Building remote sensing interpretation method and device and electronic equipment

Abstract

The invention relates to the technical field of data processing, and discloses a building remote sensing interpretation method and device and electronic equipment, and the method comprises the steps: obtaining multi-source heterogeneous geographic data in a target region, dividing the multi-source heterogeneous geographic data into a plurality of target sub-rectangular regions based on a preset reference coordinate system, generating a multi-band sample of each target sub-rectangle in parallel; performing AI model reasoning on the multi-band sample to obtain an interpretation result of each target sub-rectangular grid; the grid interpretation result comprises a graphic grid and an attribute grid; converting the grid interpretation result into vector graphs, and fusing the vector graphs located at the boundaries of the target sub-rectangles to obtain a geometrically complete vector graph; and carrying out attribute assignment on the geometrically complete vector graph based on the attribute grids, and generating an attribute vector interpretation result of the target region. According to the method, the problem of engineering bottleneck of low reasoning speed when massive high-resolution remote sensing data is processed in the prior art is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method, apparatus, and electronic device for remote sensing interpretation of buildings. Background Technology

[0002] Significant engineering performance bottlenecks arise when applying large-scale remote sensing models for large-scale, high-resolution remote sensing interpretation. Taking nationwide building interpretation as an example (building location, roof, height, and use), it is necessary to create the samples required for the large-scale remote sensing model based on 350 million Level 18 remote sensing images and other heterogeneous multi-source data, perform inference, and then resolve the vectors on the raster inference results to form the final interpretation results.

[0003] The following problems mainly arise during the interpretation and reasoning process of this massive amount of remote sensing data: The inference speed is slow. Even with the purchase of 8 A6000 machines, on a large remote sensing model with 100 million parameters in the ViT (Vision Transformer, computer vision) architecture, without performance optimization, it still takes more than 2 years for continuous inference, which is too slow and cannot meet the timeliness requirements of the business.

[0004] Therefore, there is an urgent need for a computationally efficient method for interpreting building remote sensing data. Summary of the Invention

[0005] This invention provides a method, apparatus, and electronic device for interpreting building remote sensing data, in order to solve the engineering bottleneck problem of slow inference speed faced by existing technologies when processing massive amounts of high-resolution remote sensing data.

[0006] In a first aspect, the present invention provides a method for remote sensing interpretation of buildings, the method comprising: Acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel. AI model inference is performed on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster. The interpretation results of each target sub-rectangle raster are converted into vector graphics, and the vector graphics located at the boundary of the target sub-rectangle are merged to obtain a geometrically complete vector graphic; Attributes are assigned to geometrically complete vector graphics based on attribute raster, generating attribute vector interpretation results for the target area.

[0007] This invention provides a building remote sensing interpretation method that achieves efficient and complete attribute-based vector interpretation by using a parallel sample generation mechanism under a unified coordinate system, a boundary-aware vector fusion strategy, and a synchronous attribute fusion technology designed for non-streaming scenes, while ensuring that both graphic and attribute rasters are available. This method significantly improves the processing efficiency of massive remote sensing data, shortening a task that originally required more than two years of continuous inference to be completed within two months. It provides a feasible and efficient non-streaming solution for large-scale remote sensing interpretation and effectively solves the engineering bottleneck problem of slow inference speed faced by existing technologies when processing massive high-resolution remote sensing data.

[0008] In one optional implementation, based on a preset reference coordinate system, the multi-source heterogeneous geographic data is divided into multiple target sub-rectangular regions, and multi-band samples for each target sub-rectangular region are generated in parallel, including: Multi-source heterogeneous geographic data is uniformly aligned to a preset reference coordinate system, and the multi-source heterogeneous geographic data is divided into multiple independent data files according to a preset rectangle size for storage. Each data file is uniquely identified and indexed by its row and column numbers in the preset reference coordinate system. The boundaries of the target area are determined. Based on the input image size required by the AI ​​model, the boundaries are divided into a regular grid, where each regular grid cell is a target sub-rectangle. The row and column numbers of each target sub-rectangle in the preset reference coordinate system are calculated. Based on the row and column numbers of each target sub-rectangle, target sub-rectangles that intersect with the geographical boundary of the target area are selected from all grid cells to generate the final target sub-rectangle task list; The target sub-rectangle task list is traversed in parallel. Based on the row and column number of each target sub-rectangle, the required multi-source data files are retrieved and extracted. Through splicing and synthesis, the corresponding multi-band samples are generated.

[0009] This invention provides a building remote sensing interpretation method that significantly improves the sample generation efficiency under massive data through precise row and column number mapping and rapid retrieval. Combined with task list filtering and parallel processing, it ensures high utilization of computing resources, provides a stable and high-speed sample stream for subsequent AI model inference, and realizes efficient organization and parallel reading of multi-source heterogeneous geographic data, thereby supporting the high-performance execution of large-scale remote sensing interpretation tasks from the source.

[0010] In one optional implementation, the target sub-rectangle task list is traversed in parallel, and the required multi-source data files are retrieved and extracted based on the row and column numbers of each target sub-rectangle. These files are then spliced ​​and synthesized to generate corresponding multi-band samples, including: Based on the row and column numbers of each target sub-rectangle, all relevant data files are located from each source data, and all located data files are spliced ​​together to form the source data large rectangle corresponding to each source data. Calculate the coordinate offset of the target sub-rectangle relative to each source data rectangle, and based on the offset and the size of the target sub-rectangle, cut out the corresponding target area data block from each source data rectangle; The target area data blocks from different source data are spliced ​​and fused along the band dimension to generate the final multi-band sample of the corresponding target sub-rectangle.

[0011] This invention provides a building remote sensing interpretation method that reduces I / O overhead from frequent reading of small files by batch stitching large rectangles of source data; achieves pixel-level precise positioning and cropping through a coordinate offset mechanism, ensuring the spatial accuracy of the samples; and finally achieves seamless fusion of multi-source data in the band dimension, ensuring the integrity of the input data and perfectly adapting to the requirements of subsequent AI models for multi-band input formats. Thus, it supports the stable operation of large-scale interpretation tasks in terms of both accuracy and efficiency, and realizes the efficient conversion of multi-source geographic data to standard samples.

[0012] In one optional implementation, AI model inference is performed on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result for each target sub-rectangle, including: Input the multi-band samples of each target sub-rectangle into the preset AI model to perform preset task inference and output the multi-class prediction results of each target sub-rectangle; The multi-class prediction results of each target sub-rectangle are converted into raster interpretation results of the corresponding category according to the preset category.

[0013] This invention provides a building remote sensing interpretation method that fully utilizes the generalization ability of AI models to simultaneously output multiple prediction results, such as building outline, height, and usage, during a single inference process, significantly improving interpretation efficiency. Furthermore, by generating structured raster data through category conversion, it provides a unified and standardized intermediate expression for subsequent vector transformation and attribute fusion, effectively supporting a smooth transition from pixel-level prediction to object-level vectors. This step achieves efficient and automated semantic inference for each target sub-rectangle by directly inputting standardized multi-band samples into a preset AI model.

[0014] In one optional implementation, the interpretation result of each target sub-rectangle raster is converted into a vector graphic, and the vector graphics located at the boundary of the target sub-rectangle are merged to obtain a geometrically complete vector graphic, including: Extract the graphic raster data of each target sub-rectangle, and filter it according to the preset invalid raster values ​​of building targets, deleting graphic rasters that do not contain valid building information; Perform a raster-to-vector conversion operation on all valid graphic rasters to generate an initial set of vector graphics; the initial set of vector graphics includes interior vectors that are completely inside the sub-rectangle, as well as boundary vectors that intersect the boundaries of the sub-rectangle; Identify and mark all boundary vectors, and perform vector intersection fusion operation on the boundary vectors of all target sub-rectangles to merge broken vectors that cross target sub-rectangles and belong to the same feature into a complete geometry; The internal vectors of each target sub-rectangle are merged with the fused boundary vectors to output the geometrically complete building outline vector within the target sub-rectangle area.

[0015] The present invention provides a building remote sensing interpretation method that greatly reduces computational complexity and memory consumption by fusing only boundary vectors, avoiding the risk of program crashes caused by direct full fusion; at the same time, it ensures the geometric integrity and continuity of the same ground feature across sub-rectangles, fundamentally eliminating the vector fragmentation problem caused by block processing, and providing a reliable guarantee for generating high-quality, seamless vector results.

[0016] In one optional implementation, attribute values ​​are assigned to geometrically complete vector graphics based on an attribute raster to generate a vector interpretation result of the target area with attributes, including: Convert the building attribute raster of each target sub-rectangle into an attribute vector, with each attribute vector carrying an attribute label; A spatial index is constructed for the attribute vectors. Based on the spatial index, all attribute vectors that intersect with each building outline vector are retrieved. The first intersection area of ​​the intersection between each retrieved attribute vector and the currently processed building outline vector is calculated one by one. By accumulating the first intersection area with the same attribute label, the attribute with the largest total area is determined as the final attribute vector of the corresponding building outline vector; Assign the attribute labels of the final attribute vector to the corresponding building outline vector to generate an attributed building outline vector, which is the final vector interpretation result.

[0017] This invention provides a building remote sensing interpretation method that utilizes spatial indexing and intersection area calculation to achieve automatic spatial matching of attribute values ​​and building outlines. By determining the dominant attribute based on the "area dominance" principle, it significantly improves the accuracy and rationality of attribute assignment. This step, through establishing an attribute assignment mechanism of attribute raster vectorization, spatial intersection analysis, and area-dominant decision-making, effectively solves the challenge of accurately and efficiently integrating massive vector data and multi-source attribute information.

[0018] In one optional implementation, the attribute labels of the final attribute vector are assigned to the corresponding building outline vector to generate an attributed building outline vector, which serves as the final vector interpretation result, including: Collect all attributed building outline vectors generated from the target sub-rectangles, and divide the attributed building outline vectors into internal vectors and boundary vectors; Perform adjacent or intersecting fusion operations on the boundary vectors of all target sub-rectangles to obtain a fused set of boundary vectors; The fused boundary vector set is merged with the internal vectors of all target sub-rectangles to generate the target area building vector remote sensing interpretation results.

[0019] This invention provides a method for interpreting building remote sensing data. It performs cross-block fusion on vectors with already assigned attributes, simultaneously ensuring geometric integrity (eliminating gaps through boundary fusion) and attribute consistency (directly inheriting assigned attributes). Finally, it rapidly merges seamless boundary vectors with internal vectors that do not require processing, achieving efficient and lossless convergence from block-based parallel processing to a unified global result, ensuring the direct usability of the interpretation results in large-scale display and analysis.

[0020] In one optional implementation, the fused boundary vector set is merged with the internal vectors of all target sub-rectangles to generate remote sensing interpretation results of building vectors in the target area, including: A spatial index is established for the fused boundary vector set, and all original boundary vectors intersecting with each fused boundary vector are retrieved. The second intersection area of ​​each fused boundary vector with its intersecting original boundary vector is calculated. Based on the proportion of the second intersection area to the total area of ​​the original boundary vectors, calculate the area of ​​attribute values ​​contributed by each original boundary vector; The area of ​​all contributing attribute values ​​is summed according to the attribute value category, and the attribute value with the highest summation value is determined as the final attribute of the corresponding fused boundary vector. The fused boundary vector set with completed attribute assignments is merged with the internal vectors of all target sub-rectangles, and the output is the building vector remote sensing interpretation result of the target area.

[0021] This invention provides a building remote sensing interpretation method that accurately calculates the attribute contribution value of the original vector to the fused new vector based on the area ratio, ensuring the traceability and rationality of attribute information in the geometric fusion process; by selectively inheriting the dominant attribute, it maintains the consistency of attribute semantics and avoids complex attribute conflict adjudication; ultimately, it achieves the unity of geometric seamlessness and attribute accuracy, ensuring the attribute integrity and reliability of large-scale interpretation results at the fusion boundary.

[0022] In a second aspect, the present invention provides a building remote sensing interpretation device, the device comprising: The data acquisition and multi-band sample generation module is used to acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel. The raster interpretation module is used to perform AI model inference on multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster; The raster-to-vector module is used to convert the interpretation results of each target sub-rectangle raster into vector graphics, and to merge the vector graphics located at the boundary of the target sub-rectangle to obtain geometrically complete vector graphics; The attribute vector generation module is used to assign attribute values ​​to geometrically complete vector graphics based on attribute rasters, and generate attribute vector interpretation results for the target area.

[0023] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the building remote sensing interpretation method described in the first aspect or any corresponding embodiment thereof.

[0024] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the building remote sensing interpretation method described in the first aspect or any corresponding embodiment thereof.

[0025] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the building remote sensing interpretation method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0026] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the first process of the building remote sensing interpretation method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the second process of the building remote sensing interpretation method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the third process of the building remote sensing interpretation method according to an embodiment of the present invention; Figure 4 This is a flowchart of multi-band sample generation in the building remote sensing interpretation method according to an embodiment of the present invention; Figure 5This is a flowchart illustrating the generation of multi-raster file contour vectors in the building remote sensing interpretation method according to an embodiment of the present invention; Figure 6 This is a schematic diagram of efficient generation of attribute-based vectors in the building remote sensing interpretation method according to an embodiment of the present invention; Figure 7 This is a structural block diagram of a building remote sensing interpretation device according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0030] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0031] In existing remote sensing interpretation schemes, based on the input image size of the AI ​​model, a large target area is divided into many sub-rectangles of the same size as the AI ​​model image input. After inference on these small rectangles, the raster results are fused into vector results, and then the vector results from multiple small rectangles are merged into a large vector result, which is the interpretation result of the target area. The existing scheme has the following problems: when merging multiple small rectangle vector results, the same vector (such as the roof of a building) is divided into multiple rectangles, and no literature has clearly addressed how to fuse them back together. Furthermore, some literature directly presents the unfused results; although it may be a single building, after merging, it will actually be divided into multiple vector blocks, resulting in slow inference speed. A solution for vector merging after large-scale raster inference is currently unclear.

[0032] This invention provides a method for interpreting building remote sensing data. By using a parallel sample generation mechanism under a unified coordinate system, a boundary-aware vector fusion strategy, and a synchronous attribute fusion technology designed for non-streaming scenes, it achieves an efficient and complete attribute-based vector interpretation process while simultaneously using graphic and attribute rasters, significantly improving the processing efficiency of massive amounts of remote sensing data.

[0033] According to an embodiment of the present invention, a method for remote sensing interpretation of buildings is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0034] This embodiment provides a building remote sensing interpretation method, which can be used in the aforementioned electronic equipment. Figure 1 This is a flowchart of a building remote sensing interpretation method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Obtain multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel.

[0035] Specifically, multi-source heterogeneous geographic data refers to geospatial data acquired from multiple different sources, differing in data format, structure, and physical meaning, used as input to an AI model to improve interpretation accuracy in order to complete architectural interpretation tasks. Specifically, it mainly includes the following two categories: Remote sensing image data: This refers to high-resolution (sub-meter resolution) satellite or aerial remote sensing images covering a target area (such as the entire country). It provides visual, textural, and spectral information about ground features (such as buildings).

[0036] Other heterogeneous multi-source data: This is supplementary data introduced to assist AI models in making more accurate judgments, including: POI data: that is, Point of Interest data, which contains location and type information of specific locations such as schools, shopping malls, and factories. It can help the model infer the purpose of a building (for example, if a building outline overlaps with multiple restaurant POIs, it is very likely to be a restaurant).

[0037] Probability maps / color temperature maps: These are raster data that have undergone preliminary processing or characterization of other features. For example, a probability map may be the probability of building presence derived from other models; a color temperature map may reflect the intensity and color of urban lights at night, indirectly indicating the commercial or industrial activity in a region.

[0038] The preset reference coordinate system can be the Web Mercator coordinate system (3857), GPS coordinate system (4326), etc. The purpose of the preset reference coordinate system is to provide a unified spatial benchmark for multi-source heterogeneous geographic data, thereby achieving accurate data alignment, efficient indexing, and seamless fusion, laying the foundation for subsequent parallel processing and result synthesis. Since AI models can only process a small, fixed-size image (e.g., 224×224 pixels) at a time, an entire "national map" cannot be processed. Therefore, the entire large multi-source heterogeneous geographic data is pre-divided into smaller pieces. Specifically: Dividing multi-source heterogeneous geographic data into multiple target sub-rectangular regions based on a preset reference coordinate system refers to using a regular grid to lay out a national map, dividing the entire target area (such as the whole country) into countless small, regular rectangular grids of identical size. Each small grid is a target sub-rectangular region. Simultaneously, multiple computing processes are launched to quickly retrieve the data required for the corresponding target sub-rectangular regions from different data source warehouses based on the previously established unified coordinate system and index, and then assemble them into the final multi-band sample.

[0039] Step S102: Perform AI model inference on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster.

[0040] Specifically, the raster interpretation result refers to the preliminary recognition result in a pixel grid shape output by the AI ​​model after inference on the input multi-band samples. That is, standardized multi-band samples are input into the AI ​​model for parallel inference, and the model simultaneously outputs a graphic raster describing the building outline and an attribute raster describing the building characteristics.

[0041] A graphic raster is a binary or labeled raster image in which each pixel is assigned a value. For example, in building outline recognition, a pixel with a value of 1 might indicate that it is a building, while a pixel with a value of 0 might indicate that it is not a building (i.e., background). The graphic raster defines the spatial location, shape, and outline of features (such as buildings) and forms the basis for subsequent generation of vector boundaries.

[0042] An attribute raster is a multi-valued or continuous-valued raster map where each pixel is assigned a value representing a specific attribute. For example, in building use interpretation, value 1 might represent residential, value 2 commercial, and value 3 industrial. In the building height interpretation of this embodiment, the pixel value directly represents the estimated metric height. The attribute raster assigns semantic information or quantitative attributes to each location in the above image (especially locations identified as buildings). The raster interpretation result is a preliminary answer, expressed in pixel form, understood by the AI ​​model, clearly separating the geometric information (graphic raster) and semantic information (attribute raster) of the features, laying a solid foundation for the next step of converting this pixel-level information into accurate, usable vector polygons and attribute tables.

[0043] Step S103: Convert the interpretation result of each target sub-rectangle raster into a vector graphic, and merge the vector graphics located at the boundary of the target sub-rectangle to obtain a geometrically complete vector graphic.

[0044] Specifically, vector graphics are a data format that represents geographic features using geometric shapes (such as points, lines, and polygons) defined by mathematical formulas. For a building, its vector graphic is typically a polygon, composed of a series of ordered vertices connected by geographic coordinates. The specific steps involved in its formation are as follows: The process involves identifying all connected pixel regions with the same values ​​(such as pixel blocks representing a single building) within the raster (e.g., a graphic raster) of each target sub-rectangle. Using a specific algorithm, the boundaries of these pixel regions are extracted and converted into polygon vectors composed of a series of geographic coordinate points. Each pixel cluster identified as a building in the raster is transformed into an independent initial polygon vector graphic, providing the foundation for subsequent geometric blending and attribute assignment.

[0045] The initial vector graphics formed after vectorizing each target sub-rectangle are classified, and all boundary vectors intersecting the sub-rectangle boundaries are identified and marked. Simultaneously, internal vectors completely located within the sub-rectangles are confirmed. A vector fusion operation is performed specifically on all marked boundary vectors, merging fragmented polygons spanning different sub-rectangles but belonging to the same actual feature into a single, complete geometric entity. The seamless boundary vector formed after fusion is then merged with the internal vectors in each sub-rectangle that do not require fusion, thereby outputting a geometrically complete building outline vector graphic covering the entire target area. This significantly reduces computational complexity and memory consumption, achieving efficient conversion from raster to vector.

[0046] Step S104: Assign attribute values ​​to the geometrically complete vector graphics based on the attribute raster to generate the attribute vector interpretation result of the target area.

[0047] Specifically, raster data representing building attributes (such as usage and height) is converted into vector surfaces with clear attribute labels. Spatially queryable attribute source data is established. By constructing a spatial index, the intersection area of ​​each building outline vector with all attribute vector surfaces is quickly located, and the area of ​​each intersection area is accurately calculated. The intersection areas of the same attribute within the same building outline are summarized, and the attribute with the largest total area is determined as the final attribute of the building using the "area dominance" principle. The determined attribute labels are permanently associated with the corresponding building outline vectors, ultimately forming attribute vector interpretation results that combine complete geometric form and accurate semantic information.

[0048] Based on spatial location relationships, attribute raster information is accurately assigned to geometrically complete vector graphics, and attribute matching is achieved through spatial indexing and the area dominance principle. This step effectively solves the accuracy problem of vector attribute fusion, ultimately generating a complete interpretation result that combines geometric accuracy and semantic information.

[0049] For example, for a large area across the country, the data is first divided into districts and counties, with each district or county serving as an inference unit; after inference is completed, these are then merged into a complete region. The core innovation lies in the efficient generation of multi-band samples, the efficient vectorization of inference results, and the fusion of inference results.

[0050] The building remote sensing interpretation method provided in this embodiment achieves efficient and complete attribute-based vector interpretation by using a parallel sample generation mechanism under a unified coordinate system, a boundary-aware vector fusion strategy, and a synchronous attribute fusion technology designed for non-streaming scenes, while ensuring that both graphic raster and attribute raster are available. This method significantly improves the processing efficiency of massive remote sensing data, shortening a task that originally required more than two years of continuous inference to be completed within two months, and providing a feasible and efficient non-streaming solution for large-scale remote sensing interpretation.

[0051] This embodiment provides a building remote sensing interpretation method, which can be used in the aforementioned electronic equipment. Figure 2 This is a flowchart of a building remote sensing interpretation method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel.

[0052] Specifically, there are several geographic data sources, each stored in blocks of rectangles of varying sizes. These data sources need to be merged and aligned based on geographic location to quickly form new rectangles of multi-band target size. This ensures that even with streaming generation, there are sufficient samples for AI model inference on GPU cards, enabling rapid inference without waiting.

[0053] For example, in the national building use interpretation task, it is necessary to prepare remote sensing data input and POI (Place of Intetest) distance color temperature map data input. Both of these data inputs are rectangular rasters, but the rasters are different sizes. Therefore, it is necessary to generate multi-band samples. The flowchart for generating multi-band samples is as follows. Figure 4 As shown, step S201 above includes: Step S2011: Align the multi-source heterogeneous geographic data to a preset reference coordinate system, and divide the multi-source heterogeneous geographic data into multiple independent data files according to the preset rectangle size for storage. Each data file is uniquely identified and indexed by its row and column numbers in the preset reference coordinate system.

[0054] Specifically, select a reference coordinate system: coordinate origin, projected coordinate system selection, and resolution (pixel size). The selection of the reference coordinate system should be compatible with the coordinate characteristics of various source data and target output data, such as the Web Mercator coordinate system (3857) and the GPS coordinate system (4326).

[0055] Based on the selected region coordinate system, the files from each data source are aligned to form rectangular regions of appropriate size within that coordinate system. Each file's name indicates its row and column number within the selected coordinate system and rectangular size (facilitating subsequent fast indexing). Specifically, when the number of files is exceptionally large (e.g., hundreds of millions), they are stored in two-level directories, which can be based on either row or column numbers. The row number spacing is determined by the input image size; here, it is assumed that all images are of equal size, with no overlap or omissions.

[0056] Step S2012: Determine the boundaries of the target area. Based on the input image size required by the preset AI model, divide the boundaries into a regular grid, where each regular grid cell is a target sub-rectangle. Calculate the row and column numbers of each target sub-rectangle in the preset reference coordinate system.

[0057] Specifically, this step generates a list of target sub-rectangles of all target sizes based on the overall target area and reference coordinate system, and according to the input image size required by the AI ​​model, to facilitate parallel sample generation.

[0058] Determine the grid extent: First, obtain the circumscribed rectangle of the target area, i.e., its four boundaries (westernmost, easternmost, northernmost, and southernmost coordinates). This circumscribed rectangle defines the maximum spatial extent covered by the regular grid.

[0059] Calculate the row and column range: Based on the input image size required by the AI ​​model (i.e., the size of the target sub-rectangle), combined with the coordinates and resolution of the data source origin (usually the top left corner of the coordinate system), the calculation is performed using predefined mathematical relationships: Row number = ((source origin y-coordinate - current y-value) / target sub-rectangle height); Column number = integer((current x value - x coordinate of data source origin) / width of target sub-rectangle); Calculate the row and column numbers corresponding to the coordinates of the top left and bottom right corners of the rectangle that the target region is bounded by, thereby determining the range of row and column numbers for all potential grid cells covering the region.

[0060] Step S2013: Based on the row and column numbers of each target sub-rectangle, select the target sub-rectangles that intersect with the geographical boundary of the target area from all grid cells to generate the final target sub-rectangle task list.

[0061] Specifically, the initial grid is generated by traversing the range of row and column numbers calculated in the previous step and generating the number (e.g., <row number, column number>) of each grid cell (i.e., target sub-rectangle) within that range, thus forming an initial and complete list of task grids.

[0062] Filtering valid tasks: For each grid cell number, determine the actual geographic rectangle range and its spatial relationship with the precise geographic boundary of the target area (which may be irregular, such as national or provincial borders). Only retain grid cells whose geographic range is entirely within the target area or intersects with the target area boundary; filter out grid cells completely outside the target area.

[0063] Step S2014: Parallel traverse the target sub-rectangle task list, retrieve and extract the required multi-source data files based on the row and column numbers of each target sub-rectangle, and generate corresponding multi-band samples through splicing and synthesis.

[0064] In an optional implementation, step S2014 includes: Step a1: Based on the row and column numbers of each target sub-rectangle, locate all relevant data files from each source data, and then stitch all the located data files together to form the source data large rectangle corresponding to each source data.

[0065] For each target sub-rectangle, based on its known row and column numbers, calculate the range of source data files that need to be accessed for each type of source data (such as remote sensing imagery or POI map). Specifically, calculate the row and column numbers of the source data at the top left and bottom right corners of the target sub-rectangle, and combine the row and column number ranges corresponding to these two row and column numbers to obtain the file number to be retrieved for each data source.

[0066] Based on the calculated list of file numbers, the corresponding individual files are retrieved from each source data set. Within each source data set, these adjacent small files are then concatenated in memory into a temporary, larger rectangle of source data with a wider coverage area (by concatenation). This step transforms numerous small file I / O operations into continuous matrix operations, preparing for subsequent precise pruning.

[0067] Step a2: Calculate the coordinate offset of the target sub-rectangle relative to each source data rectangle. Based on the offset and the size of the target sub-rectangle, cut out the corresponding target area data block from each source data rectangle.

[0068] For each large rectangle in the source data, calculate the pixel offset (x, y) of the top left corner of the target sub-rectangle relative to the top left corner of the large rectangle.

[0069] Based on the pixel offset (x, y) and the size of the target sub-rectangle, the target rectangular region can be quickly selected from the large rectangles found in the source data.

[0070] Specifically, after finding the pixel offset, the new pixel is found using the following calculation formula: new_matrix = old_matrix[x:x+L, y:y+L]; Where new_matrix represents the new pixel, old_matrix represents the original pixel, and L represents the pixel offset (i.e., length).

[0071] Based on this offset and the known dimensions (L, W) of the target sub-rectangle, data blocks that perfectly correspond to the target sub-rectangle region are quickly and accurately cropped from each source data rectangle using efficient array slicing operations.

[0072] Step a3 involves stitching and fusing target area data blocks from different source data along the band dimension to generate the final multi-band sample of the corresponding target sub-rectangle.

[0073] Pre-cropped data blocks from different sources (each block may already be multi-band, such as 3 bands in RGB imagery) are treated as different band groups and stitched and aligned along the band dimension. Ultimately, these data blocks are fused into a unified multi-channel two-dimensional matrix with more bands (e.g., 6 or more), which represents the multi-band samples required for the target sub-rectangle and can be directly input into the AI ​​model for inference.

[0074] Step S202: Perform AI model inference on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster.

[0075] Specifically, step S202 includes: Step S2021: Input the multi-band samples of each target sub-rectangle into the preset AI model to perform preset task inference and output the multi-class prediction results of each target sub-rectangle.

[0076] Specifically, each target sub-rectangle multi-band sample generated in parallel is used as input data and fed in batches into a pre-set remote sensing interpretation model, i.e., a pre-set AI model (such as a remote sensing large model based on the ViT architecture). Based on the knowledge obtained through its pre-training, this model performs forward inference on the input samples and simultaneously completes multiple pre-set tasks (such as building outline recognition, usage classification, height estimation, etc.).

[0077] The model outputs multi-class prediction results for each target sub-rectangle, represented as multiple two-dimensional arrays, where each pixel position contains the predicted score or regression value for different categories.

[0078] Step S2022: Convert the multi-class prediction results of each target sub-rectangle into the corresponding category raster interpretation results according to the preset category.

[0079] Specifically, the various prediction results output by the above model are analyzed and reconstructed according to the preset interpretation categories (e.g., background, building outline, residential, commercial, industrial, height value, etc.).

[0080] A separate raster file is generated for each type of prediction result (e.g., building outline, building use, building height). During this process, for classification tasks, operations such as argmax are used to convert the predicted pixel scores into specific class labels; for regression tasks (e.g., height), the estimated continuous values ​​are directly output. Finally, a set of structured raster interpretation results, strictly aligned with the spatial location of the input samples, is generated for each target sub-rectangle, including but not limited to: Graphical raster: A binary or labeled image that clearly indicates the location and outline of a building.

[0081] Attribute Raster: A thematic layer that stores attribute information such as building use and height.

[0082] Step S203: Convert the interpretation result of each target sub-rectangle raster into vector graphics, and merge the vector graphics located at the boundaries of the target sub-rectangles to obtain geometrically complete vector graphics. For details, please refer to [link to details]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0083] Step S204: Assign attribute values ​​to the geometrically complete vector graphics based on the attribute raster to generate the attribute vector interpretation result for the target area. For details, please refer to [link to details]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0084] The building remote sensing interpretation method provided in this embodiment makes full use of the generalization ability of AI models, and outputs multiple prediction results such as building outline, height, and use in a single inference process, which significantly improves the interpretation efficiency; and generates structured raster data by category conversion.

[0085] This embodiment provides a building remote sensing interpretation method, which can be used in the aforementioned electronic equipment. Figure 3 This is a flowchart of a building remote sensing interpretation method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps: Step S301: Acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, generating multi-band samples for each target sub-rectangular area in parallel. For details, please refer to [link to details]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0086] Step S302: Perform AI model inference on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result for each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster. For details, please refer to... Figure 2 Step S202 of the illustrated embodiment will not be described again here.

[0087] Step S303: Convert the interpretation result of each target sub-rectangle raster into a vector graphic, and merge the vector graphics located at the boundary of the target sub-rectangle to obtain a geometrically complete vector graphic.

[0088] Specifically, suppose there are several rectangular grids. Within each rectangular grid area, multiple internal vectors can be formed by aggregating pixels with the same value. The same surface object (such as a building roof) may be scattered across vector graphics converted from multiple sub-rectangular areas. How can we quickly and memory-efficiently merge the vector results from multiple rectangles?

[0089] This embodiment divides vectors into internal rectangular vectors and rectangular boundary vectors by observing the characteristics of the data. Only on the rectangular boundary vectors will the above-mentioned problem of the same surface object being segmented in multiple vector graphics exist. It is only necessary to judge the vectors on the rectangular boundary and perform boundary intersection fusion.

[0090] The flowchart for generating multi-raster file outline vectors is as follows: Figure 5 As shown, step S303 above includes: Step S3031: Extract the graphic raster data of each target sub-rectangle, and filter it according to the preset invalid raster values ​​of building targets, deleting graphic rasters that do not contain valid building information.

[0091] Specifically, the graphical raster data (such as building outline raster) of each target sub-rectangle is first extracted. Based on the preset invalid raster values ​​for building targets (for example, a value of 0 represents a non-building area), the raster is quickly scanned and filtered. Graphical rasters with all invalid raster values ​​(i.e., containing no valid building information) are identified, skipped, or deleted directly, thereby avoiding subsequent calculations on meaningless data and significantly improving processing efficiency.

[0092] For example, extract the graphic raster containing the roof, and first filter out the raster image without the roof (i.e., graphic raster with all raster values ​​of 0).

[0093] Step S3032: Perform a raster-to-vector conversion operation on all valid graphic rasters to generate an initial vector graphic set; the initial vector graphic set includes internal vectors that are completely inside the sub-rectangle, and boundary vectors that intersect with the boundary of the sub-rectangle.

[0094] Specifically, for all valid graphic rasters retained after filtering, a raster-to-vector conversion operation is performed. This operation aggregates adjacent pixels with the same value in the raster to form independent vector polygons, which together constitute the initial vector graphic set.

[0095] Simultaneously with generation, the system initially categorizes each vector polygon into two types based on its positional relationship with the boundary of the current target sub-rectangle: Internal vector: A vector that is completely located inside the sub-rectangle.

[0096] Boundary vector: The vector that intersects any boundary of the sub-rectangle.

[0097] Step S3033: Identify and mark all boundary vectors, and perform a vector intersection and fusion operation on the boundary vectors of all target sub-rectangles to merge broken vectors that cross target sub-rectangles and belong to the same feature into a complete geometry.

[0098] Specifically, all target sub-rectangles are traversed, and each boundary vector is identified and marked. Then, a vector intersection and fusion operation (such as union_all) is performed specifically on the set of boundary vectors of all sub-rectangles. The core purpose of this operation is to merge the fragmented boundary vectors that are divided into different sub-rectangles due to area division, but actually belong to the same feature (such as a building), into a single, complete geometry.

[0099] Step S3034: Merge all the internal vectors of each target sub-rectangle with the fused boundary vectors to output the geometrically complete building outline vector within the target sub-rectangle area.

[0100] Specifically, the obtained, already merged set of boundary vectors is directly merged with the internal vectors of all sub-rectangles. Since internal vectors do not have cross-region issues, this merging operation requires no complex spatial calculations and is extremely fast. The final output is a geometrically complete building outline vector file covering the entire target area, in which every real-world feature is represented with its complete geometric form.

[0101] It should be noted that if the target area contains a large number of sub-rectangles, or if the physical device performing the conversion and merging has limited memory, the entire target area can be divided into several medium-sized rectangles, each containing multiple target sub-rectangles. First, perform steps S3032 to S3034 on the target sub-rectangles within the medium-sized rectangles. Then, perform steps S3033 and S3034 on each medium-sized rectangle, completing the final merging within the available memory. As can be seen, the above process enables fast and memory-efficient vector result generation, supporting the final vector result generation.

[0102] Step S304: Assign attribute values ​​to the geometrically complete vector graphics based on the attribute raster to generate the attribute vector interpretation result of the target area.

[0103] Specifically, given several vector graphic (e.g., rooftop) geographic raster images and several geographic raster images with vector attributes (purposes), how can we quickly and memory-efficiently generate a unified attributed vector? Only if this fusion generation is fast enough can the AI ​​model's interpretation results be converted into the final attributed vector result.

[0104] For non-stream inference scenarios, vector graphic rasters and vector attribute rasters are generated simultaneously. In this case, the vector image rasters and vector attribute rasters on specific sub-rectangular regions can be parsed and fused separately first, then the boundary vectors of each sub-rectangle can be fused, and finally the vector results interpreted by each rectangle can be fused as a whole to obtain the attribute vector interpretation result of the target region.

[0105] A schematic diagram of efficient generation of attribute vectors is shown below. Figure 6 As shown, the process includes two stages. The first stage involves vectorizing and merging the graphic and attribute grids of each target rectangle sub-rectangle; the second stage involves merging the vector results of the sub-rectangles within the target area.

[0106] Step S304 above includes: Step S3041: Convert the building attribute raster of each target sub-rectangle into an attribute vector, with each attribute vector carrying an attribute label.

[0107] Step S3042: Construct a spatial index for the attribute vectors, retrieve all attribute vectors that intersect with each building outline vector based on the spatial index, and calculate the first intersection area of ​​the intersection between each retrieved attribute vector and the currently processed building outline vector.

[0108] Specifically, steps S3041 and S3042 above belong to the first stage, and steps S3041 and S3042 are described uniformly as follows: Step S3041 pertains to the data preparation stage for attribute fusion. Its core operation is converting raster-format attribute information into vector-format spatial objects. Specifically, the system reads the building attribute raster of each target sub-rectangle (e.g., a raster map where each pixel value represents residential, commercial, or industrial), and uses a vectorization algorithm to aggregate pixels with the same attribute values ​​and spatially adjacent pixels, forming continuous, uniform polygon vectors. Each such generated vector polygon carries a specific attribute label to record the attribute category it represents.

[0109] Within a single target sub-rectangular region, after vectorizing both the geometric raster and the geometric attribute raster, adjacent rasters with the same attributes will be converted into attribute vectors with the same value. The process of merging multiple graphic vectors (roof) and multiple attribute vectors (such as purpose) within the same target sub-rectangular region is as follows: 1. Establish spatial indexes for attribute vectors to facilitate quick calculation of overlap between graphic vectors and attribute vectors. The aforementioned spatial index refers to establishing a spatial query structure similar to a "book catalog" for all attribute vector polygons, which allows for the quick selection of a few attribute vectors that may intersect with the building outline vector based on its position, without having to traverse all the data.

[0110] 2. Traverse each building outline vector, and find all attribute vectors that overlap (intersect) with it based on the spatial index. That is, for each building outline vector, use this spatial index to quickly retrieve all attribute vectors that have a spatial overlap (intersection) relationship with it. For each retrieved attribute vector, accurately calculate the area of ​​the intersection between it and the current building outline vector, i.e., the first intersection area, which specifically includes: (1) Based on the first intersection area, obtain the area of ​​each overlapping attribute vector on the graphic vector.

[0111] (2) Calculate the area of ​​intersection of the same attributes to obtain the area of ​​each intersecting attribute value on the vector.

[0112] 3. Traverse each building outline vector, calculate whether it intersects with the target sub-region, and add an attribute to the graphic vector to annotate the calculation result.

[0113] Step S3043: By accumulating the first intersection area with the same attribute label, the attribute with the largest total area is determined as the final attribute vector of the corresponding building outline vector.

[0114] Specifically, for a given building outline vector, all intersecting attribute vectors, their first intersection areas, and corresponding attribute labels are calculated. All first intersection areas are grouped and summed according to their attribute labels. For example, the total area of ​​the commercial attribute is obtained by summing the intersection areas of all vectors labeled "commercial"; similarly, the total area of ​​attributes such as residential and industrial is calculated. These summed total areas are compared, and a decision is made based on the principle of area dominance. That is, the attribute label with the largest summed area is officially determined as the final attribute of the building outline vector.

[0115] Step S3044: Assign the attribute labels of the final attribute vector to the corresponding building outline vector to generate an attributed building outline vector, which is the final vector interpretation result.

[0116] Specifically, the determined final attribute label is written as a new field (such as building use) and permanently associated with the attribute table of the corresponding building outline vector.

[0117] After assigning attribute values ​​to all building outline vectors, these vectors are no longer just geometric shapes, but have become attributed building outline vectors that simultaneously contain precise geometric forms (outlines) and key semantic information (attributes). This set is the final vector interpretation result, which can be directly used for map visualization, spatial querying, and various geographic analysis applications.

[0118] In an optional implementation, step S3044 includes: Step b1: Collect all the attributed building outline vectors generated by the target sub-rectangles, and divide the attributed building outline vectors into internal vectors and boundary vectors.

[0119] Step b2: Perform adjacent or intersecting merging operations on the boundary vectors of all target sub-rectangles to obtain a merging set of boundary vectors.

[0120] Step b3: The fused boundary vector set is merged with the internal vectors of all target sub-rectangles to generate the target area building vector remote sensing interpretation results.

[0121] In one optional implementation, step b3 above includes: Step b31: Establish a spatial index for the fused boundary vector set, retrieve all original boundary vectors that intersect with each fused boundary vector, and calculate the second intersection area of ​​each fused boundary vector with its intersecting original boundary vector.

[0122] Step b32: Based on the proportion of the second intersection area to the total area of ​​the original boundary vectors, calculate the area of ​​attribute values ​​contributed by each original boundary vector.

[0123] Step b33: Accumulate the area of ​​all contributing attribute values ​​according to the attribute value category, and determine the attribute value with the highest accumulated value as the final attribute of the corresponding fused boundary vector.

[0124] Step b34: Merge the fused boundary vector set with the completed attribute assignments and the internal vectors of all target sub-rectangles, and output the building vector remote sensing interpretation results of the target area.

[0125] In summary, steps S3043 and S3044 belong to stage two, the purpose of which is to determine whether to perform adjacent merging on the boundary vectors on the target sub-rectangle boundary, and to directly merge the vectors within the target sub-rectangle. The specific boundary vector attribute merging logic is as follows: 1. Create a spatial index for the boundary vectors to facilitate quick indexing of the attribute values ​​of the intersecting vectors before fusion to the merged boundary vectors.

[0126] 2. Perform vector fusion on boundary vectors with adjacent / intersecting fusion. 3. The vector result after traversal and fusion: (1) Based on the spatial index constructed for the boundary vector, find the set of vectors that intersect with the merged vector. (2) Calculate the intersection area of ​​each intersecting vector. Based on the intersection area and the attribute value area of ​​the intersecting vector itself, obtain the attribute value area corresponding to the intersection part of each intersecting vector. (3) Accumulate the attribute value areas corresponding to the intersection areas of each intersecting vector to obtain the attribute area of ​​the merged vector.

[0127] After completing the vector fusion of the sub-rectangle boundaries of the target region, the attribute vectors within the sub-rectangles of the target region are directly merged to obtain the final parsed fusion result vector.

[0128] The building remote sensing interpretation method provided in this embodiment utilizes spatial indexing and intersection area calculation to achieve automatic spatial matching of attribute values ​​and building outlines; it determines the dominant attribute by the "area dominance" principle, which significantly improves the accuracy and rationality of attribute assignment; and finally directly outputs vector results with attributes, which supports the efficient generation of large-scale vector interpretation results while ensuring accuracy.

[0129] This embodiment also provides a building remote sensing interpretation device, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0130] This embodiment provides a building remote sensing interpretation device, such as Figure 7 As shown, it includes: The data acquisition and multi-band sample generation module 701 is used to acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel.

[0131] The raster interpretation module 702 is used to perform AI model inference on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster.

[0132] The raster-to-vector module 703 is used to convert the interpretation result of each target sub-rectangle raster into a vector graphic, and to merge the vector graphics located at the boundary of the target sub-rectangle to obtain a geometrically complete vector graphic.

[0133] The attribute vector generation module 704 is used to assign attribute values ​​to geometrically complete vector graphics based on attribute rasters, and generate attribute vector interpretation results for the target area.

[0134] In some alternative implementations, the data acquisition and multi-band sample generation module 701 includes: The data preprocessing unit is used to uniformly align multi-source heterogeneous geographic data to a preset reference coordinate system, and divide the multi-source heterogeneous geographic data into multiple independent data files according to a preset rectangle size for storage. Each data file is uniquely identified and indexed by its row and column numbers in the preset reference coordinate system.

[0135] The range determination and target sub-rectangle division unit is used to determine the four boundaries of the target area. Based on the input image size required by the AI ​​model, the four boundaries are divided into regular grids, where each regular grid unit is a target sub-rectangle, and the row and column numbers of each target sub-rectangle in the preset reference coordinate system are calculated.

[0136] The task list generation unit is used to filter out the target sub-rectangles that intersect with the geographic boundary of the target area from all grid cells based on the row and column numbers of each target sub-rectangle, and generate the final target sub-rectangle task list.

[0137] The multi-band sample generation unit is used to traverse the target sub-rectangle task list in parallel, retrieve and extract the required multi-source data files based on the row and column numbers of each target sub-rectangle, and generate the corresponding multi-band samples through splicing and synthesis.

[0138] In some optional implementations, the multi-band sample generation unit includes: The data source file splicing sub-unit is used to locate all relevant data files from each source data based on the row and column numbers of each target sub-rectangle, and then splice all the located data files to form the source data large rectangle corresponding to each source data.

[0139] The target region data block generation sub-unit is used to calculate the coordinate offset of the target sub-rectangle relative to each source data large rectangle. Based on the offset and the size of the target sub-rectangle, the corresponding target region data block is cut out from each source data large rectangle.

[0140] The multi-band sample generation subunit is used to stitch and fuse target area data blocks from different source data along the band dimension to generate the final multi-band sample of the corresponding target sub-rectangle.

[0141] In some alternative implementations, the raster interpretation module 702 includes: The inference unit is used to input multi-band samples of each target sub-rectangle into a preset AI model to perform preset task inference and output multi-class prediction results for each target sub-rectangle.

[0142] The raster interpretation unit is used to convert the multi-class prediction results of each target sub-rectangle into the corresponding category raster interpretation results according to the preset category.

[0143] In some alternative implementations, the raster-to-vector module 703 includes: The invalid raster filtering unit is used to extract the graphic raster data of each target sub-rectangle and filter it according to the preset invalid raster value of the building target, deleting graphic rasters that do not contain valid building information.

[0144] The raster-to-vector unit is used to perform raster-to-vector conversion operations on all valid graphic rasters, generating an initial vector graphic set; the initial vector graphic set includes internal vectors that are completely inside the sub-rectangle, and boundary vectors that intersect the boundary of the sub-rectangle. The boundary vector fusion unit is used to identify and mark all boundary vectors, and perform vector intersection fusion operation on the boundary vectors of all target sub-rectangles, merging broken vectors that cross target sub-rectangles and belong to the same feature into a complete geometry.

[0145] The merging unit is used to merge all the internal vectors of each target sub-rectangle with the merged boundary vector, and output the geometrically complete building outline vector within the target sub-rectangle area.

[0146] In some alternative implementations, the attribute vector generation module 704 includes: The attribute vectorization unit is used to convert the building attribute raster of each target sub-rectangle into an attribute vector, and each attribute vector carries an attribute label.

[0147] The spatial overlap analysis unit is used to construct a spatial index for the attribute vectors, retrieve all attribute vectors that intersect with each building outline vector based on the spatial index, and calculate the first intersection area of ​​the intersection between each retrieved attribute vector and the currently processed building outline vector.

[0148] The dominant attribute determination unit is used to determine the attribute with the largest total area by accumulating the first intersection area of ​​the same attribute labels, and then determining it as the final attribute vector of the corresponding building outline vector.

[0149] The attribute assignment and output unit is used to assign the attribute labels of the final attribute vector to the corresponding building outline vector, generating an attributed building outline vector as the final vector interpretation result.

[0150] In some optional implementations, the attribute assignment and output unit includes: Vector partitioning sub-units are used to aggregate the attributed building outline vectors generated by all target sub-rectangles and divide the attributed building outline vectors into internal vectors and boundary vectors.

[0151] The fusion sub-unit is used to perform adjacent or intersecting fusion operations on the boundary vectors of all target sub-rectangles to obtain a fused set of boundary vectors.

[0152] The merged sub-unit is used to merge the fused boundary vector set with the internal vectors of all target sub-rectangles to generate remote sensing interpretation results of building vectors in the target area.

[0153] In some optional implementations, the fused boundary vector set is merged with the interior vectors of all target sub-rectangles to generate remote sensing interpretation results of building vectors in the target area, including: A spatial index is established for the fused boundary vector set, and all original boundary vectors intersecting with each fused boundary vector are retrieved. The second intersection area of ​​each fused boundary vector with its intersecting original boundary vector is calculated. Based on the proportion of the second intersection area to the total area of ​​the original boundary vectors, calculate the area of ​​attribute values ​​contributed by each original boundary vector; The area of ​​all contributing attribute values ​​is summed according to the attribute value category, and the attribute value with the highest summation value is determined as the final attribute of the corresponding fused boundary vector. The fused boundary vector set with completed attribute assignments is merged with the internal vectors of all target sub-rectangles, and the output is the building vector remote sensing interpretation result of the target area.

[0154] The building remote sensing interpretation device provided in this embodiment of the invention can execute the building remote sensing interpretation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0155] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0156] The following is a detailed reference. Figure 8 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 801, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 802 or a program loaded from memory 808 into random access memory (RAM) 803. The RAM 803 also stores various programs and data required for the operation of the electronic device. The processor 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0157] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 808 including, for example, magnetic tapes, hard disks, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0158] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 809, or installed from a memory 808, or installed from a ROM 802. When the computer program is executed by the processor 801, it performs the functions defined in the building remote sensing interpretation method of the embodiments of the present invention.

[0159] Figure 8The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0160] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the building remote sensing interpretation method shown in the above embodiments is implemented.

[0161] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0162] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for remote sensing interpretation of buildings, characterized in that, The method includes: Acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel. AI model inference is performed on the multi-band samples of each target sub-rectangle to obtain the raster interpretation result of each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster. The interpretation results of each target sub-rectangle raster are converted into vector graphics, and the vector graphics located at the boundary of the target sub-rectangle are merged to obtain a geometrically complete vector graphic; The attribute raster is used to assign attribute values ​​to the geometrically complete vector graphic, generating the target area with attribute vector interpretation results.

2. The building remote sensing interpretation method according to claim 1, characterized in that, Based on a preset reference coordinate system, the multi-source heterogeneous geographic data is divided into multiple target sub-rectangular regions, and multi-band samples for each target sub-rectangle are generated in parallel, including: The multi-source heterogeneous geographic data is uniformly aligned to a preset reference coordinate system, and the multi-source heterogeneous geographic data is divided into multiple independent data files according to a preset rectangle size for storage. Each data file is uniquely identified and indexed by its row and column numbers in the preset reference coordinate system. The boundaries of the target area are determined. Based on the input image size required by the AI ​​model, the boundaries are divided into a regular grid, where each regular grid cell is a target sub-rectangle. The row and column numbers of each target sub-rectangle in the preset reference coordinate system are calculated. Based on the row and column numbers of each target sub-rectangle, target sub-rectangles that intersect with the geographical boundary of the target area are selected from all grid cells to generate the final target sub-rectangle task list; The target sub-rectangle task list is traversed in parallel. Based on the row and column number of each target sub-rectangle, the required multi-source data files are retrieved and extracted. Through splicing and synthesis, the corresponding multi-band samples are generated.

3. The building remote sensing interpretation method according to claim 2, characterized in that, The parallel traversal of the target sub-rectangle task list, based on the row and column numbers of each target sub-rectangle, retrieves and extracts the required multi-source data files, and generates corresponding multi-band samples through splicing and synthesis, including: Based on the row and column numbers of each target sub-rectangle, all relevant data files are located from each source data, and all located data files are spliced ​​together to form the source data large rectangle corresponding to each source data. Calculate the coordinate offset of the target sub-rectangle relative to each source data rectangle, and based on the offset and the size of the target sub-rectangle, cut out the corresponding target area data block from each source data rectangle; The target region data blocks from different source data are spliced ​​and fused along the band dimension to generate the final multi-band sample of the corresponding target sub-rectangle.

4. The building remote sensing interpretation method according to claim 1, characterized in that, AI model inference is performed on multi-band samples of each target sub-rectangle to obtain the raster interpretation results for each target sub-rectangle, including: Input the multi-band samples of each target sub-rectangle into the preset AI model to perform preset task inference and output the multi-class prediction results of each target sub-rectangle; The multi-class prediction results of each target sub-rectangle are converted into corresponding category raster interpretation results according to the preset categories.

5. The building remote sensing interpretation method according to claim 1, characterized in that, The process of converting the interpretation results of each target sub-rectangle raster into vector graphics and fusing the vector graphics located at the boundaries of the target sub-rectangles to obtain geometrically complete vector graphics includes: Extract the graphic raster data of each target sub-rectangle, and filter it according to the preset invalid raster values ​​of building targets, deleting graphic rasters that do not contain valid building information; Perform a raster-to-vector conversion operation on all valid graphic rasters to generate an initial vector graphic set; the initial vector graphic set includes internal vectors that are completely inside the sub-rectangle, and boundary vectors that intersect the boundary of the sub-rectangle; Identify and label all the boundary vectors, and perform a vector intersection fusion operation on the boundary vectors of all target sub-rectangles to merge the broken vectors that cross the target sub-rectangles and belong to the same feature into a complete geometry; The internal vectors of each target sub-rectangle are merged with the fused boundary vectors to output the geometrically complete building outline vector within the target sub-rectangle area.

6. The building remote sensing interpretation method according to claim 1, characterized in that, Attributes are assigned to the geometrically complete vector graphic based on the attribute raster, generating a vector interpretation result of the target area with attributes, including: Convert the building attribute raster of each target sub-rectangle into an attribute vector, with each attribute vector carrying an attribute label; A spatial index is constructed for the attribute vectors. Based on the spatial index, all attribute vectors that intersect with each building outline vector are retrieved, and the first intersection area of ​​the intersection between each retrieved attribute vector and the currently processed building outline vector is calculated one by one. By accumulating the first intersection area with the same attribute label, the attribute with the largest total area is determined as the final attribute vector of the corresponding building outline vector; Assign the attribute labels of the final attribute vector to the corresponding building outline vector to generate an attributed building outline vector, which is the final vector interpretation result.

7. The building remote sensing interpretation method according to claim 1, characterized in that, Assigning attribute labels to the corresponding building outline vectors to generate attributed building outline vectors, which serve as the final vector interpretation result, including: Collect all attributed building outline vectors generated from the target sub-rectangles, and divide the attributed building outline vectors into internal vectors and boundary vectors; Perform adjacent or intersecting fusion operations on the boundary vectors of all target sub-rectangles to obtain a fused set of boundary vectors; The fused boundary vector set is merged with the internal vectors of all target sub-rectangles to generate the target area building vector remote sensing interpretation result.

8. The building remote sensing interpretation method according to claim 7, characterized in that, The fused boundary vector set is merged with the internal vectors of all target sub-rectangles to generate remote sensing interpretation results of building vectors in the target area, including: A spatial index is established for the fused boundary vector set, and all original boundary vectors intersecting with each fused boundary vector are retrieved. The second intersection area of ​​each fused boundary vector with its intersecting original boundary vector is calculated. Based on the proportion of the second intersection area to the total area of ​​the original boundary vectors, calculate the area of ​​attribute values ​​contributed by each original boundary vector; The area of ​​all contributing attribute values ​​is summed according to the attribute value category, and the attribute value with the highest summation value is determined as the final attribute of the corresponding fused boundary vector. The fused boundary vector set with completed attribute assignments is merged with the internal vectors of all target sub-rectangles, and the output is the building vector remote sensing interpretation result of the target area.

9. A building remote sensing interpretation device, characterized in that, The device includes: The data acquisition and multi-band sample generation module is used to acquire multi-source heterogeneous geographic data within the target area, and based on a preset reference coordinate system, divide the multi-source heterogeneous geographic data into multiple target sub-rectangular areas, and generate multi-band samples for each target sub-rectangular area in parallel. The raster interpretation module is used to perform AI model inference on multi-band samples of each target sub-rectangle to obtain the raster interpretation result for each target sub-rectangle; the raster interpretation result includes graphic raster and attribute raster. The raster-to-vector module is used to convert the interpretation results of each target sub-rectangle raster into vector graphics, and to merge the vector graphics located at the boundary of the target sub-rectangle to obtain geometrically complete vector graphics; The attribute vector generation module is used to assign attribute values ​​to the geometrically complete vector graphics based on the attribute raster, and generate attribute vector interpretation results for the target area.

10. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the building remote sensing interpretation method according to any one of claims 1 to 8 by executing the computer instructions.

Images