Remote sensing image geographic element automatic identification method and system based on large model analysis

Abstract

The invention discloses a remote sensing image geographic element automatic identification method and system based on large model analysis, and relates to the image processing related technical field, and the method comprises the steps: obtaining remote sensing image data according to a plurality of remote sensing data sources based on a remote sensing data processing rule; obtaining a geographic element identification instruction; performing multi-level image segmentation constraint registration, and establishing an image segmentation constraint tree; performing segmentation processing on the remote sensing image data according to the SAM image segmentation large model to obtain a multi-level remote sensing image segmentation result; performing segmentation loss optimization to obtain a multi-level remote sensing image segmentation optimization result; and carrying out vectorization element identification to obtain a multi-level geographic element identification result. The technical problems that in the prior art, remote sensing image multi-level geographic element recognition precision is insufficient, efficiency is not high, and image segmentation is not accurate are solved, efficient and accurate geographic element recognition is achieved, and the technical effect of improving geographic element recognition efficiency and accuracy is achieved.

Description

Technical Field

[0001] The present application relates to the technical field related to image processing, and specifically to a method and system for automatically identifying geographic elements in remote sensing images based on large model analysis. Background Art

[0002] Remote sensing image geographic feature recognition is an important application of remote sensing technology in geographic information system (GIS), and is widely used in urban planning, agricultural monitoring, environmental monitoring, resource exploration and other fields. As an important means of obtaining ground information, remote sensing images collect image data of the earth's surface through sensors such as satellites and aerial platforms, which can provide rich information for geographic feature recognition. However, since remote sensing images often have complex geographical backgrounds, different types of objects and different spatial resolutions, traditional remote sensing image geographic feature recognition methods mostly rely on pixel-based image classification methods. Generally, feature extraction of image data, such as spectral features, texture features, shape features, etc., is used to classify and recognize geographic features. It is difficult to handle complex image features, especially when the image quality is poor or there is noise. The spatial context information of remote sensing images cannot be fully utilized, resulting in low classification accuracy in complex scenes. In addition, for complex multi-level and multi-scale geographic feature information, such as urban buildings, farmland, water bodies, forests, etc., traditional geographic feature recognition methods cannot handle such diverse and complex remote sensing images. For example, it is difficult to automatically identify and accurately locate geographic features in nature reserves, post-disaster assessment areas and other areas.

[0003] Therefore, in the current relevant technologies, there are technical problems such as insufficient accuracy in identifying multi-level geographic elements in remote sensing images, low efficiency, and inaccurate image segmentation. Summary of the invention

[0004] This application solves the technical problems of insufficient precision, low efficiency and inaccurate image segmentation in the prior art of multi-level geographic element recognition in remote sensing images by providing a method and system for automatic identification of geographic elements in remote sensing images based on large model analysis, thereby achieving efficient and accurate identification of geographic elements and achieving the technical effect of improving the efficiency and accuracy of geographic element recognition.

[0005] The present application provides a method for automatic identification of geographic elements in remote sensing images based on large model analysis, the method comprising: obtaining remote sensing image data based on remote sensing data processing rules and according to multiple remote sensing data sources; obtaining geographic element identification instructions, wherein the geographic element identification instructions include multi-level element identification targets, and the multi-level element identification targets include full elements, prompt word elements and point of interest elements; performing multi-level image segmentation constraint alignment on the remote sensing image data according to the multi-level element identification targets, and establishing an image segmentation constraint tree; based on the image segmentation constraint tree, performing segmentation processing on the remote sensing image data according to the SAM image segmentation large model, and obtaining a multi-level remote sensing image segmentation result; performing segmentation loss optimization on the multi-level remote sensing image segmentation result according to an image segmentation loss detection model, and obtaining a multi-level remote sensing image segmentation optimization result; performing vectorized element recognition according to the multi-level remote sensing image segmentation optimization result, and obtaining a multi-level geographic element recognition result.

[0006] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis also performs the following processing: performing image segmentation constraint parameter analysis on the remote sensing image data according to the full elements to obtain a first image segmentation constraint condition; performing image segmentation constraint parameter analysis on the remote sensing image data according to the prompt word elements to obtain a second image segmentation constraint condition; performing image segmentation constraint parameter analysis on the remote sensing image data according to the point of interest elements to obtain a third image segmentation constraint condition; using the full elements as the first node for geographic element identification, the prompt word elements as the second node for geographic element identification, and the point of interest elements as the third node for geographic element identification; constructing an element identification node tree according to the first geographic element identification node, the second geographic element identification node, and the third geographic element identification node; performing image segmentation constraint configuration on the element identification node tree according to the first image segmentation constraint condition, the second image segmentation constraint condition, and the third image segmentation constraint condition to generate the image segmentation constraint tree.

[0007] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis further performs the following processing: performing twin data retrieval according to the remote sensing image data to obtain Q remote sensing image twin samples, where Q is a positive integer greater than 1; Collect historical parameters of the full-factor image segmentation constraints corresponding to the Q remote sensing image twin samples, and obtain Q full-factor segmentation constraint groups, wherein each full-factor segmentation constraint group includes M full-factor image segmentation constraint samples corresponding to each remote sensing image twin sample, and M is a positive integer greater than 1; based on the global confidence evaluation channel, perform global confidence evaluation on the Q full-factor segmentation constraint groups to obtain global confidences of the Q constraint groups; based on the global confidences of the Q constraint groups, select the Q full-factor segmentation constraint groups according to a predetermined global confidence, and establish a full-factor segmentation constraint space; perform centralized value calculation based on the full-factor segmentation constraint space to generate the first image segmentation constraint condition, wherein the first image segmentation constraint condition includes M image segmentation constraint centralized values.

[0008] In a possible implementation, the method for automatic identification of geographical elements in remote sensing images based on large model analysis further performs the following processing: the global confidence evaluation channel includes a global confidence evaluation formula, and the global confidence evaluation formula is: ; in, Represents the global confidence of the qth constraint group corresponding to the qth full-factor segmentation constraint group, q is a positive integer, 1≤q≤Q, Characterizes the support of the qth constraint group corresponding to the qth full-element segmentation constraint group, Characterizes the support of the mth full-element image segmentation constraint sample in the qth full-element segmentation constraint group.

[0009] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis further performs the following processing: the prompt word element includes multiple geographic element prompt words; the multiple geographic element prompt words are synonymously expanded according to the prompt word expansion large model to obtain multiple prompt word clusters; the multiple prompt word clusters and the remote sensing image data are input into an image target detection model to obtain multiple target element annotation frame sets corresponding to the multiple prompt word clusters; coupling optimization is performed based on the multiple target element annotation frame sets to generate the second constraint condition for image segmentation.

[0010] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis also performs the following processing: the point of interest element includes multiple geographical locations of points of interest; a coordinate system is constructed based on the remote sensing image data to obtain a remote sensing image coordinate system; coordinates of the multiple geographical locations of points of interest are converted based on the remote sensing image coordinate system to obtain spatial data of the point of interest of the element; CSV conversion is performed based on the spatial data of the point of interest of the element to generate the third constraint condition for image segmentation.

[0011] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis also performs the following processing: performing segmentation loss detection on the multi-level remote sensing image segmentation results according to the image segmentation loss detection model to obtain each image segmentation loss detection result, wherein the image segmentation loss detection model includes a multi-dimensional indicator for image segmentation loss detection, and the multi-dimensional indicator for image segmentation loss detection includes segmentation accuracy, over-segmentation rate and under-segmentation rate; judging whether each image segmentation loss detection result satisfies the multi-dimensional constraint of segmentation loss detection; if each image segmentation loss detection result does not satisfy the multi-dimensional constraint of segmentation loss detection, generating a segmentation loss optimization instruction; based on the segmentation loss optimization instruction, performing segmentation loss correction on the multi-level remote sensing image segmentation results according to the multi-dimensional constraint of segmentation loss detection to obtain a multi-level remote sensing image segmentation optimization result.

[0012] In a possible implementation, the method for automatic identification of geographic elements in remote sensing images based on large model analysis also performs the following processing: the remote sensing data processing rules include remote sensing data quality constraints, a predetermined remote sensing data format and remote sensing data processing factors, and the remote sensing data processing factors include radiation correction, geometric correction and image enhancement; according to the multiple remote sensing data sources, first remote sensing image data are loaded; according to the remote sensing data quality constraints, the first remote sensing image data are filtered to obtain second remote sensing image data; according to the predetermined remote sensing data format, the second remote sensing image data is format converted to obtain third remote sensing image data; the third remote sensing image data is preprocessed according to the remote sensing data processing factors to generate the remote sensing image data.

[0013] The present application also provides a remote sensing image geographic element automatic recognition system based on large model analysis, including: a remote sensing image data acquisition module, which is used to obtain remote sensing image data based on remote sensing data processing rules and multiple remote sensing data sources; a geographic element recognition instruction acquisition module, which is used to obtain geographic element recognition instructions, wherein the geographic element recognition instructions include multi-level element recognition targets, and the multi-level element recognition targets include full-volume elements, prompt word elements and point of interest elements; an image segmentation constraint tree establishment module, which is used to perform multi-level image segmentation constraint alignment on the remote sensing image data according to the multi-level element recognition targets and establish an image segmentation constraint tree; a remote sensing image segmentation result acquisition module, which is used to perform segmentation processing on the remote sensing image data according to the SAM image segmentation large model based on the image segmentation constraint tree to obtain a multi-level remote sensing image segmentation result; a remote sensing image segmentation optimization result acquisition module, which is used to optimize the segmentation loss of the multi-level remote sensing image segmentation result according to the image segmentation loss detection model to obtain a multi-level remote sensing image segmentation optimization result; a geographic element recognition result acquisition module, which is used to perform vectorized element recognition according to the multi-level remote sensing image segmentation optimization result to obtain a multi-level geographic element recognition result.

[0014] The method and system for automatic identification of geographic elements in remote sensing images based on large model analysis proposed in this application is intended to obtain remote sensing image data based on remote sensing data processing rules and multiple remote sensing data sources; obtain geographic element identification instructions; perform multi-level image segmentation constraint registration and establish an image segmentation constraint tree; perform segmentation processing on remote sensing image data based on the SAM image segmentation large model to obtain multi-level remote sensing image segmentation results; perform segmentation loss optimization to obtain multi-level remote sensing image segmentation optimization results; perform vectorized element identification to obtain multi-level geographic element identification results. The technical problems of insufficient precision, low efficiency and inaccurate image segmentation in remote sensing image multi-level geographic element identification in the prior art are solved, and efficient and accurate identification of geographic elements is achieved, achieving the technical effect of improving the efficiency and accuracy of geographic element identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0016] Figure 1 A schematic flow chart of a method for automatic identification of geographic elements in remote sensing images based on large model analysis provided in an embodiment of the present application.

[0017] Figure 2 A schematic diagram of the structure of a system for automatic identification of geographic elements in remote sensing images based on large model analysis provided in an embodiment of the present application.

[0018] Explanation of the accompanying drawings: remote sensing image data acquisition module 10, geographic element recognition instruction acquisition module 20, image segmentation constraint tree establishment module 30, remote sensing image segmentation result acquisition module 40, remote sensing image segmentation optimization result acquisition module 50, geographic element recognition result acquisition module 60. DETAILED DESCRIPTION

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0020] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0021] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0022] The present application embodiment provides a method for automatically identifying geographic elements in remote sensing images based on large model analysis, such as Figure 1 As shown, the method includes: Step S100, based on remote sensing data processing rules, remote sensing image data is obtained according to multiple remote sensing data sources.

[0023] Step S100 further includes step S110, wherein the remote sensing data processing rules include remote sensing data quality constraints, a predetermined remote sensing data format and a remote sensing data processing factor, and the remote sensing data processing factor includes radiation correction, geometric correction and image enhancement; step S120, loading the first remote sensing image data according to the multiple remote sensing data sources; step S130, filtering the first remote sensing image data according to the remote sensing data quality constraints to obtain second remote sensing image data; step S140, format converting the second remote sensing image data according to the predetermined remote sensing data format to obtain third remote sensing image data; step S150, preprocessing the third remote sensing image data according to the remote sensing data processing factor to generate the remote sensing image data.

[0024] Preferably, based on remote sensing data processing rules, multiple remote sensing data sources are processed to obtain remote sensing image data, wherein remote sensing data processing rules refer to rules for acquiring, processing and analyzing remote sensing data, including remote sensing data quality constraints, predetermined remote sensing data formats and remote sensing data processing factors. Remote sensing data quality constraints define how to evaluate and ensure the quality of remote sensing data. The quality of remote sensing data may be affected by many factors, including sensor performance, atmospheric conditions, cloud cover, severe blur, etc. The validity of the data is screened and evaluated by setting some quality standards (such as signal-to-noise ratio, resolution requirements, etc.); predetermined remote sensing data formats refer to predetermined data format requirements to ensure that remote sensing image data from different data sources can be processed and analyzed consistently. Common remote sensing data formats include GeoTIFF, HDF, NetCD F, etc., to ensure the standardized processing of the spatial information and attribute data of the data; remote sensing data processing factors refer to the operating techniques that need to be applied in the preprocessing process. Remote sensing data processing factors include radiation correction, geometric correction and image enhancement. Specifically, the radiation value of the remote sensing image may be affected by factors such as the sensor, time, and ground conditions. Radiation correction eliminates these influences by adjusting the radiation value in the image, so that the ground information reflected by the image is more accurate; remote sensing images may produce geometric distortions during the acquisition process due to factors such as the sensor angle and the curvature of the earth. Geometric correction corrects these distortions by performing geometric transformations on the image (such as plane projection) to make the geographic coordinates of the image consistent with the real ground coordinates; enhancement technology is used to improve the visual effect of the image or highlight the features of certain objects, for example, enhancing the contrast and brightness of the image or removing noise through filtering.

[0025] Preferably, the remote sensing data source includes remote sensing image data from different platforms and sensors, which may include images from satellite remote sensing platforms, for example, image data from satellites such as Landsat, Sentinel, WorldView, etc., with different resolutions, band combinations and imaging methods; images from aerial remote sensing platforms, remote sensing image data taken by aircraft or drones, usually with high spatial resolution, suitable for fine-grained regional analysis; laser radar data (LiDAR), the laser radar system obtains high-precision three-dimensional terrain data by scanning the ground with lasers; synthetic aperture radar (SAR) data, which uses radar bands to obtain ground reflection signals, can penetrate clouds and lighting conditions, and is particularly suitable for post-disaster monitoring or surface deformation monitoring; these remote sensing data sources are processed into a standardized data format through remote sensing data processing rules to obtain remote sensing image data, that is, the original remote sensing data after preliminary processing (such as radiation correction, atmospheric correction, etc.), which is used for the identification, classification and analysis of geographic elements.

[0026] Preferably, first remote sensing image data is selected and loaded from a plurality of remote sensing data sources. The first remote sensing image data may be derived from satellite images, aerial images, etc., and has not undergone any quality control or format conversion. The first remote sensing image is filtered according to remote sensing data quality constraints (for example, signal-to-noise ratio, image clarity, etc.), and parts that do not meet the quality standards (such as low-quality images, cloud interference areas, etc.) are removed to obtain second remote sensing image data that meets the quality requirements; and then, the second remote sensing image data is format converted according to a predetermined remote sensing data format, wherein different remote sensing data sources may use different file formats, that is, the second remote sensing image data is converted into a predetermined remote sensing data format. Standard data format is used to unify the data structure, ensure that the data is compatible with subsequent processing, and obtain the third remote sensing image data (meeting the quality requirements and format requirements); finally, the third remote sensing image data is preprocessed according to the remote sensing data processing factor, including the application of remote sensing data processing factors (such as radiation correction, geometric correction, image enhancement, etc.), adjusting the radiation information in the image, so that the ground features can be accurately reflected in the image, eliminating the influence of the atmosphere and sensors, and transforming the image coordinates to ensure that the image is aligned with the ground real coordinate system, eliminating geometric distortion, and highlighting important ground features or improving the image's recognizability by enhancing the image's visual performance. Through quality filtering, format conversion and preprocessing, remote sensing data from different data sources are converted into standardized, high-quality remote sensing image data to ensure the accuracy, consistency and reliability of remote sensing image data, which can be used for effective geographic feature identification and analysis.

[0027] Step S200, obtaining a geographic element recognition instruction, wherein the geographic element recognition instruction includes a multi-level element recognition target, and the multi-level element recognition target includes a full element, a prompt word element, and an interest point element.

[0028] Preferably, a geographic element recognition instruction is obtained for identifying and extracting geographic elements according to different target levels. Each level of recognition target has a specific function and content. Specifically, the geographic element recognition instruction guides how the geographic elements in the remote sensing image are identified and classified. The geographic element recognition instruction includes a multi-level element recognition target, that is, according to the needs of the application scenario, different types and levels of geographic elements can be specified for recognition. Among them, the multi-level element recognition target indicates that geographic element recognition is not a single target, but is divided into tasks at multiple levels or levels, with different recognition requirements and details. The multi-level element recognition target includes full elements, prompt word elements and point of interest elements. Full elements refer to all possible geographic element categories, covering various types and features of land objects that may appear in the entire remote sensing image. At this level, all geographic elements in the image are identified and classified, regardless of their quantity, type or importance. For example, various types of land objects such as urban buildings, farmlands, forests, and water bodies are all full elements. Full elements require the system to identify all land objects and natural features appearing in the image, and are usually used for geographic element analysis of the entire region, such as large-scale land use, environmental monitoring, etc.

[0029] Preferably, the prompt word element refers to guiding the identification of geographic elements according to the pre-set prompt words or keywords during the identification process to help focus on specific land object categories or areas. For example, the prompt word may be "bridge", "road", "river", etc. There are often multiple prompt word elements, each prompt word points to a specific category or area. If there is a specific bridge or road in a remote sensing image, these specific structures are identified based on prompt words such as "bridge" or "road", without having to care about other irrelevant areas in the image. The addition of prompt words enhances the accuracy of image analysis and avoids interference from irrelevant parts during the identification process; the point of interest element refers to multiple areas or land objects that are of special concern in a specific application scenario. The point of interest can be a specific geographical location, a specific land object (such as a historical site, a special building, a disaster-affected area, an important urban facility, etc.), or a sign of a special event (such as a fire or flood-affected area). Identifying the point of interest element can help improve the monitoring and management of specific areas and meet the needs of the application level. Through such hierarchical identification, remote sensing image analysis can be more accurate and efficient to meet the needs of different fields.

[0030] Step S300 , performing multi-level image segmentation constraint registration on the remote sensing image data according to the multi-level element recognition target, and establishing an image segmentation constraint tree.

[0031] Preferably, according to the multi-level element recognition target, the acquired remote sensing image data is subjected to multi-level image segmentation, that is, the remote sensing image data is segmented step by step in accordance with the recognition requirements of different geographic elements (different scales, different levels). Specifically, firstly, a rough overall segmentation of all elements is performed to mark different major types of objects in the image (such as forests, farmlands, water bodies, etc.). Then, based on specific targets such as prompt words and points of interest, the image is further finely segmented, focusing on specific areas or specific types of objects, such as identifying bridges, roads, buildings, etc. Through multi-level segmentation, it is ensured that geographic elements can be identified from the global to the local, from the rough to the fine. Among them, image segmentation is a commonly used technology in computer vision and remote sensing image processing. Its purpose is to divide the image into several regions with certain similar features (such as color, texture, shape, etc.), each region representing a specific geographic feature or object.

[0032] Preferably, registration is to align remote sensing images from different sources, at different times or from different sensors to the same coordinates. The purpose of image registration is to eliminate geometric deformations caused by different shooting angles, shooting times, sensor differences, etc., and to ensure consistency between images. Specifically, when segmenting remote sensing images, the images are accurately aligned so that the segmented areas can accurately correspond to the real geographic elements on the ground. Constrained registration can ensure image accuracy and consistency between different segmentation levels, especially in the case of fusion of multi-source data (such as satellite images, aerial images, etc.), to avoid recognition errors caused by differences in data sources. Some objects may have relative position relationships between different levels, or there may be spatial dependencies and constraints between recognition targets at different levels (such as full elements, prompt word elements, etc.). These constraint registrations can ensure that corresponding ground features can be correctly aligned and associated at different levels of image segmentation. By organizing the segmentation results and constraint relationships at different levels in a tree structure, an image segmentation constraint tree is established to represent the image segmentation process at different levels and their mutual constraint relationships. That is, the image segmentation constraint tree includes nodes at each segmentation level (that is, different segmentation results of the image) and the constraint relationships between them. The hierarchical structure of the tree reflects the segmentation process from the rough segmentation of all elements to the detailed segmentation of prompt word elements or interest point elements. The edges or connections in the tree represent the constraint relationships between different segmentation levels. For example, some ground objects should maintain consistent spatial positions or forms at different segmentation levels (for example, roads and buildings in urban areas should maintain consistent segmentation results at the same location).

[0033] Furthermore, step S300 also includes step S310, performing image segmentation constraint parameter parsing on the remote sensing image data according to the full amount elements to obtain a first image segmentation constraint condition; step S320, performing image segmentation constraint parameter parsing on the remote sensing image data according to the prompt word elements to obtain a second image segmentation constraint condition; step S330, performing image segmentation constraint parameter parsing on the remote sensing image data according to the point of interest elements to obtain a third image segmentation constraint condition; step S340, using the full amount elements as the first node of geographic element identification, using the prompt word elements as the second node of geographic element identification, and using the point of interest elements as the third node of geographic element identification; step S350, constructing an element identification node tree according to the first node of geographic element identification, the second node of geographic element identification, and the third node of geographic element identification; step S360, performing image segmentation constraint configuration on the element identification node tree according to the first constraint condition of image segmentation, the second constraint condition of image segmentation, and the third constraint condition of image segmentation to generate the image segmentation constraint tree.

[0034] Preferably, each geographic element (such as all elements, cue word elements and point of interest elements) corresponds to a set of image segmentation constraint parameters, which define how to segment the image, thereby ensuring that the segmentation results of different elements can accurately and finely reflect the geographic information in the image, and performing image segmentation constraint parameter analysis on remote sensing image data based on all geographic elements, including parsing the first constraint condition for image segmentation through the characteristics of all elements (such as spectral characteristics, shape characteristics, etc. of the objects), that is, how to segment according to the categories of all objects in the image (such as forests, farmlands, urban buildings, etc.), to ensure that each object can be accurately identified and distinguished in the image; performing constraint analysis on object recognition guided by specific "cue words" (such as "bridges" and "roads") to generate the second constraint condition for image segmentation, to ensure that the objects related to these cue words in the image can be accurately segmented; and generating the third constraint condition for image segmentation by performing constraint analysis on point of interest elements, to ensure that the segmentation results can highlight and accurately identify these special areas or objects.

[0035] Preferably, the identified geographic elements are assigned to different nodes and organized into a tree structure based on these nodes. Each node represents a level of geographic element recognition task. There may be some dependencies or hierarchical relationships between the nodes. Specifically, the first node for geographic element recognition is the full amount of elements, which represents the most basic full amount of geographic element recognition for remote sensing images; the second node for geographic element recognition is the prompt word element, which represents the recognition task for a specific type of land object; the third node for geographic element recognition is the point of interest element, which represents the geographic features of special concern in the image, such as certain specific buildings, urban areas or natural landscapes; the multi-level geographic element recognition process in the image is organized and managed through nodes to construct an element recognition node tree, the root node of the tree represents the full amount of element recognition, and the other child nodes represent more detailed The hierarchical relationship of the tree structure shows how the recognition tasks at different levels are interdependent. Then, according to the first constraint condition of image segmentation, the second constraint condition of image segmentation and the third constraint condition of image segmentation, the feature recognition node tree is configured with image segmentation constraints, that is, these constraints are applied to the feature recognition node tree to adjust and configure the segmentation task of each node, ensuring that the image segmentation process of each node can be carried out efficiently and accurately. Finally, an image segmentation constraint tree is generated, which not only defines how to identify geographic elements at different levels, but also stipulates how to affect the details of segmentation through constraints, ensuring that the boundaries and categories of objects are accurately identified in the segmentation of different levels, thereby ensuring that the recognition of geographic elements in remote sensing images can be carried out more efficiently and accurately.

[0036] Furthermore, step S310 also includes step S311, performing twin data retrieval according to the remote sensing image data to obtain Q remote sensing image twin samples, wherein Q is a positive integer greater than 1; step S312, collecting the full-factor image segmentation constraint historical parameters corresponding to the Q remote sensing image twin samples to obtain Q full-factor segmentation constraint groups, wherein each full-factor segmentation constraint group includes M full-factor image segmentation constraint samples corresponding to each remote sensing image twin sample, and M is a positive integer greater than 1; step S313, based on the full A local confidence evaluation channel is used to perform a global confidence evaluation on the Q full-factor segmentation constraint groups to obtain global confidences of the Q constraint groups; step S314 is used to select the Q full-factor segmentation constraint groups based on the global confidences of the Q constraint groups according to a predetermined global confidence to establish a full-factor segmentation constraint space; step S315 is used to perform a concentration value calculation on the full-factor segmentation constraint space to generate the first constraint condition for image segmentation, wherein the first constraint condition for image segmentation includes M image segmentation constraint concentration values.

[0037] Step S313 further includes that the global confidence evaluation channel includes a global confidence evaluation formula, and the global confidence evaluation formula is: ; in, Represents the global confidence of the qth constraint group corresponding to the qth full-factor segmentation constraint group, q is a positive integer, 1≤q≤Q, Characterizes the support of the qth constraint group corresponding to the qth full-element segmentation constraint group, Characterizes the support of the mth full-element image segmentation constraint sample in the qth full-element segmentation constraint group.

[0038] Preferably, twin data retrieval refers to retrieving Q twin samples (Q is a positive integer greater than 1) similar to the current remote sensing image data from historical data sets. They are historical image samples that are similar to the current remote sensing image data, and have similar land features, geographical locations or environmental backgrounds. Each remote sensing image twin sample has a corresponding historical segmentation constraint. The historical parameters of the full-factor segmentation constraints extracted from the Q remote sensing image twin samples constitute Q full-factor segmentation constraint groups. Each segmentation constraint group contains M full-factor image segmentation constraint samples (M is a positive integer greater than 1), representing a variety of possible constraints when a full-factor is segmented in a historical image. For example, different parameter setting strategies should be adopted for different extraction methods, extraction object refinement, and actual hardware. The parameters include minimum mask area, filtering threshold, and stability threshold. Among them, the minimum mask area parameter must be greater than 0, and the filtering threshold and stability threshold range between 0 and 1.

[0039] Preferably, a global confidence evaluation channel is used to perform a global confidence evaluation on the Q full-factor segmentation constraint groups, that is, the reliability and effectiveness of each constraint group are evaluated by certain standards (such as segmentation precision, accuracy, matching degree of historical data, etc.). A higher global confidence means that the constraint group may be more effective and reliable in practical applications, and then the global confidence of the Q constraint groups is obtained, which reflects the effect of each group of constraints. The global confidence evaluation channel includes a global confidence evaluation formula, which is: ; in, Represents the global confidence of the qth constraint group corresponding to the qth full-factor segmentation constraint group, q is a positive integer, 1≤q≤Q, Characterizes the support of the qth constraint group corresponding to the qth full-element segmentation constraint group, Characterizes the support of the mth full-element image segmentation constraint sample in the qth full-element segmentation constraint group.

[0040] Preferably, optimization and screening are performed based on the global confidence of each full-factor segmentation constraint group. Constraint groups with high confidence will be considered as more reliable segmentation standards, and constraint groups with lower confidence may be eliminated or adjusted. The full-factor segmentation constraint groups corresponding to the global confidence of the constraint groups greater than or equal to the predetermined global confidence are added to the full-factor segmentation constraint space to ensure the accuracy and reliability of subsequent image segmentation. In the full-factor segmentation constraint space, the concentrated value of the constraint conditions is obtained by calculation. The concentrated value usually refers to the average value and median of the constraint conditions, which represents the best constraint standard. The concentrated value calculation can balance the influence of different constraints to ensure that the constraint conditions finally selected are the most reasonable in practical applications. Finally, the first constraint condition for image segmentation is obtained, which includes M concentrated values ​​of image segmentation constraints, which is a comprehensive evaluation result of all historical data constraints. Each concentrated value of image segmentation constraints represents a segmentation rule, which provides an accurate constraint basis for image segmentation tasks.

[0041] Furthermore, step S320 also includes step S321, wherein the prompt word elements include multiple geographic element prompt words; step S322, performing synonymous expansion on the multiple geographic element prompt words according to the prompt word expansion model to obtain multiple prompt word clusters; step S323, inputting the multiple prompt word clusters and the remote sensing image data into the image target detection model to obtain multiple target element annotation frame sets corresponding to the multiple prompt word clusters; step S324, performing coupling optimization according to the multiple target element annotation frame sets to generate the second constraint condition for image segmentation.

[0042] Preferably, the prompt word refers to a keyword used to guide image analysis, and the prompt word elements include multiple geographic element prompt words, indicating different types of land feature categories, and each prompt word corresponds to a specific geographic element, such as "building", "road", "river", "construction", "cultivated land", "green space", "vegetation", etc. Then, the multiple geographic element prompt words are synonymously expanded according to the prompt word expansion model, that is, based on natural language processing (NLP) technology, synonym expansion is performed through existing prompt words. For example, based on a large-scale language model (such as BERT, etc.), each geographic element prompt word is expanded into a synonym cluster. For example, the prompt word "river" is expanded to "waterway", "stream", "water system", etc., and the prompt word "road" is expanded to "street", "lane", "highway", etc., thereby obtaining multiple prompt word clusters, each cluster containing a specific geographic element category and all its synonyms.

[0043] Preferably, multiple prompt word clusters and remote sensing image data are input into an image target detection model, and the image target detection model (such as YOLO, Faster R-CNN, etc.) automatically detects the target objects in the image (i.e., specific geographical features in the remote sensing image) based on the input prompt word clusters and remote sensing image data. Specifically, the target detection model uses the synonym information in the prompt word clusters to search for matching areas in the image and generate corresponding annotation boxes. It can find the area belonging to a certain feature category in the image, and then obtain multiple target feature annotation box sets corresponding to multiple prompt word clusters. These annotation box sets contain the location information of all target features identified in the image (i.e., their bounding boxes in the image), thereby identifying the exact location and category of each feature. The annotation boxes of image target detection may overlap, be redundant, or be misdetected. Finally, coupling optimization is performed based on multiple target feature annotation box sets, that is, multiple annotation boxes are associated, integrated, and optimized to ensure that they cover the correct feature area more accurately, such as merging redundant annotation boxes, correcting erroneous annotation boxes, or optimizing the size and position of the boxes. Finally, the second constraint condition for image segmentation is generated to ensure that the bounding boxes of target features (such as roads, buildings, rivers, etc.) are more accurate during the image segmentation process, thereby improving the accuracy of image segmentation.

[0044] Furthermore, step S330 also includes step S331, wherein the point of interest element includes multiple geographical locations of points of interest; step S332, constructing a coordinate system based on the remote sensing image data to obtain a remote sensing image coordinate system; step S333, performing coordinate conversion on the multiple geographical locations of points of interest based on the remote sensing image coordinate system to obtain element point of interest spatial data; step S334, performing CSV conversion based on the element point of interest spatial data to generate the third constraint condition for image segmentation.

[0045] Preferably, a point of interest element refers to a specific geographic location or area that a user or task pays special attention to, for example, certain specific buildings, facilities, natural landscapes, disaster areas, etc. Each point of interest element corresponds to one or more specific geographic location coordinates, indicating an area that needs to be focused on in the remote sensing image. A coordinate system is constructed based on the remote sensing image data, for example, based on a geographic coordinate system (such as WGS84, UTM, etc.) to determine the precise geographic location corresponding to each pixel in the image on the earth's surface, and the remote sensing image coordinate system obtained is the geographic coordinate system of each pixel in the image.

[0046] Preferably, coordinate transformation is performed on the geographical locations of multiple points of interest based on the remote sensing image coordinate system, and the coordinates are transformed into the coordinate system used by the remote sensing image to ensure that these points of interest correspond to the actual locations of the objects in the remote sensing image, and the spatial data of the element points of interest are obtained, that is, the spatial location of each point of interest will match the remote sensing image data and be transformed into spatial data consistent with the remote sensing image coordinate system, which includes the accurate location and spatial information of the point of interest in the image. Finally, CSV transformation is performed based on the spatial data of the element points of interest, that is, CSV transformation is performed based on the spatial data of the element points of interest, and each point of interest will be represented as a table row, including its location (coordinates) and related attributes (such as type, label, etc.) in the image, wherein CSV (comma separated values) is a common data storage format, which can convert geographic spatial data into a table form, each row represents a point of interest, and includes relevant information of the point of interest (such as latitude and longitude coordinates, object type, other attributes, etc.), and finally the third constraint condition for image segmentation is obtained, which guides the image segmentation algorithm to accurately identify and process these point of interest areas in the image, and ensure that the segmentation result accurately captures the area where the point of interest is located, avoiding mis-segmentation or missed segmentation.

[0047] Step S400, based on the image segmentation constraint tree, the remote sensing image data is segmented according to the SAM image segmentation model to obtain a multi-level remote sensing image segmentation result.

[0048] Preferably, the remote sensing image data is segmented and processed based on the image segmentation constraint tree according to the SAM image segmentation model, wherein SAM is a powerful image segmentation model based on deep learning, which usually adopts large-scale pre-training data and powerful model architecture (such as convolutional neural network CNN, Transformers, etc.) for fine image segmentation. It can achieve efficient and high-precision segmentation on different types of images, and is particularly suitable for complex remote sensing image analysis. Specifically, the remote sensing image data is segmented and processed using the SAM model and the image segmentation constraint tree, that is, the SAM model refines the segmentation hierarchy and strategy according to the constraints in the image segmentation constraint tree, accurately extracts the boundaries of different geographical elements (such as forests, urban buildings, water bodies, etc.) in the remote sensing image, and assigns corresponding category labels to each land object.

[0049] Preferably, for full-factor extraction, firstly, image segmentation threshold parameters, such as the minimum mask area, are input into the SAM image segmentation model, and then the image element segmentation reasoning function of the SAM model is used to perform segmentation prediction processing on all geographic elements within the remote sensing image; for prompt word extraction, firstly, the target element annotation box detected by the target detection model according to the prompt word is input into the SAM image segmentation model, and the target element annotation box is used as the constraint condition of the SAM model, and then the image element segmentation reasoning function of the SAM model is used to perform segmentation prediction processing on the geographic target elements detected in the remote sensing image; for interest point extraction, the CSV file with X and Y geographic coordinates of the interest annotation points is input into the SAM image segmentation model, and the coordinate points in the CSV file are used as the constraint condition of the SAM model, and then the image element segmentation reasoning function of the SAM model is used to perform segmentation prediction processing on the geographic elements annotated by the interest points in the remote sensing image.

[0050] Preferably, since the image segmentation constraint tree is multi-level, it can support the refined segmentation of the remote sensing image layer by layer, and finally obtain a multi-level remote sensing image segmentation result. Each level focuses on different types of objects (such as all objects, specific prompt word objects, interest point objects, etc.), ensuring the final image segmentation accuracy and precision. For complex remote sensing image analysis tasks, it can meet the recognition requirements of different levels and improve the efficiency and accuracy of remote sensing image processing.

[0051] Step S500, performing segmentation loss optimization on the multi-level remote sensing image segmentation result according to the image segmentation loss detection model to obtain a multi-level remote sensing image segmentation optimization result.

[0052] Preferably, the segmentation loss of the multi-level remote sensing image segmentation results is optimized by an image segmentation loss detection model to improve the quality of image segmentation and make the segmentation results more accurate and precise. Specifically, the image segmentation loss detection model is a model for evaluating the quality of remote sensing image segmentation results. The image segmentation results are usually judged based on certain preset standards. The loss detection model evaluates the accuracy and rationality of the segmentation results by detecting different errors in the image segmentation process, such as insufficient segmentation accuracy, over-segmentation and under-segmentation. The segmentation loss (such as low segmentation accuracy, over-segmentation or under-segmentation) detected by the image segmentation loss detection model is optimized and adjusted, and fine adjustments are made in the segmentation of multiple levels to ensure that the segmentation of each level can correctly and accurately reflect the characteristics of the ground objects, thereby obtaining the multi-level remote sensing image segmentation optimization results, which are more accurate and effective at multiple levels, including ensuring that the segmentation of each ground object category is more accurate, reducing classification errors, and reducing the over-segmentation rate and under-segmentation rate, reducing redundant or missed segmentation areas, improving the overall segmentation quality, and ensuring the effectiveness and accuracy in remote sensing image data.

[0053] Furthermore, step S500 also includes step S510, performing segmentation loss detection on the multi-level remote sensing image segmentation results according to the image segmentation loss detection model to obtain each image segmentation loss detection result, wherein the image segmentation loss detection model includes a multi-dimensional indicator for image segmentation loss detection, and the multi-dimensional indicator for image segmentation loss detection includes segmentation accuracy, over-segmentation rate and under-segmentation rate; step S520, judging whether each image segmentation loss detection result satisfies the multi-dimensional constraint of segmentation loss detection; step S530, if each image segmentation loss detection result does not satisfy the multi-dimensional constraint of segmentation loss detection, generating a segmentation loss optimization instruction; step S540, based on the segmentation loss optimization instruction, performing segmentation loss correction on the multi-level remote sensing image segmentation results according to the multi-dimensional constraint of segmentation loss detection to obtain a multi-level remote sensing image segmentation optimization result.

[0054] Preferably, segmentation loss detection is performed on the multi-level remote sensing image segmentation results according to an image segmentation loss detection model, the image segmentation loss detection model includes image segmentation loss detection multidimensional indicators, the image segmentation loss detection multidimensional indicators include segmentation accuracy, over-segmentation rate and under-segmentation rate, wherein the segmentation accuracy reflects the degree of matching between the segmentation result and the actual object, over-segmentation means that a ground object in the image is mistakenly divided into multiple small parts, over-segmentation will lead to additional calculations and unnecessary area divisions, under-segmentation means that the ground objects that should be separated are mistakenly merged into one area, under-segmentation will lead to information loss and inaccurate ground object classification, and each image segmentation loss detection result is obtained to describe whether each segmentation result has problems in accuracy, over-segmentation and under-segmentation; it is judged whether each image segmentation loss detection result satisfies the segmentation loss detection multidimensional constraints, wherein the segmentation loss detection multidimensional constraints are preset standards for ensuring the quality of the segmentation results, including segmentation accuracy constraints (segmentation accuracy should reach a certain threshold), over-segmentation rate constraints and under-segmentation rate constraints (over-segmentation rate and under-segmentation rate should be lower than a certain threshold).

[0055] Preferably, if the segmentation loss detection results of each image do not meet the multi-dimensional constraints of the segmentation loss detection, a segmentation loss optimization instruction is generated to make adjustments for segmentation accuracy, over-segmentation or under-segmentation, such as adjusting the parameters of the segmentation model to improve segmentation accuracy, optimizing the model, avoiding under-segmentation, etc.; finally, based on the segmentation loss optimization instruction, the multi-level remote sensing image segmentation results are corrected for segmentation loss according to the multi-dimensional constraints of the segmentation loss detection, that is, the original segmentation results are adjusted, which may include modifying the parameters or structure in the algorithm to make the segmentation results more accurate in terms of accuracy and segmentation boundaries; reducing the over-segmentation rate to avoid unnecessary division of objects in the image into multiple regions; at the same time, avoiding under-segmentation to ensure that different objects are correctly segmented, and finally obtaining a multi-level remote sensing image segmentation optimization result with higher segmentation accuracy, more reasonable region division, and lower over-segmentation and under-segmentation problems, ensuring that the recognition and segmentation of objects in the image are both accurate and reasonable.

[0056] Step S600, performing vectorized element recognition based on the multi-level remote sensing image segmentation optimization result to obtain a multi-level geographic element recognition result.

[0057] Preferably, vectorization is performed based on the multi-level remote sensing image segmentation optimization result, which means converting the raster data (i.e., an image composed of pixels) of the multi-level remote sensing image segmentation optimization result into vector data, wherein each pixel in the raster data usually corresponds to a small area in the image, and the vector data uses geometric figures such as points, lines and surfaces to represent the boundaries of objects. Specifically, converting the segmentation results (usually regions or blocks) into more expressive vector boundaries, such as polygons, line segments, etc., helps to improve the accuracy of object representation. For example, through vectorization, the boundaries of urban areas, river basins, etc. can be converted into vectors with clear boundaries and geometric features. The vectorized result is then used to perform feature recognition, that is, to identify specific geographic features in the image (such as buildings, roads, water bodies, etc.), and finally obtain multi-level geographic feature recognition results. Each level corresponds to a different level of geographic feature recognition tasks. For example, each level corresponds to a different level of geographic feature recognition tasks including all significant land object categories in the image (such as water bodies, cities, agricultural areas, etc.); the prompt word feature recognition results further refine the recognition of specific land object categories; the point of interest feature recognition results perform fine recognition based on specific areas or key land objects, thereby improving the accuracy and reliability of remote sensing image analysis.

[0058] In the above, refer to Figure 1 The automatic identification method of geographical elements in remote sensing images based on large model analysis according to an embodiment of the present invention is described in detail. Figure 2 The following describes an automatic recognition system for geographical elements in remote sensing images based on large model analysis according to an embodiment of the present invention.

[0059] The automatic identification system of geographical elements in remote sensing images based on large model analysis according to the embodiment of the present invention is used to solve the technical problems of insufficient precision, low efficiency and inaccurate image segmentation of multi-level geographical elements in remote sensing images in the prior art, realize efficient and accurate identification of geographical elements, and achieve the technical effect of improving the efficiency and accuracy of geographical element identification. The automatic identification system of geographical elements in remote sensing images based on large model analysis includes: a remote sensing image data acquisition module 10, a geographical element recognition instruction acquisition module 20, an image segmentation constraint tree establishment module 30, a remote sensing image segmentation result acquisition module 40, a remote sensing image segmentation optimization result acquisition module 50, and a geographical element recognition result acquisition module 60.

[0060] A remote sensing image data acquisition module 10 is used to obtain remote sensing image data based on remote sensing data processing rules and multiple remote sensing data sources; a geographic element recognition instruction acquisition module 20 is used to obtain geographic element recognition instructions, wherein the geographic element recognition instructions include multi-level element recognition targets, and the multi-level element recognition targets include full-volume elements, prompt word elements and interest point elements; an image segmentation constraint tree establishment module 30 is used to perform multi-level image segmentation constraint alignment on the remote sensing image data according to the multi-level element recognition targets and establish an image segmentation constraint tree; a remote sensing image segmentation result acquisition module 40 is used to perform segmentation processing on the remote sensing image data according to the SAM image segmentation large model based on the image segmentation constraint tree to obtain a multi-level remote sensing image segmentation result; a remote sensing image segmentation optimization result acquisition module 50 is used to perform segmentation loss optimization on the multi-level remote sensing image segmentation result according to an image segmentation loss detection model to obtain a multi-level remote sensing image segmentation optimization result; a geographic element recognition result acquisition module 60 is used to perform vectorized element recognition based on the multi-level remote sensing image segmentation optimization result to obtain a multi-level geographic element recognition result.

[0061] The specific configuration of the image segmentation constraint tree establishment module 30 will be described in detail below. The image segmentation constraint tree establishment module 30 further includes: performing image segmentation constraint parameter analysis on the remote sensing image data according to the full amount elements to obtain the first image segmentation constraint condition; performing image segmentation constraint parameter analysis on the remote sensing image data according to the prompt word elements to obtain the second image segmentation constraint condition; performing image segmentation constraint parameter analysis on the remote sensing image data according to the interest point elements to obtain the third image segmentation constraint condition; using the full amount elements as the first node of geographic element identification, using the prompt word elements as the second node of geographic element identification, and using the interest point elements as the third node of geographic element identification; constructing an element identification node tree according to the first node of geographic element identification, the second node of geographic element identification, and the third node of geographic element identification; performing image segmentation constraint configuration on the element identification node tree according to the first constraint condition of image segmentation, the second constraint condition of image segmentation, and the third constraint condition of image segmentation to generate the image segmentation constraint tree.

[0062] The specific configuration of the image segmentation constraint tree establishment module 30 will be described in detail below. The image segmentation constraint tree establishment module 30 further includes: performing twin data retrieval according to the remote sensing image data to obtain Q remote sensing image twin samples, wherein Q is a positive integer greater than 1; collecting the full-element image segmentation constraint history parameters corresponding to the Q remote sensing image twin samples to obtain Q full-element segmentation constraint groups, wherein each full-element segmentation constraint group includes M full-element image segmentation constraint samples corresponding to each remote sensing image twin sample, and M is a positive integer greater than 1; based on the global confidence evaluation channel, performing global confidence evaluation on the Q full-element segmentation constraint groups to obtain the global confidence of the Q constraint groups; based on the global confidence of the Q constraint groups, selecting the Q full-element segmentation constraint groups according to the predetermined global confidence, and establishing the full-element segmentation constraint space; performing centralized value calculation according to the full-element segmentation constraint space to generate the first image segmentation constraint condition, wherein the first image segmentation constraint condition includes M image segmentation constraint centralized values.

[0063] The specific configuration of the image segmentation constraint tree building module 30 will be described in detail below. The image segmentation constraint tree building module 30 further includes: the global confidence evaluation channel includes a global confidence evaluation formula, and the global confidence evaluation formula is: ; in, Represents the global confidence of the qth constraint group corresponding to the qth full-factor segmentation constraint group, q is a positive integer, 1≤q≤Q, Characterizes the support of the qth constraint group corresponding to the qth full-element segmentation constraint group, Characterizes the support of the mth full-element image segmentation constraint sample in the qth full-element segmentation constraint group.

[0064] The specific configuration of the image segmentation constraint tree establishment module 30 will be described in detail below. The image segmentation constraint tree establishment module 30 further includes: the prompt word element includes multiple geographic element prompt words, and the multiple geographic element prompt words are synonymously expanded according to the prompt word expansion model to obtain multiple prompt word clusters; the multiple prompt word clusters and the remote sensing image data are input into the image target detection model to obtain multiple target element annotation frame sets corresponding to the multiple prompt word clusters; coupling optimization is performed according to the multiple target element annotation frame sets to generate the second image segmentation constraint condition.

[0065] The specific configuration of the image segmentation constraint tree establishment module 30 will be described in detail below. The image segmentation constraint tree establishment module 30 further includes: the interest point element includes a plurality of interest point geographical locations; a coordinate system is constructed according to the remote sensing image data to obtain a remote sensing image coordinate system; coordinate conversion is performed on the plurality of interest point geographical locations based on the remote sensing image coordinate system to obtain element interest point spatial data; CSV conversion is performed according to the element interest point spatial data to generate the image segmentation third constraint condition.

[0066] The specific configuration of the remote sensing image segmentation optimization result acquisition module 50 will be described in detail below. The remote sensing image segmentation optimization result acquisition module 50 further includes: performing segmentation loss detection on the multi-level remote sensing image segmentation results according to the image segmentation loss detection model to obtain each image segmentation loss detection result, wherein the image segmentation loss detection model includes a multi-dimensional index for image segmentation loss detection, and the multi-dimensional index for image segmentation loss detection includes segmentation accuracy, over-segmentation rate, and under-segmentation rate; judging whether each image segmentation loss detection result satisfies the multi-dimensional constraint of segmentation loss detection; if each image segmentation loss detection result does not satisfy the multi-dimensional constraint of segmentation loss detection, generating a segmentation loss optimization instruction; based on the segmentation loss optimization instruction, performing segmentation loss correction on the multi-level remote sensing image segmentation results according to the multi-dimensional constraint of segmentation loss detection to obtain a multi-level remote sensing image segmentation optimization result.

[0067] The specific configuration of the remote sensing image data acquisition module 10 will be described in detail below. The remote sensing image data acquisition module 10 further includes: the remote sensing data processing rules include remote sensing data quality constraints, a predetermined remote sensing data format and remote sensing data processing factors, and the remote sensing data processing factors include radiation correction, geometric correction and image enhancement; according to the multiple remote sensing data sources, the first remote sensing image data is loaded; according to the remote sensing data quality constraints, the first remote sensing image data is filtered to obtain the second remote sensing image data; according to the predetermined remote sensing data format, the second remote sensing image data is format converted to obtain the third remote sensing image data; according to the remote sensing data processing factor, the third remote sensing image data is preprocessed to generate the remote sensing image data.

[0068] The automatic identification system of geographical elements in remote sensing images based on large model analysis provided in the embodiments of the present invention can execute the automatic identification method of geographical elements in remote sensing images based on large model analysis provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0069] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0070] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for automatically identifying geographic elements in remote sensing images based on large model analysis, characterized in that: The method comprises: Based on remote sensing data processing rules, remote sensing image data is obtained according to multiple remote sensing data sources; Obtaining a geographic element recognition instruction, wherein the geographic element recognition instruction includes a multi-level element recognition target, and the multi-level element recognition target includes a full element, a prompt word element, and an interest point element; Perform multi-level image segmentation constraint registration on the remote sensing image data according to the multi-level element recognition target, and establish an image segmentation constraint tree; Based on the image segmentation constraint tree, the remote sensing image data is segmented according to the SAM image segmentation model to obtain a multi-level remote sensing image segmentation result; Performing segmentation loss optimization on the multi-level remote sensing image segmentation result according to the image segmentation loss detection model to obtain a multi-level remote sensing image segmentation optimization result; Vectorized element recognition is performed based on the multi-level remote sensing image segmentation optimization result to obtain a multi-level geographic element recognition result.

2. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 1, characterized in that: Performing multi-level image segmentation constraint registration on the remote sensing image data according to the multi-level element recognition target, and establishing an image segmentation constraint tree, including: Performing image segmentation constraint parameter analysis on the remote sensing image data according to the total amount elements to obtain a first constraint condition for image segmentation; Perform image segmentation constraint parameter analysis on the remote sensing image data according to the prompt word elements to obtain a second image segmentation constraint condition; Performing image segmentation constraint parameter analysis on the remote sensing image data according to the point of interest elements to obtain a third image segmentation constraint condition; A first node is identified by taking the total amount element as a geographical element, a second node is identified by taking the prompt word element as a geographical element, and a third node is identified by taking the point of interest element as a geographical element; Constructing a feature identification node tree according to the geographic feature identification first node, the geographic feature identification second node and the geographic feature identification third node; The image segmentation constraint tree is generated by configuring the element recognition node tree for image segmentation constraints according to the first image segmentation constraint condition, the second image segmentation constraint condition and the third image segmentation constraint condition.

3. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 2, characterized in that: The remote sensing image data is subjected to image segmentation constraint parameter analysis according to the total amount of elements to obtain a first image segmentation constraint condition, including: Perform twin data retrieval based on the remote sensing image data to obtain Q remote sensing image twin samples, where Q is a positive integer greater than 1; Collect the full-factor image segmentation constraint historical parameters corresponding to the Q remote sensing image twin samples to obtain Q full-factor segmentation constraint groups, wherein each full-factor segmentation constraint group includes M full-factor image segmentation constraint samples corresponding to each remote sensing image twin sample, where M is a positive integer greater than 1; Based on the global confidence evaluation channel, a global confidence evaluation is performed on the Q full-factor segmentation constraint groups to obtain global confidences of the Q constraint groups; Based on the global confidences of the Q constraint groups, the Q full-factor segmentation constraint groups are optimally selected according to a predetermined global confidence to establish a full-factor segmentation constraint space; The concentrated value calculation is performed according to the full-element segmentation constraint space to generate the first image segmentation constraint condition, wherein the first image segmentation constraint condition includes M image segmentation constraint concentrated values.

4. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 3, characterized in that: The global confidence evaluation channel includes a global confidence evaluation formula, and the global confidence evaluation formula is: ； in, Represents the global confidence of the qth constraint group corresponding to the qth full-factor segmentation constraint group, q is a positive integer, 1≤q≤Q, Characterizes the support of the qth constraint group corresponding to the qth full-element segmentation constraint group, Characterizes the support of the mth full-element image segmentation constraint sample in the qth full-element segmentation constraint group.

5. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 2, characterized in that: The remote sensing image data is analyzed for image segmentation constraint parameters according to the prompt word elements to obtain a second image segmentation constraint condition, including: The prompt word element includes a plurality of geographic element prompt words; Synonymically expanding the plurality of geographic element prompt words according to the prompt word expansion model to obtain a plurality of prompt word clusters; Inputting the plurality of prompt word clusters and the remote sensing image data into an image target detection model to obtain a plurality of target element annotation frame sets corresponding to the plurality of prompt word clusters; Coupling optimization is performed according to the multiple target element annotation frame sets to generate the second constraint condition for image segmentation.

6. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 2, characterized in that: The remote sensing image data is subjected to image segmentation constraint parameter parsing according to the interest point elements to obtain a third image segmentation constraint condition, including: The point of interest elements include a plurality of geographical locations of points of interest; Constructing a coordinate system according to the remote sensing image data to obtain a remote sensing image coordinate system; Based on the remote sensing image coordinate system, coordinate transformation is performed on the geographical locations of the plurality of interest points to obtain spatial data of element interest points; The third constraint condition for image segmentation is generated by performing CSV conversion according to the spatial data of the element interest points.

7. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 1, characterized in that: The multi-level remote sensing image segmentation result is optimized according to the image segmentation loss detection model to obtain the multi-level remote sensing image segmentation optimization result, including: Performing segmentation loss detection on the multi-level remote sensing image segmentation results according to the image segmentation loss detection model to obtain segmentation loss detection results of each image, wherein the image segmentation loss detection model includes a multi-dimensional index for image segmentation loss detection, and the multi-dimensional index for image segmentation loss detection includes segmentation accuracy, over-segmentation rate, and under-segmentation rate; Determine whether the segmentation loss detection results of each image meet the multi-dimensional constraints of segmentation loss detection; If the segmentation loss detection results of each image do not satisfy the segmentation loss detection multi-dimensional constraint, generating a segmentation loss optimization instruction; Based on the segmentation loss optimization instruction, segmentation loss correction is performed on the multi-level remote sensing image segmentation result according to the segmentation loss detection multi-dimensional constraint to obtain a multi-level remote sensing image segmentation optimization result.

8. The method for automatic identification of geographical elements in remote sensing images based on large model analysis according to claim 1, characterized in that: Based on remote sensing data processing rules, remote sensing image data is obtained from multiple remote sensing data sources, including: The remote sensing data processing rules include remote sensing data quality constraints, a predetermined remote sensing data format and remote sensing data processing factors, and the remote sensing data processing factors include radiation correction, geometric correction and image enhancement; Loading first remote sensing image data according to the multiple remote sensing data sources; filtering the first remote sensing image data according to the remote sensing data quality constraint to obtain second remote sensing image data; According to the predetermined remote sensing data format, format conversion is performed on the second remote sensing image data to obtain third remote sensing image data; The third remote sensing image data is preprocessed according to the remote sensing data processing factor to generate the remote sensing image data.

9. The automatic identification system of geographical elements in remote sensing images based on large model analysis is characterized by: The system is used to implement the method for automatically identifying geographical elements in remote sensing images based on large model analysis according to any one of claims 1 to 8, and the system comprises: A remote sensing image data acquisition module is used to obtain remote sensing image data based on remote sensing data processing rules and multiple remote sensing data sources; A geographic element recognition instruction acquisition module, used to obtain a geographic element recognition instruction, wherein the geographic element recognition instruction includes a multi-level element recognition target, and the multi-level element recognition target includes a full element, a prompt word element, and an interest point element; An image segmentation constraint tree establishment module is used to perform multi-level image segmentation constraint registration on the remote sensing image data according to the multi-level element recognition target to establish an image segmentation constraint tree; A remote sensing image segmentation result acquisition module is used to perform segmentation processing on the remote sensing image data based on the image segmentation constraint tree and the SAM image segmentation large model to obtain a multi-level remote sensing image segmentation result; A remote sensing image segmentation optimization result acquisition module is used to optimize the segmentation loss of the multi-level remote sensing image segmentation result according to the image segmentation loss detection model to obtain the multi-level remote sensing image segmentation optimization result; The geographic element recognition result acquisition module is used to perform vectorized element recognition according to the multi-level remote sensing image segmentation optimization result to obtain the multi-level geographic element recognition result.

Images