Single scene high resolution SAR image high building and height extraction method

Abstract

This invention provides a method for extracting tall buildings and their heights from a single-view high-resolution SAR image. Specifically, the method involves preprocessing the SAR image; extracting salient features from the SAR image, including bright spots and bright lines; generating facade regularity components, facade bright line components, and secondary scattering components based on semantic information and salient features; formulating production system rules based on the spatial relationships between different components to generate tall building hypotheses; removing unacceptable tall building hypotheses to obtain the final tall building extraction result; calculating the range length of each tall building in the SAR image based on the final extraction result; and converting the range length of the tall buildings into their actual heights based on the imaging geometry in the SAR image. This method does not require a large amount of training data, only a single-view SAR image and its incident angle information, to quickly extract complex and diverse tall buildings and their heights.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a method for extracting the height of tall buildings from a single high-resolution SAR image. Background Technology

[0002] Synthetic Aperture Radar (SAR) is an active microwave remote sensing method that plays a crucial role in Earth observation due to its all-weather, all-day capability. In recent years, SAR sensors have developed rapidly, resulting in a surge of high-resolution SAR images. While providing more detailed scene information, this also places higher demands on information extraction methods. High-resolution SAR images typically refer to SAR images with a resolution at the meter or decimeter level. High-rise buildings refer to residential buildings with ten or more stories, or buildings with a height greater than 28 meters, and other high-rise civil buildings with a height greater than 24 meters. Compared to large or low-rise buildings, high-rise buildings exhibit more complex features in high-resolution SAR images. Firstly, the increased resolution causes high-rise buildings to no longer appear as a single entity in SAR images, but rather as a combination of discrete features such as bright spots and bright lines, making individual building extraction difficult. Secondly, the diverse structures of high-rise buildings result in inconsistent representations in SAR images, making it impossible to describe them using fixed templates.

[0003] Existing methods for extracting tall buildings from single-scene high-resolution SAR images are mostly designed for specific building types and are not suitable for most scenes with diverse and densely distributed buildings. The method described in Chen J, Zhang B, Wang C, et al. Building detection from urban high-resolution SAR image based on facade regularities[C] / / Proceedings of EUSAR 2016:11th European Conference on Synthetic ApertureRadar.2016:1-4 first extracts candidate bright spots, then combines these bright spots into bright lines according to their directions, and finally extracts individual buildings based on these bright lines. However, this method is only effective for tall buildings whose facades exhibit uniformly distributed bright spots in both the range and azimuth directions. In the same year, Tang K, Liu B, and Zou B. High-rise building detection in dense urban area based on high resolution SAR images[C] / / 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).2016:1568-1571 first detected the bright lines of the facade, then detected the secondary scattered bright lines at its far end to obtain the building facade, and finally obtained the high-rise building extraction result through facade texture analysis. Liu Kang, Balz T, and Liao Mingsheng. Extracting building height from high-resolution SAR images using backscattering characteristics[J]. Journal of Wuhan University (Information Science Edition), 2012, 37(7):806-809. Under the condition of known building azimuth, incident angle and related medium scattering characteristics, the building height is obtained by inverting the intensity of the secondary scattered bright lines of the building through the geometrical optics (GO) approximation model.In 2022, Recla M, Schmitt M. Deep-learning-based single-image height reconstruction from very-high-resolution SAR intensity data[J].ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 183:496-509, a training dataset was constructed by associating SAR image pixels with height, subtracting DEM and DSM data from experimental scenes to obtain building heights, and using a CNN network to extract building heights from SAR amplitude images.

[0004] In summary, existing methods for extracting the height of tall buildings from single-view SAR images fail to fully utilize the scattering characteristics of tall buildings in SAR images. They are only effective for a few facade types and cannot be applied to situations where tall buildings have diverse facade forms in real-world scenarios, resulting in poor adaptability and practicality. Furthermore, existing methods for extracting the height of tall buildings from single-view SAR images all require a large amount of pre-information, such as building azimuth and incident angle, or require a large amount of data for network training, making it impossible to achieve rapid extraction of tall building heights. Summary of the Invention

[0005] Based on the above, the purpose of this invention is to provide a method for extracting tall buildings and their heights from a single high-resolution SAR image. Specifically, this method is a component-based extraction method that combines the characteristics of the tall buildings themselves and their various manifestations in high-resolution SAR images. This method can quickly extract complex and diverse tall buildings and their heights without requiring a large amount of training data, only a single SAR image and its incident angle information.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, embodiments of this application provide a method for extracting tall buildings and their heights from a single high-resolution SAR image, comprising the following steps:

[0008] The SAR image is preprocessed;

[0009] The salient features are extracted from the SAR image, including bright spots and bright lines;

[0010] Based on semantic information and the aforementioned salient features, generate a facade regularity component, a facade bright line component, and a secondary scattering component;

[0011] Based on the spatial relationships between different components, production system rules are formulated to generate high-rise building assumptions;

[0012] Remove the high-rise building assumptions that do not meet the requirements to obtain the final high-rise building extraction results;

[0013] Based on the final high-rise building extraction results, the range length of each high-rise building in the SAR image is calculated;

[0014] Based on the imaging geometry in the SAR image, the distance-direction length of the high-rise building is converted into the actual building height.

[0015] In one alternative approach, the preprocessing of the SAR image specifically includes amplitude correction and speckle filtering, employing a nonlocal speckle filtering method.

[0016] In one alternative approach, the extraction of salient features from the SAR image, including bright spots and bright lines, specifically involves:

[0017] A modified spot filter is used for salient feature extraction. The filtered value is:

[0018] S a (p(x,y))=p(x,y)-min(R) step ), step = 1, 2, 3…

[0019] R step =max(p(x+i,y+j)),max(|i|,|j|)=step and s a (i,j)=1

[0020] Among them, s a For filters with different directions, a = 0, 45, 90, 135 represents the filter direction, p(x,y) is the center point, x is the range position, and y is the azimuth position. step represents the Chebyshev distance from the point to the center point p(x,y);

[0021] The obtained filtering results are then integrated:

[0022] F1(x,y)=max(S0(x,y),S 45 (x,y),S 135 (x,y))

[0023] F2(x,y)=S 90 (x,y)

[0024] By selecting an appropriate threshold and binarizing F1 and F2, the extracted bright spots and bright lines can be obtained.

[0025]

[0026] In one alternative approach, generating the facade regularity component, facade bright line component, and secondary scattering component based on semantic information and the salient features specifically involves:

[0027] Generate the facade pattern component based on the semantic information: take the distance to the center point for each orientation of each connected domain in F1;

[0028] The system iteratively detects each point, checking whether the bright spot still exists within a certain range (x:x+t1,y) from the point (x,y) to the far end. If it does, the distance between the two points is calculated as the floor interval, and compared with the existing floor interval to see if it matches. If it matches, the point is added to the group, and the system continues to detect to the far end based on this point until the bright spot is not detected twice in a row or the interval does not match the group.

[0029] Set the minimum number of bright spots per group to four; otherwise, do not record them.

[0030] Finally, the groups that are adjacent in orientation and have the same floor spacing are integrated to form the facade pattern component;

[0031] Generate the facade highlighting component based on the semantic information: calculate the distance length of each connected component in F2, which is greater than a set value t. f The connected component detection is the facade bright line component.

[0032] The secondary scattering component is generated based on the semantic information: the mean range width and amplitude of each connected component in F1 are calculated, and the range width is greater than a set value t. b And the average amplitude is greater than the set value A t The connected component detection is the secondary scattering component.

[0033] In one alternative approach, the production system rules are formulated based on the spatial relationships between different components to generate the high-rise building hypothesis. Specifically, the production system rules are:

[0034] The azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade components can be combined to form a high-rise building;

[0035] The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The multiple facade regularity components and the distance to the direction is less than the set value D b The aforementioned facade patterns and secondary scattering components can be combined to form a high-rise building;

[0036] The azimuth distance is less than the set value D f The coverage rate is greater than the set value C.f The plurality of the aforementioned facade regularity components and facade bright line components, and the distance to the direction less than the set value D b The aforementioned facade patterns, facade bright lines, and secondary scattering components can be combined to form a high-rise building;

[0037] The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The aforementioned facade patterns and facade bright line components can be combined to form a high-rise building;

[0038] The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f Multiple facade bright line components and distances less than the set value D b The aforementioned facade bright lines and secondary scattering components can be combined to form a high-rise building;

[0039] The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The multiple facade lighting components can be combined to form a high-rise building.

[0040] The directional distance refers to the distance between the rightmost side of the facade pattern component or facade bright line component and the leftmost side of the first right facade pattern component or facade bright line component.

[0041] The coverage rate refers to the larger value of the ratio of the length of the overlapping part of the facade regular component, the facade bright line component and the adjacent component to the length of the two components in the upward direction;

[0042] Distance in the direction of distance refers to the closest distance in the same direction of the facade pattern component and the facade bright line component, or the distance between the midpoint of the distance direction of the secondary scattering component and the far end of the facade pattern component and the facade bright line component.

[0043] In one alternative approach, removing the high-rise building assumptions that do not meet the requirements to obtain the final high-rise building extraction result specifically involves: the same component may belong to different building assumptions, only the high-rise building assumption containing the largest number of components is retained to obtain the most complete high-rise building extraction result, and building assumptions with a building area or floor height less than a set value are removed.

[0044] In one alternative approach, the step of calculating the range length of each high-rise building in the SAR image based on the final high-rise building extraction result specifically involves: for each final high-rise building extraction result, calculating the range length corresponding to each azimuth direction of the entire building, including the facade regularity component, the facade bright line component, and the secondary scattering component, in the SAR image area occupied by the building as a whole;

[0045] For a high-rise building with the aforementioned secondary scattering component, the distance from the midpoint of the connected domain of the secondary scattering component is taken as the bottom of the building.

[0046] When the extraction results are incomplete, the longest range length in the same final high-rise building extraction result is taken as the range length H of that high-rise building in the SAR image. r .

[0047] In one alternative approach, converting the distance-oriented length of the high-rise building into its actual height based on the imaging geometry of the SAR image specifically involves:

[0048] H = H r / cosθ

[0049] Among them, H r H is the distance length of the high-rise building, H is the actual building height of the high-rise building, and θ is the incident angle of the SAR image.

[0050] Secondly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements the steps of any of the above methods.

[0051] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the program to implement the steps of any of the above methods.

[0052] The present invention discloses the following technical effects:

[0053] (1) The inherent structural features of high-rise buildings are described by components (local features) in SAR images. By combining semantic information, each building component is extracted, which breaks through the limitations of the fixed template method.

[0054] (2) Production rules are used to describe the spatial topological relationship between components, which conforms to human visual cognition. The method is transparent and the results are highly interpretable. It can adapt to the complex and diverse forms of high-rise buildings in SAR images and realize the extraction of various types of high-rise buildings.

[0055] (3) Based on the results of high-rise building extraction, no large amount of training data or other scene information is required. Only the incident angle information of SAR image can be used to quickly extract the height of high-rise buildings in a single scene SAR image. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the content of the embodiments of the present invention and these drawings without creative effort.

[0057] Figure 1 This is a flowchart of a method for extracting the height of tall buildings in a single high-resolution SAR image, provided by an embodiment of the present invention.

[0058] Figure 2 This is a schematic diagram of spot filters in different directions provided in an embodiment of the present invention;

[0059] Figure 3 This is a schematic diagram of the generation system rules for high-rise buildings provided in an embodiment of the present invention;

[0060] Figure 4 This is a schematic diagram of the spatial relationship between different components provided in an embodiment of the present invention;

[0061] Figure 5 This is a schematic diagram of distance-to-length extraction for high-rise buildings provided in an embodiment of the present invention;

[0062] Figure 6 This is a schematic diagram of the computer device structure provided in an embodiment of the present invention;

[0063] Figure 7 This is a schematic diagram of a computer-readable storage medium structure provided in an embodiment of the present invention. Detailed Implementation

[0064] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0065] In the description of this invention, unless otherwise explicitly specified and limited, the terms "connected," "linked," and "fixed" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0066] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0067] In the description of this embodiment, the terms "upper," "lower," "left," and "right," etc., refer to the orientation or positional relationship shown in the accompanying drawings. They are used solely for ease of description and simplification of operation, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. In the description of the present invention, unless otherwise stated, "a plurality of" means two or more. Furthermore, the terms "first" and "second" are merely used for descriptive distinction and have no special meaning.

[0068] This application provides a method for extracting the height of tall buildings from a single high-resolution SAR image, such as... Figure 1 As shown, it includes the following steps:

[0069] S1. Preprocess the SAR image.

[0070] Specifically, the preprocessing includes amplitude correction and speckle filtering. Considering the complexity of urban scenes, this application employs a nonlocal speckle filtering method to better preserve architectural details while removing speckles.

[0071] S2. Extract salient features from SAR images, including bright spots and bright lines.

[0072] Specifically, in order to meet the requirement of simultaneously extracting bright spots of different sizes and bright lines of different widths and directions in the scene, such as Figure 2As shown, an improved spot filter is used for salient feature extraction. The gray background represents the points involved in the calculation. The filtered value of the center point p(x,y) is:

[0073] S a (p(x,y))=p(x,y)-min(R) step ), step = 1, 2, 3…

[0074] R step =max(p(x+i,y+j)),max(|i|,|j|)=step and s a (i,j)=1

[0075] Among them, s a For filters with different directions, a = 0, 45, 90, 135 represents the filter direction, p(x,y) is the center point, x is the range position, and y is the azimuth position. step represents the Chebyshev distance from this point to the center point p(x,y);

[0076] To better extract components, the obtained filtering results are integrated:

[0077] F1(x,y)=max(S0(x,y),S 45 (x,y),S 135 (x,y))

[0078] F2(x,y)=S 90 (x,y)

[0079] By selecting an appropriate threshold and binarizing F1 and F2, the extraction results of bright spots and bright lines can be obtained.

[0080]

[0081] By improving the spot filter, this application can extract bright spots of different sizes and bright lines of different widths from SAR images, meeting the requirements for extracting building components.

[0082] S3. Generate facade regularity components, facade bright line components, and secondary scattering components based on semantic information and salient features.

[0083] Interpretatively, taking into account the characteristics of high-rise buildings themselves and their various manifestations in high-resolution SAR images, this application selects a combination of facade regularity components, facade bright line components, and secondary scattering components to represent high-rise buildings.

[0084] Air conditioning units, window sills, or external protruding structures of high-rise buildings will appear as bright spots or bright lines in SAR images. These structures usually repeat on every floor or every two floors, forming bright spot and bright line modules that appear regularly along the distance direction in SAR images, which are defined as facade regular components.

[0085] The energy returned from the facade of a high-rise building parallel to the SAR incident angle is superimposed to form bright lines distributed along the distance direction, which are defined as facade bright line components; when the resolution is insufficient, bright spots that are regularly distributed on each floor will also be detected as facade bright line components.

[0086] The dihedral structure formed by the building's base and the ground will produce secondary or multiple scattering under SAR illumination, resulting in a thicker bright line at the building's base, which is defined as a secondary scattering component.

[0087] Specifically, based on semantic information, a facade pattern component is generated: for each directional direction of each connected component in F1, the distance to the center point is taken; for each point (x, y), iteratively detect whether there are still bright spots within a certain range (x: x+t1, y) from the far end; if so, calculate the distance between the two points as the floor interval and compare it with the existing floor interval; if it matches, add the point to the group and continue to probe the far end based on the point until no bright spots are detected twice in a row or the interval does not match the group; set the minimum number of bright spots in a group to four, otherwise do not record them; finally, integrate the groups that are adjacent in the directional direction and have the same floor interval to form the facade pattern component;

[0088] Generate a surface highlighting component based on semantic information: calculate the distance length of each connected component in F2, which is greater than the set value t. f The connected domain detection is for the facade bright line component.

[0089] Generate a secondary scattering component based on semantic information: calculate the mean of the range width and amplitude of each connected component in F1, where the range width is greater than a set value t. b And the average amplitude is greater than the set value A t The connected component detection is a secondary scattering component.

[0090] This application analyzes and extracts the scattering features of the inherent structure of high-rise buildings in SAR images, enabling the extraction of components of different types of high-rise buildings, and further realizing the extraction of complex and diverse high-rise buildings and their heights.

[0091] S4. Based on the spatial relationships between different components, formulate production system rules to generate high-rise building assumptions.

[0092] Specifically, a production system is a type of expert system that uses a set of rules to transform a problem from an initial state to a target state. The basic rules of the production system used in this application are as follows: Figure 3The rules are defined as follows:

[0093] P1: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade components can be combined to form a high-rise building;

[0094] P2: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade regularity components and distances less than the set value D b The facade patterns and secondary scattering components can be combined to form high-rise buildings;

[0095] P3: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade regularity components and facade bright line components, as well as distances less than the set value D. b Facade patterns, facade bright lines, and secondary scattering components can be combined to form high-rise buildings;

[0096] P4: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f The facade patterns and facade bright line components can be combined to form a high-rise building;

[0097] P5: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade bright line components and distances less than the set value D b The facade's bright lines and secondary scattering components can be combined to form a high-rise building;

[0098] P6: Azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade lighting components can be combined to form a high-rise building.

[0099] like Figure 4 As shown, the directional distance refers to the distance between the rightmost side of the facade pattern component or facade bright line component and the leftmost side of the first facade pattern component or facade bright line component on the right.

[0100] Coverage rate refers to the larger of the following ratios of the length of the overlapping portion between a facade regular component, a facade bright line component, and an adjacent component, up to the length of the two components;

[0101] Distance in the direction of distance refers to the closest distance in the same direction of the facade pattern component and the facade bright line component, or the distance between the midpoint of the distance in the direction of the secondary scattering component and the far end of the facade pattern component and the facade bright line component.

[0102] The production rules are formulated in accordance with human cognitive processes and are easy to understand. Through the use of production rules, the spatial topological information between components of high-rise buildings can be fully utilized. The flexibility of production rules allows this method to extract information from various high-rise building types.

[0103] S5. Remove the assumptions about high-rise buildings that do not meet the requirements and obtain the final high-rise building extraction results.

[0104] Specifically, in step S4, the same component may belong to different building assumptions. Only the high-rise building assumption containing the maximum number of components is retained to obtain the most complete high-rise building extraction results. Building assumptions with a building area of ​​less than 200 square meters or a floor height of less than 24 meters are removed. At this point, the final high-rise building extraction results can be obtained.

[0105] S6. Based on the final high-rise building extraction results, calculate the range length of each high-rise building in the SAR image.

[0106] Specifically, such as Figure 5 As shown, for each final high-rise building extraction result, the distance length in each azimuth direction of the building as a whole, including the facade regularity component, the facade bright line component and the secondary scattering component, is calculated in the SAR image area occupied by the building as a whole.

[0107] For high-rise buildings with secondary scattering components, the distance from the midpoint of the connected domain of the secondary scattering components is taken as the bottom of the building.

[0108] When the extraction results are incomplete, the longest range length in the same final high-rise building extraction results is taken as the range length H of that high-rise building in the SAR image. r .

[0109] S7. Based on the imaging geometry in the SAR image, convert the distance to length of the high-rise building into the actual building height.

[0110] Specifically, the calculation formula is as follows:

[0111] H = H r / cosθ

[0112] Among them, H r Let H be the distance length of the high-rise building, H be the actual building height, and θ be the incident angle of the SAR image. Based on the extracted high-rise building, this application can extract the height of the high-rise building using only the incident angle of the SAR image, without requiring network training or other scene information, thus exhibiting high efficiency and practicality.

[0113] Based on the same inventive concept, this embodiment provides a computer device, such as... Figure 6As shown, it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any step of the method described above.

[0114] Based on the same inventive concept, this embodiment provides a computer-readable storage medium, such as... Figure 7 As shown, a computer program is stored thereon, which, when executed by a processor, implements any step of the method described above.

[0115] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0116] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for extracting the height of tall buildings from a single high-resolution SAR image, characterized in that, Includes the following steps: The SAR image is preprocessed; The salient features are extracted from the SAR image, including bright spots and bright lines; Based on semantic information and the aforementioned salient features, generate a facade regularity component, a facade bright line component, and a secondary scattering component; Based on the spatial relationships between different components, production system rules are formulated to generate high-rise building assumptions; Remove the high-rise building assumptions that do not meet the requirements to obtain the final high-rise building extraction results; Based on the final high-rise building extraction results, the range length of each high-rise building in the SAR image is calculated; Based on the imaging geometry in the SAR image, the distance-direction length of the high-rise building is converted into the actual building height.

2. The method according to claim 1, characterized in that, The preprocessing of the SAR image specifically includes amplitude correction and speckle filtering, employing a nonlocal speckle filtering method.

3. The method according to claim 1, characterized in that, The extraction of salient features from the SAR image, including bright spots and bright lines, specifically involves: A modified spot filter is used for salient feature extraction. The filtered value is: S a (p(x,y))=p(x,y)-min(R step ),step=1,2,3… R step = max(p(x + i, y + j)), max(|i|, |j|) = step and s a (i, j) = 1 Among them, s a For filters with different directions, a = 0, 45, 90, 135 represents the filter direction, p(x,y) is the center point, x is the range position, and y is the azimuth position. step represents the Chebyshev distance from the point to the center point p(x,y); The obtained filtering results are then integrated: F1(x,y)=max(S0(x,y),S 45 (x,y),S 135 (x,y)) F2(x,y)=S 90 (x,y) By selecting an appropriate threshold and binarizing F1 and F2, the extracted bright spots and bright lines can be obtained.

4. The method according to claim 3, characterized in that, The generation of the facade regularity component, facade bright line component, and secondary scattering component based on semantic information and the salient features specifically includes: Generate the facade pattern component based on the semantic information: take the distance to the center point for each orientation of each connected domain in F1; The system iteratively detects each point, checking whether the bright spot still exists within a certain range (x:x+t1,y) from the point (x,y) to the far end. If it does, the distance between the two points is calculated as the floor interval, and compared with the existing floor interval to see if it matches. If it matches, the point is added to the group, and the system continues to detect to the far end based on this point until the bright spot is not detected twice in a row or the interval does not match the group. Set the minimum number of bright spots per group to four; otherwise, do not record them. Finally, the groups that are adjacent in orientation and have the same floor spacing are integrated to form the facade pattern component; Generate the facade highlighting component based on the semantic information: calculate the distance length of each connected component in F2, which is greater than a set value t. f The connected component detection is the facade bright line component. The secondary scattering component is generated based on the semantic information: the mean range width and amplitude of each connected component in F1 are calculated, and the range width is greater than a set value t. b And the average amplitude is greater than the set value A t The connected component detection is the secondary scattering component.

5. The method according to claim 4, characterized in that, The production system rules are formulated based on the spatial relationships between different components to generate the high-rise building hypothesis. The production system rules are specifically as follows: The azimuth distance is less than the set value D f Coverage rate greater than the set value C f Multiple facade components can be combined to form a high-rise building; The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The multiple facade regularity components and the distance to the direction is less than the set value D b The aforementioned facade patterns and secondary scattering components can be combined to form a high-rise building; The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The plurality of the aforementioned facade regularity components and facade bright line components, and the distance to the direction less than the set value D b The aforementioned facade patterns, facade bright lines, and secondary scattering components can be combined to form a high-rise building; The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The aforementioned facade patterns and facade bright line components can be combined to form a high-rise building; The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f Multiple facade bright line components and distances less than the set value D b The aforementioned facade bright lines and secondary scattering components can be combined to form a high-rise building; The azimuth distance is less than the set value D f The coverage rate is greater than the set value C. f The multiple facade lighting components can be combined to form a high-rise building. The directional distance refers to the distance between the rightmost side of the facade pattern component or facade bright line component and the leftmost side of the first right facade pattern component or facade bright line component. The coverage rate refers to the larger value of the ratio of the length of the overlapping part of the facade regular component, the facade bright line component and the adjacent component to the length of the two components in the upward direction; Distance in the direction of distance refers to the closest distance in the same direction of the facade pattern component and the facade bright line component, or the distance between the midpoint of the distance direction of the secondary scattering component and the far end of the facade pattern component and the facade bright line component.

6. The method according to claim 5, characterized in that, The process of removing non-compliant high-rise building assumptions to obtain the final high-rise building extraction result involves the following steps: the same component may belong to different building assumptions; only the high-rise building assumption containing the largest number of components is retained to obtain the most complete high-rise building extraction result; and building assumptions with a building area or floor height less than a set value are removed.

7. The method according to claim 6, characterized in that, Based on the final high-rise building extraction results, the range length of each high-rise building on the SAR image is calculated. Specifically, for each final high-rise building extraction result, the range length corresponding to each azimuth direction of the entire building, including the facade regularity component, the facade bright line component, and the secondary scattering component, is calculated in the SAR image area occupied by the building as a whole in the SAR image area. For a high-rise building with the aforementioned secondary scattering component, the distance from the midpoint of the connected domain of the secondary scattering component is taken as the bottom of the building. When the extraction results are incomplete, the longest range length in the same final high-rise building extraction result is taken as the range length H of that high-rise building in the SAR image. r .

8. The method according to claim 7, characterized in that, The step of converting the distance-oriented length of the high-rise building into its actual height based on the imaging geometry of the SAR image specifically involves: H=H r / cosθ Among them, H r H is the distance length of the high-rise building, H is the actual building height of the high-rise building, and θ is the incident angle of the SAR image.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-8.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1-8.

Images