Method, system and storage medium for extracting building height by multi-source data fusion

Abstract

The application discloses a method for extracting building height by fusing multi-source data, and the method comprises the following steps: acquiring laser point cloud data and 1:500 basic topographic map data of a target city; pre-processing house surface data and ground data to obtain effective house surface and ground data; classifying and thinning the laser point cloud data, obtaining the information of the height of the sky surface and the ground surface by point-surface analysis, and eliminating the height values outside the house surface or the ground surface; extracting the height values in the house surface and the height values in the ground surface by the maximum likelihood method; and calculating the height difference between the height of the sky surface and the height of the ground surface to obtain the height of the corresponding building. The application also discloses a system for extracting building height by fusing multi-source data and a storage medium. The application can quickly, automatically and low-costly acquire the building height of a city based on various existing basic geographic information data, has low requirements on data sources, and is faster, more efficient and more accurate in data processing, and can meet the existing business requirements.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geographic information, and in particular to a method and system for extracting building height by fusing multi-source data and a storage medium. BACKGROUND

[0002] Building height information plays an important role in city management, city planning, smart city construction, disaster emergency, communication, etc. At present, there are many studies on building height information extraction, and certain results have been achieved. The studies can be mainly summarized into three categories: 1. building height inversion using high-resolution satellite images; 2. obtaining building height by using aerial photogrammetry to obtain stereo pairs; and 3. field measurement by using GPS and total station.

[0003] However, when building height inversion is performed using high-resolution satellite images, specific shooting time and satellite attitude parameters need to be known, and complex formula calculation needs to be performed. When stereo pairs are obtained by using stereo satellite images or multiple high-resolution SAR image data, there is a high demand for data sources, the cost is high, data processing is complex, and the efficiency is not satisfactory. When a traditional measurement method such as field measurement by using GPS and total station is used, the precision is high, but the cost is high and the efficiency is extremely low.

[0004] Airborne LIDAR point cloud data is widely used in city measurement and other work because it can quickly obtain three-dimensional coordinates of ground objects. The data has the characteristics of high-density and high-precision three-dimensional coordinate points of ground objects and can penetrate vegetation to obtain ground coordinate points. With the urgent demand of three-dimensional modeling, city planning and other aspects for large-scale building height information, it is necessary to study a set of automatic and low-cost city building height acquisition method based on various existing basic geographic information data. SUMMARY

[0005] Embodiments of the present application provide a method for extracting building height by fusing multi-source data, which uses laser point cloud data and basic terrain data, adopts a set of improved height extraction algorithm, calculates building roof elevation and ground elevation, thereby obtains building height, and realizes automatic calculation of a large amount of data by using FME. On the one hand, the demand for rapid acquisition of large-scale building height information can be met, and on the other hand, a large amount of cost can be saved.

[0006] Embodiment one of the present application provides a method for extracting building height by fusing multi-source data, which comprises the following steps.

[0007] Obtaining laser point cloud data and 1:500 basic topographic map data of a target area; the basic topographic map data comprises a building roof data set;

[0008] The building surface dataset is cleaned of invalid data and subjected to topology processing to obtain a valid building surface dataset; the valid building surface dataset is then subjected to buffer analysis to obtain a building surface buffer; and the building surface buffer and the valid building surface dataset are then subjected to topology processing to obtain a ground dataset.

[0009] The laser point cloud data is classified to obtain a first building roof point cloud set and a first ground point cloud set; the first building roof point cloud set and the first ground point cloud set are thinned to obtain a second building roof point cloud set and a second ground point cloud set.

[0010] The effective building surface dataset and the second building roof point cloud are used to perform point-surface analysis to obtain the building roof elevation set; the ground dataset and the second ground point cloud are used to perform point-surface analysis to obtain the ground elevation set.

[0011] The maximum likelihood method is used to obtain the roof elevation value of the target building from the set of building roof elevations; the maximum likelihood method is used to obtain the target ground elevation value from the set of ground elevations.

[0012] The height of the target building is obtained by calculating the difference between the elevation of the roof of the target building and the elevation of the target ground.

[0013] Specifically, the process of cleaning invalid data and performing topological processing on the building facade dataset to obtain a valid building facade dataset includes:

[0014] By filtering out building data whose structural attributes are "shed", "damaged", "built" or empty, the attribute fields of the building's floor area dataset are used.

[0015] By performing topology processing, the building facades visible under orthographic projection are retained, i.e., the facades with higher floors, while overlapping portions of the facades below are removed, resulting in a valid building facade dataset.

[0016] Furthermore, the step of performing buffer analysis on the valid house surface dataset to obtain a house surface buffer, and then performing topological processing on the house surface buffer and the valid house surface dataset to obtain a ground dataset, specifically includes:

[0017] The effective house surface dataset is used to perform buffer analysis based on the buffer distance to obtain the house surface buffer;

[0018] The building surface buffer and the valid building surface dataset are subjected to topological processing to obtain the ground dataset.

[0019] Specifically, classifying the laser point cloud data to obtain a first building rooftop point cloud set and a first ground point cloud set; and thinning the first building rooftop point cloud set and the first ground point cloud set to obtain a second building rooftop point cloud set and a second ground point cloud set, includes:

[0020] The point cloud is classified using the PointCloudFilter converter to obtain the first building roof point cloud set and the first ground point cloud set;

[0021] The first building rooftop point cloud and the first ground point cloud are thinned using the PointCloudCombiner converter and the PointCloudCoercer converter to obtain the second building rooftop point cloud and the second ground point cloud.

[0022] Specifically, obtaining the roof elevation value of the target building from the building roof elevation set using the maximum likelihood method includes:

[0023] Using the maximum likelihood method and setting a first threshold, the elevation interval with the highest frequency of occurrence in the roof elevation set of the target building is obtained from the roof elevation set of the building, i.e. the maximum probability interval. Then, the roof elevation value of the target building is obtained by combining one of the minimum, maximum, average, median or maximum probability values.

[0024] Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

[0025] Furthermore, the step of obtaining the target ground elevation value from the ground elevation set using the maximum likelihood method specifically includes:

[0026] The maximum likelihood method is used to obtain the elevation interval with the highest frequency in the target ground elevation set from the ground elevation set by setting a second threshold. Then, the target ground elevation value is obtained by combining one of the following methods: minimum value, maximum value, average value, median or maximum probability value.

[0027] Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

[0028] Accordingly, this invention provides a system for extracting building height through multi-source data fusion, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a method for extracting building height through multi-source data fusion as described in Embodiment 1 of this invention.

[0029] Accordingly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform a method for extracting building height by multi-source data fusion as described in Embodiment 1 of the present invention.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] This invention utilizes airborne laser point cloud data and basic topographic map data widely used in the surveying and mapping industry. It employs an improved height extraction algorithm to calculate building rooftop and ground elevations, thereby obtaining building heights. Large-scale automated data calculations are achieved through FME (Fiber Optic Meta-Ray). The invention features low data source requirements; laser point cloud data and basic topographic data are regularly updated in major cities. Airborne laser point cloud data offers high density and high precision in identifying three-dimensional coordinates of ground objects and can penetrate vegetation to obtain ground coordinates, simultaneously acquiring three-dimensional coordinates of building rooftops and surrounding ground. Utilizing existing data allows for broader application support, improving data utilization efficiency and saving financial expenditure. Furthermore, the data processing is more convenient and efficient. Employing the powerful data processing capabilities of FME software, various converters are combined, parameters are designed, and an automated processing flow is built according to the designed technical route. This facilitates efficient processing of large volumes of data, meeting the need for rapid acquisition of building height information over a wide area. The algorithm for extracting rooftop and ground elevations has been optimized, employing multiple measures to improve the accuracy of the results. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the method for extracting building height by multi-source data fusion according to an embodiment of the present invention;

[0033] Figure 2 It is a commercial and residential building project in 1:500 topographic map data. Detailed Implementation

[0034] 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, and 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.

[0035] See Figure 1 This is a flowchart illustrating an embodiment of the method for extracting building height through multi-source data fusion provided by the present invention. The method includes steps S101 to S106, as detailed below:

[0036] S101, acquire laser point cloud data of the target area and its 1:500 basic topographic map data; the basic topographic map data includes building facade datasets;

[0037] Specifically, the data sources in this embodiment of the invention are airborne laser point cloud data and basic topographic map data, which are widely used in the surveying and mapping industry. Laser point cloud data and basic topographic data are updated regularly in major cities. For example, the "Guangzhou Surveying and Mapping Management Measures" stipulates that the 1:500 topographic map of the built-up area and key construction area of ​​Guangzhou City should be updated at least once a year, and the 1:500 topographic map of other construction areas should be updated at least once every two years. Aerial photography is carried out every two years, which will acquire laser point cloud data once.

[0038] In this embodiment of the invention, the test area is the historical urban area of ​​Guangzhou, covering an area of ​​20.39 square kilometers, involving the three districts of Yuexiu, Liwan and Haizhu. It uses high-precision laser point cloud data from the results of the 2020-2021 "Guangzhou High-Resolution Aerial Photogrammetry Project" as well as existing 1:500 topographic map data of Guangzhou.

[0039] S102, perform invalid data cleaning and topology processing on the building surface dataset to obtain a valid building surface dataset; perform buffer analysis on the valid building surface dataset to obtain a building surface buffer; and perform topology processing on the building surface buffer and the valid building surface dataset to obtain a ground dataset.

[0040] Specifically, the process of cleaning invalid data and performing topological processing on the building facade dataset to obtain a valid building facade dataset includes:

[0041] By filtering out building data whose structural attributes are "shed", "damaged", "built" or empty, the attribute fields of the building's floor area dataset are used.

[0042] By performing topology processing, the building facades visible under orthographic projection are retained, i.e., the facades with higher floors, while overlapping portions of the facades below are removed, resulting in a valid building facade dataset.

[0043] In a specific embodiment, since different floors in the 1:500 topographic map building surface dataset have different building surface shapes, they will be represented by the spatial superposition of multiple surfaces. Therefore, there is an overlap between building surface graphics. However, in the actual laser point cloud data, only the information of the highest point is collected at the same location of the building. Therefore, the collected building surface dataset needs to be preprocessed first, that is, through invalid data cleaning and topology processing, so that each two-dimensional coordinate point has only one unique top building surface coverage, in order to improve the accuracy of subsequent data extraction.

[0044] It should be noted that "shed", "broken", and "built" are building structure attributes in topographic map standards and specifications, while "empty" refers to a house record where the building structure is empty.

[0045] See Figure 2 It is a commercial and residential building project in 1:500 topographic map data.

[0046] exist Figure 2 The residential building project consists of 12 rooftops, 9 of which are floors 2 to 32, and 3 are the 33rd floor. Therefore, there is an overlap between the rooftop graphics. In actual laser point cloud data, only the information of the highest point is collected at the same location of the building. Therefore, it is necessary to first perform topological processing on the cleaned rooftop data to ensure that each two-dimensional coordinate point is covered by only one rooftop, so as to improve the accuracy of subsequent data extraction.

[0047] Specifically, in this embodiment of the invention, by determining the topological relationships between lines, between lines and polygons, and further between polygons, all overlapping parts of polygons with overlapping relationships are deleted using all intersection points and edges as boundaries, thus forming non-overlapping building surface data.

[0048] Furthermore, the step of performing buffer analysis on the valid house surface dataset to obtain a house surface buffer, and then performing topological processing on the house surface buffer and the valid house surface dataset to obtain a ground dataset, specifically includes:

[0049] The effective house surface dataset is used to perform buffer analysis based on the buffer distance to obtain the house surface buffer;

[0050] The building surface buffer and the valid building surface dataset are subjected to topological processing to obtain the ground dataset.

[0051] In a specific embodiment, the buffer distance is set to 1 meter, and the closed strip buffer boundary is obtained by using the convex corner arc method. After the building surface forms a buffer, it has a mutual overlapping relationship with the existing building surface. According to the topological relationship between polygons, topological processing is performed to remove the overlapping parts and obtain the ground dataset.

[0052] S103, classify the laser point cloud data to obtain a first building rooftop point cloud set and a first ground point cloud set; perform thinning processing on the first building rooftop point cloud set and the first ground point cloud set to obtain a second building rooftop point cloud set and a second ground point cloud set.

[0053] Specifically, classifying the laser point cloud data to obtain a first building rooftop point cloud set and a first ground point cloud set; and thinning the first building rooftop point cloud set and the first ground point cloud set to obtain a second building rooftop point cloud set and a second ground point cloud set, includes:

[0054] The point cloud is classified using the PointCloudFilter converter to obtain the first building roof point cloud set and the first ground point cloud set;

[0055] The first building rooftop point cloud and the first ground point cloud are thinned using the PointCloudCombiner converter and the PointCloudCoercer converter to obtain the second building rooftop point cloud and the second ground point cloud.

[0056] In this embodiment of the invention, the laser point cloud data used is the result of classification and gross error processing from the "Guangzhou High-Resolution Aerial Photogrammetry Project". The point cloud density is not less than 4 points / square meter, and there are no large-area gaps. However, the number of point clouds is large, with a total of approximately 470 million point clouds within the test area, and includes various types of ground feature data, including roads, vegetation, and ground data. Due to the large amount of data, it seriously affects the efficiency of data reading and processing. Therefore, it is necessary to classify and thin the data.

[0057] Specifically, PointCloudCombiner merges point clouds within a certain range, and then PointCloudCoercer converts multiple points into one or a few fewer points, thereby thinning the point cloud.

[0058] S104, using the effective building surface dataset and the building roof point cloud set, point-surface analysis is performed to obtain the building roof elevation set; using the ground dataset and the ground point cloud set, point-surface analysis is performed to obtain the ground elevation set;

[0059] Specifically, using the effective building surface dataset and the building roof point cloud set, a ray-mapping method is used to perform point-surface analysis, and point cloud data not within the building surface are removed; using the ground dataset and the ground point cloud set, a ray-mapping method is used to perform point-surface analysis, and point cloud data not within the ground are removed.

[0060] S105, Obtain the roof elevation value of the target building from the building roof elevation set using the maximum likelihood method; Obtain the target ground elevation value from the ground elevation set using the maximum likelihood method;

[0061] Specifically, obtaining the roof elevation value of the target building from the building roof elevation set using the maximum likelihood method includes:

[0062] Using the maximum likelihood method and setting a first threshold, the elevation interval with the highest frequency of occurrence in the roof elevation set of the target building is obtained from the roof elevation set of the building, i.e. the maximum probability interval. Then, the roof elevation value of the target building is obtained by combining one of the minimum, maximum, average, median or maximum probability values.

[0063] Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

[0064] Furthermore, the step of obtaining the target ground elevation value from the ground elevation set using the maximum likelihood method specifically includes:

[0065] The maximum likelihood method is used to obtain the elevation interval with the highest frequency in the target ground elevation set from the ground elevation set by setting a second threshold. Then, the target ground elevation value is obtained by combining one of the following methods: minimum value, maximum value, average value, median or maximum probability value.

[0066] Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

[0067] In a specific embodiment, since the roofs of buildings in the historical urban area of ​​Guangzhou are usually covered with debris such as elevator rooms, parapet walls, insulation walls, and green vegetation, which can interfere with the accurate extraction of building roof elevation information, this study has independently developed an optimized maximum likelihood method after repeated practice.

[0068] The basic idea of ​​maximum likelihood estimation is that there are many possible values ​​for the unknown parameter in the population distribution. The goal is to find an estimate that maximizes the probability of the sample occurring. This estimate is the maximum likelihood estimate.

[0069] For a discrete population X, its probability distribution is P(X = x; θ). Let (x1, x2, ..., xn) be an observation from a sample (X1, X2, ..., Xn) of this discrete population X. The probability of this observation occurring generally depends on one or more parameters, denoted by θ. This probability can be considered as a function of θ, denoted by L(θ), also known as the likelihood function of θ, i.e.:

[0070] L(θ)=P(X1=x1, X2=x2,...,Xn=xn; θ);

[0071] Finding the maximum likelihood estimate is to find an estimate of θ = (x1, x2, ..., xn) such that L(θ) in the above equation is maximized.

[0072] In a specific implementation, the maximum likelihood method is improved by setting a threshold to eliminate false maximum probability values. For example, a building's elevation is 10, but due to various reasons such as elevator shafts, the elevations of 10 points within its range are (9, 9, 10, 10, 10, 11, 14, 14, 14, 14). When using the maximum likelihood method directly, the elevation of the building's elevation is found to be 14. However, if a threshold of 2 meters is set, by controlling the threshold interval, the interval 9-11 has more point clouds than the interval 13-15. Therefore, after using threshold control, the result will be an elevation of 10. That is, the points with a higher overall probability within the threshold interval are the valid points for calculating the elevation.

[0073] The first and second thresholds need to be set according to the specific experiment to ensure higher accuracy.

[0074] Specifically, the maximum likelihood method is used with a threshold set. The elevation interval with the highest frequency of occurrence in the roof elevation set of the target building is obtained from the building roof elevation set. Based on this, the minimum, maximum, average, median and maximum probability values ​​are calculated to obtain the corresponding roof elevation values ​​for various combinations.

[0075] Among them, the formula for calculating the minimum value of a numerical attribute within the elevation interval with the most frequent occurrences of the elevation, which combines the maximum likelihood method with the minimum value method, is as follows:

[0076] Hmin = min(H1,H2,...,Hn);

[0077] Combining the maximum likelihood method with the maximum value method, the formula for calculating the maximum value of a numerical attribute within the elevation interval where the elevation occurs most frequently is as follows:

[0078] Hmax = max(H1,H2,...,Hn);

[0079] Combining the maximum likelihood method with the average value, the sum of the values ​​within the elevation interval that occurs most frequently, divided by the number of values, is calculated using the following formula:

[0080] Hmean=sum(H1,H2,...,Hn) / n;

[0081] This method combines maximum likelihood estimation with median analysis to calculate the overall standard deviation of values ​​within the elevation interval that occurs most frequently. If the number of attributes is even, the median returns the average of the two median values; if the number of attributes is odd, the median returns the value of the median.

[0082] The maximum likelihood method is combined with the maximum probability value, which is the value that appears most frequently within the elevation interval where the elevation occurs most frequently.

[0083] Specifically, the maximum likelihood method is used with a threshold set. The elevation interval with the highest frequency in the target ground elevation set is obtained from the ground elevation set. Based on this, the minimum, maximum, average, median and maximum probability values ​​are calculated to obtain the corresponding ground elevation values ​​for various combinations of methods.

[0084] The algorithm used to obtain ground elevation values ​​is the same as that used to obtain building roof elevation values. Multiple ground elevation values ​​can be obtained by combining the maximum likelihood method with one of the minimum, maximum, average, median, or maximum probability values.

[0085] In this embodiment, the average value method and the maximum probability value method are the most reliable in terms of accuracy. The specific method can be adjusted according to the original point cloud conditions.

[0086] S106, calculate the difference between the roof elevation of the target building and the ground elevation to obtain the height of the target building.

[0087] In summary, the multi-source data fusion method for extracting building height provided by this invention utilizes airborne laser point cloud data and basic topographic map data widely used in the surveying and mapping industry. It employs an improved height extraction algorithm to calculate the building's rooftop elevation and ground elevation, thereby obtaining the building's height. Furthermore, it achieves automated calculation of large-scale data using FME. The data source requirements of this invention are low; laser point cloud data and basic topographic data are regularly updated in major cities. Airborne laser point cloud data is used because it features high-density and high-precision three-dimensional coordinate points of ground objects and can penetrate vegetation to obtain ground coordinate points, simultaneously obtaining the three-dimensional coordinate points of the building's rooftop and surrounding ground. It leverages existing data results, enabling a single set of collected data to support more applications, improving data utilization efficiency and saving financial expenditure. In addition, the data processing of this invention is more convenient and faster. It utilizes the powerful data processing capabilities of FME software, combining different converters and designing parameters to build an automated processing flow according to the designed technical route. This conveniently supports efficient processing of large-scale data, meeting the need for rapid acquisition of building height information over a wide area. The algorithm for extracting rooftop and ground elevations has been optimized, employing multiple measures to improve the accuracy of the results.

[0088] This invention also provides a system for extracting building height through multi-source data fusion, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiment of an automated method for extracting building height through multi-source data fusion.

[0089] Furthermore, embodiments of the present invention provide a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the steps of the above-described embodiment of a method for extracting building height through multi-source data fusion.

[0090] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary hardware platforms, and of course, it can also be implemented entirely by hardware. Based on this understanding, all or part of the technical solution of the present invention that contributes to the background art can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0091] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for extracting building height through multi-source data fusion, characterized in that, include: Acquire laser point cloud data and its 1:500 base topographic map data for the target area; The basic topographic map data includes a dataset of building facades; The building surface dataset is cleaned of invalid data and subjected to topology processing to obtain a valid building surface dataset; the valid building surface dataset is then subjected to buffer analysis to obtain a building surface buffer; and the building surface buffer and the valid building surface dataset are then subjected to topology processing to obtain a ground dataset. The laser point cloud data is classified to obtain a first building rooftop point cloud set and a first ground point cloud set; The cloud of points on the roof of the first building and the cloud of points on the ground are thinned to obtain the cloud of points on the roof of the second building and the cloud of points on the ground. Using the effective building surface dataset and the second building roof point cloud set, point-surface analysis is performed to obtain the building roof elevation set; Ground elevation set is obtained by performing point-area analysis using the ground dataset and the second ground point cloud set; The maximum likelihood method is used to obtain the roof elevation value of the target building from the building roof elevation set; The target ground elevation value is obtained from the ground elevation set using the maximum likelihood method; The height of the target building is obtained by calculating the difference between the elevation of the roof of the target building and the elevation of the target ground.

2. The method for extracting building height by multi-source data fusion as described in claim 1, characterized in that, The process of cleaning invalid data and performing topology processing on the building facade dataset to obtain a valid building facade dataset specifically includes: By filtering out building data whose structural attributes are "shed", "damaged", "built" or empty, based on the attribute fields of the building's floor area dataset; By performing topology processing, the building facades visible under orthographic projection are retained, i.e., the facades with higher floors, while overlapping portions of the facades below are removed, resulting in a valid building facade dataset.

3. The method for extracting building height by multi-source data fusion as described in claim 1, characterized in that, The process of performing buffer analysis on the valid building surface dataset to obtain a building surface buffer, and then performing topological processing on the building surface buffer and the valid building surface dataset to obtain a ground dataset, specifically includes: The effective house surface dataset is used to perform buffer analysis based on the buffer distance to obtain the house surface buffer; The ground dataset is obtained by performing topological processing on the building surface buffer and the valid building surface dataset.

4. The method for extracting building height by multi-source data fusion as described in claim 1, characterized in that, The laser point cloud data is classified to obtain a first building rooftop point cloud set and a first ground point cloud set; The cloud of rooftop points and the cloud of ground points of the first building are thinned to obtain the cloud of vertices of the second building and the cloud of ground points, specifically including: The point cloud is classified using the PointCloudFilter converter to obtain the first building roof point cloud set and the first ground point cloud set; The first building rooftop point cloud and the first ground point cloud are thinned using the PointCloudCombiner converter and the PointCloudCoercer converter to obtain the second building rooftop point cloud and the second ground point cloud.

5. The method for extracting building height by multi-source data fusion as described in claim 1, characterized in that, The step of obtaining the roof elevation value of the target building from the building roof elevation set using the maximum likelihood method specifically includes: Using the maximum likelihood method and setting a first threshold, the elevation interval with the highest frequency of occurrence in the roof elevation set of the target building is obtained from the roof elevation set of the building, i.e. the maximum probability interval. Then, the roof elevation value of the target building is obtained by combining one of the minimum, maximum, average, median or maximum probability values. Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

6. The method for extracting building height by multi-source data fusion as described in claim 1, characterized in that, The step of obtaining the target ground elevation value from the ground elevation set using the maximum likelihood method specifically includes: The maximum likelihood method is used to obtain the elevation interval with the highest frequency of occurrence in the target ground elevation set, i.e., the maximum probability interval, from the ground elevation set by setting a second threshold. Then, the target ground elevation value is obtained by combining one of the minimum, maximum, average, median or maximum probability value. Specifically, the StatisticCalculator converter, listbuilder converter, listrangeextractor converter, listexport converter, and tester converter are combined to obtain the maximum probability interval, and then the minimum, maximum, average, median, or elevation value of the maximum probability value in the maximum probability interval is obtained through the StatisticCalculator converter.

7. A system for extracting building height through multi-source data fusion, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the multi-source data fusion method for extracting building height as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the multi-source data fusion method for extracting building height as described in any one of claims 1 to 6.

Images