, the annotation rendering output module adjusts the annotation position according to the avoidance adjustment scheme, merges and renders the adjusted annotation with the map image data, and outputs it to the urban traffic management WebGIS platform.
[0031] In some embodiments, the semantic segmentation model can be the U-Net model, the DeepLabV3+ model, etc., and is not specifically limited; the preset threshold can be dynamically adjusted according to the display accuracy requirements of the WebGIS platform, for example, the threshold can be set to 3px in a high-definition display scenario. 2 In normal display scenarios, set to 8px 2 .
[0032] Optionally, the priority configuration module can provide a visual operation interface, through which users can directly drag and drop to adjust the priority order of map elements and annotations, or import preset priority configuration files, with no restrictions on the specific operation method.
[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
[0035] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.
Claims
1. A method for automatic obstacle avoidance in WebGIS map annotations based on semantic segmentation, characterized in that, include: The system includes a map data acquisition module, a semantic segmentation module, an annotation information extraction module, an obstacle avoidance decision-making module, and an annotation rendering output module; among which, The map data acquisition module is used to acquire map data to be processed from the WebGIS platform. The map data to be processed includes map image data and initial annotation data. The semantic segmentation module is connected to the map data acquisition module and is used to perform semantic segmentation processing on the map image data to obtain map element semantic segmentation results. The map element semantic segmentation results include spatial location and range information of different types of map elements. The annotation information extraction module is connected to the map data acquisition module and is used to extract annotation core information from the initial annotation data. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. The avoidance decision module is connected to the semantic segmentation module and the annotation information extraction module, and includes a conflict detection unit and an avoidance strategy generation unit. The conflict detection unit is used to determine whether there is a spatial conflict between the initial annotation and the map element, and between the initial annotation itself, based on the semantic segmentation results of the map element and the core information of the annotation. The avoidance strategy generation unit is used to generate a corresponding annotation avoidance adjustment scheme for annotations with spatial conflicts, by combining the priority of the map element and the priority of the annotation. The annotation rendering output module is connected to the avoidance decision module. It is used to adjust the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and then output the adjusted annotation to the WebGIS platform after fusion rendering with the map image data.
2. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation as described in claim 1, Its features are, Includes the following steps: S1. Obtain the map data to be processed from the WebGIS platform through the map data acquisition module. The map data to be processed includes map image data and initial annotation data. S2. The map image data is semantically segmented using the semantic segmentation module to obtain map element semantic segmentation results containing spatial location and range information of different types of map elements. S3. Extract annotation core information from the initial annotation data through the annotation information extraction module. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. S4. Through the conflict detection unit of the avoidance decision module, combined with the semantic segmentation results of the map elements and the core information of the annotation, determine whether there is a spatial conflict between the initial annotation and the map elements, and between the initial annotation; if there is a spatial conflict, through the avoidance strategy generation unit of the avoidance decision module, combined with the priority of the map elements and the priority of the annotation, generate an annotation avoidance adjustment scheme. S5. The annotation rendering output module adjusts the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and after the adjusted annotation is fused and rendered with the map image data, it is output to the WebGIS platform.
3. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 2, characterized in that, The specific process of semantic segmentation in step S2 is as follows: input map image data into a preset semantic segmentation model, identify different types of elements such as roads, buildings, water systems, and green spaces in the map through the semantic segmentation model, output the mask image and spatial coordinate range corresponding to each element, and form the semantic segmentation result of map elements.
4. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 3, characterized in that, The specific process for determining whether there is a spatial conflict between the initial annotation and map elements, and between the initial annotation, in step S4 is as follows: based on the spatial coordinate range of map elements and the initial position coordinates and annotation size parameters of the annotation, calculate the overlapping area between the annotation area and each map element area, as well as the overlapping area between different annotation areas. If the overlapping area is greater than a preset threshold, a spatial conflict is determined to exist.
5. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 4, characterized in that, The priorities of map elements and annotations mentioned in step S4 are either preset priorities or dynamically adjusted priorities based on the application scenario of the WebGIS platform. The priority of map elements, from highest to lowest, is roads, buildings, water systems, and green spaces; the priority of annotations is set synchronously according to the priority of the map elements corresponding to the annotations.
6. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 5, characterized in that, The annotation avoidance adjustment scheme mentioned in step S4 includes a position adjustment scheme and a size adjustment scheme; When a label conflicts with a high-priority map feature, a position adjustment scheme is adopted to shift the label to a non-conflicting area around the conflicting feature; when labels conflict with each other and have the same priority, a size adjustment scheme that reduces the size proportionally or a position adjustment scheme that shifts alternately is adopted.
7. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 6, characterized in that, The semantic segmentation module uses a deep learning-based semantic segmentation model, which is trained on a map image sample dataset containing map image samples labeled with feature types.
8. The automatic obstacle avoidance method for WebGIS map annotations based on semantic segmentation according to claim 7, characterized in that, It also includes a priority configuration module, which is connected to the avoidance decision module. The priority configuration module is used to allow users to preset or dynamically adjust the priority of map elements and annotations, and transmit the configured priority information to the avoidance decision module.
9. An automatic obstacle avoidance device for WebGIS map annotation based on semantic segmentation, characterized in that, include: The components include map data acquisition, semantic segmentation, annotation information extraction, obstacle avoidance decision-making, and annotation rendering output. The map data acquisition component is used to acquire map data to be processed from the WebGIS platform. The map data to be processed includes map image data and initial annotation data. The semantic segmentation component is connected to the map data acquisition component and is used to perform semantic segmentation processing on the map image data to obtain map element semantic segmentation results. The map element semantic segmentation results include spatial location and range information of different types of map elements. The annotation information extraction component is connected to the map data acquisition component and is used to extract annotation core information from the initial annotation data. The annotation core information includes annotation text content, annotation size parameters, and annotation initial position coordinates. The avoidance decision component is connected to the semantic segmentation component and the annotation information extraction component, respectively. It is used to determine whether there is a spatial conflict between the initial annotation and the map element, and between the initial annotation, based on the semantic segmentation result of the map element and the core information of the annotation. For annotations with spatial conflicts, it generates a corresponding annotation avoidance adjustment scheme by combining the priority of the map element and the priority of the annotation. The annotation rendering output component is connected to the avoidance decision component and is used to adjust the position or size of the initial annotation according to the annotation avoidance adjustment scheme, and then output the adjusted annotation to the WebGIS platform after fusion rendering with the map image data. The device also includes a priority configuration component, which is connected to the avoidance decision component. The priority configuration component is used to allow users to preset or dynamically adjust the priority of map elements and annotations, and to transmit the configured priority information to the avoidance decision component.