Airport detection method and device in complex scene based on parallel line cluster positioning

Abstract

The invention discloses an airport detection method and device in a complex scene based on parallel line cluster positioning. According to the method, an obtained synthetic aperture radar image is preprocessed to reduce interference information in the synthetic aperture radar image; the parallel line clusters existing in the preprocessed synthetic aperture radar image are extracted, the interested area corresponding to each parallel line cluster is determined, and the position and the range of the airport target can be still effectively delineated when the long straight line segment corresponding to the airport runway is fractured; and determining the region of interest meeting the homogeneity requirement as the airport region, and performing clustering operation on the parallel line segments in the airport region to determine all airports included in the airport region, so that the airport detection precision can be optimized, the probability of occurrence of a false alarm target can be reduced, and rapid detection of the airport region in a complex scene can be realized.

Description

Technical Field

[0001] This application relates to the field of radar remote sensing applications, specifically to an airport detection method, device, electronic equipment, and medium for complex scenarios based on parallel line cluster positioning. Background Technology

[0002] Airports are important strategic targets in both military and civilian fields. Automatic airport location and identification have significant practical value in aircraft navigation, urban planning, ground reconnaissance, and precision strikes. Synthetic Aperture Radar (SAR) has the advantage of being unaffected by weather and lighting conditions, thus possessing significant practical value in airport target search for airborne or space-based remote sensing applications.

[0003] Since airports are typically scattered across a vast spatial area, detection of airport areas is often hampered by interference from nearby road networks, large urban building areas, and other man-made targets. Designing appropriate feature extraction methods to effectively eliminate complex background interference is crucial for rapid target search and localization in large areas, enabling accurate acquisition of airport target locations using SAR imagery.

[0004] However, existing SAR image airport detection methods all have certain shortcomings, resulting in low accuracy or efficiency in airport detection in complex scenes. Therefore, there is an urgent need for an airport detection method that combines both accuracy and efficiency to meet the needs of airport target detection in complex scenarios in practical applications. Summary of the Invention

[0005] Therefore, this application provides an airport detection method, apparatus, and electronic device based on parallel line cluster positioning in complex scenarios to solve the problem of low accuracy or low efficiency of airport detection in complex scenarios caused by the lack of airport detection methods that combine both detection accuracy and detection efficiency in the prior art.

[0006] To achieve the above objectives, the first aspect of this application provides an airport detection method for complex scenarios based on parallel line cluster localization. The method includes: preprocessing an acquired synthetic aperture radar image; extracting parallel line clusters present in the preprocessed synthetic aperture radar image; determining the region of interest corresponding to each parallel line cluster; identifying the region of interest that meets the homogeneity requirement as an airport region; and performing a clustering operation on the parallel line segments within the airport region to determine all airports included in the airport region.

[0007] According to the first aspect, the preprocessing of the acquired synthetic aperture radar image includes: dividing the synthetic aperture radar image into blocks to obtain multiple image blocks; performing a three-level wavelet decomposition on each image block using a Haar wavelet basis to obtain the decomposition result corresponding to the image block; the decomposition result includes high-frequency components and low-frequency components; performing threshold truncation on the high-frequency components in the decomposition result of each image block based on a preset component threshold; performing inverse wavelet transform on the decomposition result of each image block based on the low-frequency components and the threshold-truncated high-frequency components to reconstruct the image block; and stitching the reconstructed image blocks together in their original positions to obtain the preprocessed synthetic aperture radar image.

[0008] According to the first aspect, or any implementation of the first aspect above, the extraction of parallel line clusters in the preprocessed synthetic aperture radar image includes: extracting line segment information of line segments in the preprocessed synthetic aperture radar image based on a line segment detector; the line segment information includes at least the positions of the beginning and end points, the centroid position, the slope, and the length of the line segment; retaining line segments with a length greater than or equal to a length threshold from all extracted line segments; dividing the retained line segments into multiple parallel line segment groups according to the slope; and determining parallel line clusters based on the parallel line segment groups.

[0009] According to the first aspect, or any implementation of the first aspect above, determining the region of interest corresponding to each parallel line cluster includes: for each parallel line cluster, selecting the maximum and minimum values ​​of the abscissa and the maximum and minimum values ​​of the ordinate from the coordinates of the beginning and end endpoints of each line segment of the parallel line cluster; and determining the region of interest based on the maximum and minimum values ​​of the abscissa and the maximum and minimum values ​​of the ordinate.

[0010] According to the first aspect, or any implementation of the first aspect above, determining the region of interest that meets the homogeneity requirement as the airport region includes: determining the two-dimensional entropy of each region of interest; determining the entropy interval of the two-dimensional entropy of a single ground feature image from the same sensor as the synthetic aperture radar image; and determining the region of interest whose two-dimensional entropy is greater than the maximum value of the entropy interval as the airport region.

[0011] According to the first aspect, or any implementation of the first aspect above, a clustering operation is performed on the parallel line segments within the airport area to determine all airports included in the airport area, including: clustering the parallel line segments within the airport area based on the Mean-Shift method; when the number of clusters is equal to 1, determining that the airport area is the area where an airport is located; when the number of clusters is greater than 1, determining that the airport area is the area where the number of clusters are located; and, based on the parallel line segments included in each cluster, determining the area where an airport is located.

[0012] According to the first aspect, or any implementation of the first aspect above, determining the parallel line cluster based on the parallel line segment group includes: performing line segment aggregation processing for each parallel line segment group to obtain the parallel line cluster in the parallel line segment group; the line segment aggregation processing includes: traversing the parallel line segments in the parallel line segment group and taking each parallel line segment as the center parallel line segment in turn; for each center parallel line segment, determining the centroid distance between the other parallel line segments in the parallel line segment group and the center parallel line segment, and determining the set of parallel line segments whose centroid distance is less than the distance threshold and the center parallel line segment as a parallel line cluster.

[0013] A second aspect of this application provides an airport detection device for complex scenarios based on parallel line cluster localization. The device includes: a preprocessing module for preprocessing an acquired synthetic aperture radar image; an extraction module for extracting parallel line clusters present in the preprocessed synthetic aperture radar image; a segmentation module for determining the region of interest corresponding to each parallel line cluster; an identification module for identifying the region of interest that meets the homogeneity requirement as an airport region; and a clustering module for performing clustering operations on the parallel line segments within the airport region to determine all airports included in the airport region.

[0014] A third aspect of this application provides an electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of the first aspects; and one or more I / O interfaces connected between the processors and the memory, configured to enable information interaction between the processors and the memory.

[0015] A fourth aspect of this application provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method of any one of the first aspects described above.

[0016] This application has the following advantages: The airport detection method based on parallel line cluster localization in complex scenarios provided in this application first preprocesses the acquired synthetic aperture radar image to reduce interference information in the synthetic aperture radar image; then, it extracts the parallel line clusters present in the preprocessed synthetic aperture radar image and determines the region of interest corresponding to each parallel line cluster. This method can effectively delineate the location and range of airport targets even when there are breaks in the long straight line segments corresponding to the airport runway. Furthermore, the region of interest that meets the homogeneity requirement is determined as the airport region, and the parallel line segments within the airport region are clustered to determine all airports included in the airport region. This can optimize the accuracy of airport detection, reduce the probability of false alarm targets, and achieve rapid detection of airport regions in complex scenarios. Attached Figure Description

[0017] The accompanying drawings are provided to further understand this application and form part of the specification. They are used together with the following detailed description to explain this application, but do not constitute a limitation thereof.

[0018] Figure 1 A flowchart illustrating an airport detection method in complex scenarios based on parallel line cluster localization, provided for embodiments of this application;

[0019] Figure 2 This is a flowchart illustrating a method for extracting parallel line clusters, as provided in an embodiment of this application.

[0020] Figure 3 A schematic diagram of the structure of an airport detection device in a complex scenario based on parallel line cluster positioning, provided for an embodiment of this application;

[0021] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solutions of this disclosure, exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments of this disclosure to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0023] Where there is no conflict, the various embodiments of this disclosure and the features thereof in the embodiments may be combined with each other.

[0024] As used in this application, the term "and / or" includes any and all combinations of one or more of the related enumerated entries.

[0025] The terminology used in this application is for describing specific embodiments only and is not intended to limit the application. As used herein, the singular forms "a" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0026] When the terms “comprising” and / or “made of” are used in this application, the presence of the said feature, integral, step, operation, element and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or groups thereof is not excluded.

[0027] The embodiments described in this application can be illustrated with reference to plan views and / or cross-sectional views, using the ideal schematic diagrams of this application. Therefore, the example illustrations may be modified according to manufacturing techniques and / or tolerances.

[0028] Unless otherwise specified, all terms used in this application (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this application, and will not be interpreted as having an idealized or overly formal meaning, unless expressly so defined in this application.

[0029] Airports are important strategic targets in both military and civilian fields. In modern warfare, automatic airport location and identification have significant practical value in aircraft navigation, ground reconnaissance, and precision strikes. In urban planning, airport target detection helps decision-makers in strategy formulation and urban management to identify potential risks and malfunctions in a timely manner, thereby optimizing resource allocation strategies. Synthetic Aperture Radar (SAR) has the advantage of being unaffected by weather and lighting conditions, making it of significant practical value in airport target search for airborne or space-based remote sensing applications.

[0030] Since airports are typically scattered across a vast spatial area, detection of airport areas is often hampered by interference from nearby road networks, large urban building areas, and other man-made targets. Designing appropriate feature extraction methods to effectively eliminate complex background interference is crucial for rapid target search and localization in large areas, enabling accurate acquisition of airport target locations using SAR imagery.

[0031] Existing SAR image airport detection methods can be broadly categorized into three types: methods based on line feature extraction, methods based on image segmentation, and methods based on deep learning. Among these, methods based on line feature extraction are currently the most commonly used; however, the lines detected in actual airport areas often have breaks, easily leading to confusion with other ground features. Image segmentation methods use "dark, homogeneous regions" as the basis for airport presence, and their performance is significantly affected by background complexity. Deep learning methods are constrained by the scarcity of mature datasets of SAR image airport targets. Therefore, selecting and improving appropriate detection methods to overcome their shortcomings in practical applications and establishing a complete detection workflow is essential for improving the efficiency and accuracy of airport detection algorithms and quickly outputting airport target localization results in complex scenes.

[0032] To address this, this application provides a method for airport target detection in complex scenes based on parallel line cluster localization, for rapid detection of airport targets in airborne or spaceborne SAR images. This application designs a complete detection process comprising four stages: image preprocessing, airport area detection, false alarm target exclusion, and target clustering. This achieves accurate output of the location of airport areas in complex scenes, reduces the false alarm rate of the algorithm, and distinguishes different airport areas within the same scene.

[0033] Figure 1 This document presents a flowchart illustrating an airport detection method in complex scenarios based on parallel line cluster localization, as provided in an embodiment of this application. The method is applied to electronic devices, which may be devices configured in aircraft or satellites. Figure 1 As shown, the airport detection method in complex scenarios based on parallel line cluster positioning includes the following steps S101-S104.

[0034] Step S101: Preprocess the acquired synthetic aperture radar image.

[0035] The acquired synthetic aperture radar image is an airborne or spaceborne SAR image to be detected.

[0036] In this embodiment of the application, the acquired synthetic aperture radar image is preprocessed, including the following steps one to five.

[0037] Step 1: Divide the synthetic aperture radar image into blocks to obtain multiple image blocks.

[0038] Dividing the SAR image into 4×4 pixel local blocks (which can be adjusted to, for example, 8×8 depending on the image resolution) can ensure that local features are processed while controlling computational complexity.

[0039] Step 2: For each image block, perform three-level wavelet decomposition using the Haar wavelet basis to obtain the decomposition result corresponding to the image block. The decomposition result includes high-frequency components and low-frequency components.

[0040] In wavelet decomposition, an image is decomposed into multiple frequency bands. Low-frequency components (approximation coefficients) represent the main structure and slowly changing regions of the image (such as large, uniform areas). High-frequency components (detail coefficients) contain the image's edges, textures, and noise (such as speckle noise in SAR images).

[0041] The three-level wavelet decomposition method using the Haar wavelet basis can separate high-frequency components (noise, details) and low-frequency components (main structure) at different scales. It is also computationally efficient and suitable for capturing right-angled edges in SAR images, such as airport runways and building outlines.

[0042] Step 3: Based on the preset component threshold, perform threshold truncation processing on the high-frequency component in the decomposition results of each image block.

[0043] The component threshold can be determined by multiplying the noise standard deviation (which can be estimated from the image background area) and the coefficient (usually 3-5 for SAR images; if the noise is extremely strong, it can be increased appropriately).

[0044] In this embodiment, based on a preset component threshold, threshold truncation processing is performed on the high-frequency component in the decomposition result of each image block, including hard thresholding and soft thresholding. In hard thresholding, coefficients below the component threshold are directly discarded, while coefficients greater than or equal to the component threshold are retained to maximize the preservation of significant edges, suitable for scenarios requiring sharp straight-line features (such as airport runway detection). In soft thresholding, not only are coefficients below the threshold discarded, but coefficients above the threshold are also shrunk towards zero to achieve a better smoothing effect.

[0045] Since the high-frequency component contains both real edge information and noise, threshold truncation is performed on the high-frequency component in the decomposition results of each image block. This can retain significant high-frequency coefficients (corresponding to strong edges) and filter out weak high-frequency coefficients (corresponding to noise), allowing subsequent line detection to focus more on the main structure.

[0046] Step 4: For the decomposition results of each image block, perform inverse wavelet transform based on the low-frequency components and the high-frequency components after threshold truncation to reconstruct the image block.

[0047] Step 5: The reconstructed image blocks are stitched together in their original positions to form a complete image, thereby obtaining a preprocessed synthetic aperture radar image.

[0048] In this embodiment, the image undergoes wavelet decomposition-based preprocessing to filter out noise contained in the high-frequency components, reducing potential interference in the subsequent line detector output. During the filtering process, parameters such as wavelet type, decomposition level, and decomposition window size are appropriately set. In one embodiment, the image is divided into 4×4 blocks, and a Haar wavelet basis is used within each block to perform a three-level decomposition. After setting a threshold to truncate the high-frequency components generated by the decomposition, the image is reconstructed by combining the retained low-frequency components. The different blocks are then stitched back to their respective positions to obtain the preprocessed image result.

[0049] Step S102: Extract the clusters of parallel lines present in the preprocessed synthetic aperture radar image.

[0050] Figure 2 This is a flowchart illustrating a method for extracting clusters of parallel lines, provided as an embodiment of this application. Figure 2As shown, the step of extracting parallel line clusters in the preprocessed synthetic aperture radar image (step S102) includes the following steps S201-S204.

[0051] Step S201: Based on the line segment detector, extract the line segment information of the line segments in the preprocessed synthetic aperture radar image.

[0052] The information of the line segment includes at least the positions of the beginning and end points, the position of the centroid, the slope, and the length.

[0053] In this embodiment, a line segmentation detector (LSD) is used to detect line segments and outputs basic information about the detected line segments on the graph. The output items specifically include the positions of the start and end points of the line segment, the position of the centroid, the slope, and the length. This method can balance the efficiency and accuracy of line segment detection.

[0054] Step S202: From all the extracted line segments, retain the line segments whose length is greater than or equal to the length threshold.

[0055] The length threshold can be set according to the actual application. This method of retaining line segments with a length greater than or equal to the length threshold (or removing line segments with a length less than the length threshold from the detection results) can effectively reduce background interference and create conditions for high accuracy in subsequent airport area detection.

[0056] Step S203: Divide the retained line segments into multiple parallel line segment groups according to the slope.

[0057] In this embodiment of the application, dividing the retained line segments into multiple parallel line segment groups according to the slope includes: defining the range of values ​​for the line segment slope as [0, ... A reasonable interval between groups is set, and line segments are grouped according to their slope, assuming that there is a parallel relationship between line segments in the same group. In one embodiment, the interval between groups is set to 0.02. .

[0058] In this application, dividing the parallel line segment groups can increase fault tolerance and improve the success rate of subsequent airport area detection.

[0059] Step S204: Based on the group of parallel line segments, determine the cluster of parallel lines.

[0060] In this embodiment of the application, determining the parallel line cluster based on the parallel line segment group includes: performing line segment aggregation processing for each parallel line segment group to obtain the parallel line cluster in the parallel line segment group; the line segment aggregation processing includes: step a and step b.

[0061] Step a: Traverse the parallel line segments in the group of parallel line segments, and take each parallel line segment as the center parallel line segment in turn.

[0062] Step b: For each central parallel line segment, determine the centroid distance between the other parallel line segments in the parallel line segment group and the central parallel line segment, and determine the set of parallel line segments whose centroid distance is less than the distance threshold, along with the central parallel line segment, as a parallel line cluster.

[0063] In this embodiment of the application, the process of determining the cluster of parallel lines is to determine the clustering of parallel line segments based on the position and number of each line segment in the group of parallel line segments.

[0064] Specifically, for each group of parallel line segments, the centroid distance between each member of the group and the other members is calculated. If there exists a line segment whose distance to multiple other line segments is less than a set threshold, then the group of line segments with similar distances is retained as a parallel line group. The distance threshold is set based on the maximum length of the lines in the group and is proportional to the maximum length of the lines in the group. The formula for judging the degree of line clustering is as follows:

[0065]

[0066] Assume the first The slope of the line segment is at [ Within the interval, there are a total of Parallel line segments, among which These are two line segments in group k. and These are the centroid coordinates of the m-th line segment. and These are the centroid coordinates of the nth line segment. Two parallel line segments The Euclidean distance between them For the first The first group of parallel line segments The length of the line segment, The proportional coefficient for controlling the threshold, Distance threshold For the first The number of line segments within a group that meet the aggregation criteria.

[0067] It should be noted that if no parallel line clusters can be identified in any of the parallel line segment groups, it is assumed that there is no airport area in the SAR image, and no further steps are required.

[0068] Step S103: Determine the Region of Interest (ROI) corresponding to each cluster of parallel lines.

[0069] In this embodiment of the application, determining the region of interest corresponding to each cluster of parallel lines includes: the following steps one and two.

[0070] Step 1: For each cluster of parallel lines, select the maximum and minimum values ​​of the x-coordinate, and the maximum and minimum values ​​of the y-coordinate, from the coordinates of the first and last endpoints of each line segment in the cluster.

[0071] Step 2: Based on the maximum and minimum values ​​of the horizontal axis and the maximum and minimum values ​​of the vertical axis, determine the region of interest.

[0072] The region of interest is a rectangular area.

[0073] In this embodiment of the application, the region of interest is determined based on the maximum and minimum values ​​of the horizontal axis and the maximum and minimum values ​​of the vertical axis, including the following steps a1-a4.

[0074] Step a1: Determine the minimum value of the x-coordinate as the x-coordinate value of the upper left and lower left vertices of the region of interest.

[0075] Step a2: Determine the maximum value of the x-coordinate as the x-coordinate value of the upper right and lower right vertices of the region of interest.

[0076] Step a3: Determine the minimum value of the ordinate as the ordinate value of the lower left and lower right vertices of the region of interest.

[0077] Step a4: Determine the maximum value of the ordinate as the ordinate value of the top left and top right vertices of the region of interest.

[0078] In this embodiment, the coordinates of the beginning and end of all parallel line segments within the RoI region are counted, and the minimum and maximum values ​​of the horizontal and vertical coordinates are recorded respectively, serving as the basis for calculating the four vertices of the RoI region. Specifically, the minimum value of the horizontal and vertical coordinates recorded in the line cluster is used as the horizontal and vertical coordinates of the upper left vertex of the rectangular region; the maximum value of the recorded horizontal and vertical coordinates is used as the horizontal and vertical coordinates of the lower right vertex of the rectangular region.

[0079] In this application, the regions of interest corresponding to each cluster of parallel lines can be largely identified as airport areas. This method for locating and delineating airport areas based on clusters of parallel lines can effectively detect the location and extent of airport targets even when there are breaks in the long straight segments corresponding to airport runways, achieving relatively ideal results in airport target detection under complex backgrounds. However, to further improve the accuracy of the identified airport areas, these regions of interest need to be further examined to determine whether they can be considered airport areas.

[0080] Step S104: The region of interest that meets the homogeneity requirement is identified as the airport area.

[0081] Homogeneity refers to the consistency and uniformity of the composition, properties, or distribution of a substance, system, or structure within a specific spatial range. Individual land features (such as grasslands and farmland) have high homogeneity (low two-dimensional entropy), while airport areas, due to their complex structures including runways, aprons, and aircraft, have low homogeneity (high two-dimensional entropy).

[0082] In this embodiment of the application, the region of interest that meets the homogeneity requirement is identified as the airport area, including the following steps one to three.

[0083] Step 1: Determine the two-dimensional entropy of each region of interest.

[0084] The method for calculating two-dimensional entropy is as follows:

[0085]

[0086] in, It is the number of gray levels in the region of interest. Is the grayscale value The probability of a pixel appearing in the image (region of interest).

[0087] Step 2: Determine the entropy interval of the two-dimensional entropy of a single ground feature image from the same sensor as the synthetic aperture radar image.

[0088] Step 3: Define the region of interest whose two-dimensional entropy is greater than the maximum value of the entropy interval as the airport region.

[0089] In this embodiment, by comparing the two-dimensional entropy of the RoI with the homogeneity interval of a known single ground feature, simple homogeneous areas that do not conform to airport characteristics can be filtered out. This method leverages the complex and varied fine structure within airport areas, sets adaptive thresholds for different sensor systems, and further verifies airport detection results based on regional homogeneity. This optimizes the accuracy of airport detection and reduces the probability of false alarms.

[0090] Step S105: Perform clustering operations on the parallel line segments within the airport area to determine all airports included in the airport area.

[0091] Clustering is used to classify parallel line segments within an airport area, thereby determining the number of airports within that area and the extent of each airport.

[0092] In this embodiment of the application, a clustering operation is performed on the parallel line segments within the airport area to determine all airports included in the airport area, including the following steps c1-c4.

[0093] Step c1: Cluster parallel line segments within the airport area based on the Mean-Shift method.

[0094] The Mean-Shift method requires setting the detection radius *r* and the class merging threshold *Th*. These two parameters can be determined empirically based on the average size of airports on SAR images. It is known that during urban development, different airports are generally not too close together. Therefore, the detection radius *r* can be set to the average width of the airport, and the merging threshold *Th* can be set to 2-3 times the average width of the airport.

[0095] Step c2: When the number of clusters formed is equal to 1, the airport region is determined to be the region where an airport is located.

[0096] Step c3: When the number of clusters formed is greater than 1, determine the airport region as the region where the number of airports in each cluster is located, and determine the region where an airport is located based on the parallel line segments contained in each cluster.

[0097] The steps for determining the region where an airport is located based on the parallel line segments contained in each category include: For each category of parallel line segments, perform the following operations: obtain the start and end coordinates of all parallel line segments within that category, and record the minimum and maximum values ​​of the horizontal and vertical coordinates, which will serve as the basis for calculating the four vertices of the region where the airport is located. Specifically, the minimum value of the recorded horizontal and vertical coordinates for that category will be used as the horizontal and vertical coordinates of the top-left vertex of the region (rectangular region) where the airport is located; the maximum value of the recorded horizontal and vertical coordinates will be used as the horizontal and vertical coordinates of the bottom-right vertex of the region where the airport is located.

[0098] In this embodiment, considering that when the width of the satellite image is large, there may be more than one airport target in the airport area determined in step S104, a clustering method with a custom number of categories is used to further divide the number of airport targets in the detection results, which can effectively distinguish different airport targets in the same scene. For example, if there are more than two airport areas on the image and the runway slopes are similar, after steps S101-S104, they will form the same large airport area. Using the Mean-Shift method to cluster the line segments contained within the airport area according to the centroid position can effectively deal with the above situation. When the number of clusters is greater than 1, the airport location and range will be re-determined according to the grouping results of the line segments after clustering using the method in S103.

[0099] The airport detection method for complex scenarios based on parallel line cluster localization provided in this application first preprocesses the acquired synthetic aperture radar (SAR) image to reduce interference information. Then, it extracts parallel line clusters from the preprocessed SAR image and determines the region of interest (ROI) corresponding to each parallel line cluster. This method can effectively delineate the location and range of airport targets even when there are breaks in the long straight lines corresponding to the airport runway. Furthermore, the ROI that meets the homogeneity requirement is determined as the airport region, and the parallel line segments within the airport region are clustered to identify all airports included in the airport region. This method can optimize the accuracy of airport detection, reduce the probability of false alarms, and achieve rapid detection of airport regions in complex scenarios.

[0100] This application proposes a method for airport target detection in complex scenes based on parallel line cluster localization, taking into account the quantity, slope, length, and positional distribution of line segments detected by LSD (Line Segment Detection). This application establishes a complete SAR image airport target detection workflow, enabling the determination of the presence of airport targets in SAR images without prior knowledge, and identifying their location and distribution range when they exist. Furthermore, this application improves the algorithm's detection accuracy by incorporating regional homogeneity features and considers the case where two or more airport targets with similar distribution directions exist in the same scene, taking measures to label them separately. This application eliminates the need for pre-training the model with a dataset before the detection step, providing an effective means for SAR image airport target detection in complex scenes and offering a reference method for similar problems.

[0101] The steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this application. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, but without changing the core design of the algorithm and process, are also within the scope of protection of this application.

[0102] Secondly, embodiments of this application provide an airport detection device for complex scenarios based on parallel line cluster positioning.

[0103] Figure 3 This is a schematic diagram of the structure of an airport detection device in a complex scenario based on parallel line cluster positioning, provided as an embodiment of this application. Figure 3 As shown, the device includes: a preprocessing module 31, an extraction module 32, a segmentation module 33, an identification module 34, and a clustering module 35.

[0104] The preprocessing module 31 is used to preprocess the acquired synthetic aperture radar image.

[0105] Extraction module 32 is used to extract parallel line clusters present in the preprocessed synthetic aperture radar image.

[0106] The partitioning module 33 is used to determine the region of interest corresponding to each cluster of parallel lines.

[0107] The identification module 34 is used to identify the region of interest that meets the homogeneity requirement as the airport area.

[0108] Clustering module 35 is used to perform clustering operations on parallel line segments within the airport area to determine all airports included in the airport area.

[0109] In the airport detection device for complex scenarios based on parallel line cluster localization provided in this application, the preprocessing module preprocesses the acquired synthetic aperture radar (SAR) image to reduce interference information in the SAR image; the extraction module extracts parallel line clusters present in the preprocessed SAR image; the segmentation module determines the region of interest corresponding to each parallel line cluster, which can effectively delineate the location and range of airport targets even when there are breaks in the long straight line segments corresponding to the airport runway; the identification module determines the region of interest that meets the homogeneity requirement as the airport area; and the clustering module is used to cluster the parallel line segments in the airport area to determine all airports included in the airport area, which can optimize the accuracy of airport detection, reduce the probability of false alarm targets, and realize rapid detection of airport areas in complex scenarios.

[0110] It should be clarified that this application is not limited to the specific configurations and processes described in the above embodiments and shown in the figures. For the sake of convenience and brevity, detailed descriptions of known methods are omitted here, and the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0111] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. See also... Figure 4 The electronic device includes: one or more processors 401; a memory 402 storing one or more programs, which, when executed by one or more processors, enable the one or more processors to implement the airport detection method in complex scenarios based on parallel line cluster positioning as described above; and one or more I / O interfaces 403 connected between the processors and the memory, configured to enable information interaction between the processors and the memory.

[0112] Among them, processor 401 is a device with data processing capabilities, including but not limited to central processing unit (CPU); memory 402 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory (FLASH); I / O interface (read-write interface) 403 is connected between processor 401 and memory 402, and can realize information interaction between processor 401 and memory 402, including but not limited to data bus (Bus).

[0113] In some embodiments, the processor 401, memory 402, and I / O interface 403 are interconnected via a bus, and thus connected to other components of the computing device.

[0114] This embodiment also provides a computer-readable medium having a computer program stored thereon. When the program is executed by a processor, it implements the airport detection method for complex scenarios based on parallel line cluster positioning provided in this embodiment. To avoid repetition, the specific steps of the airport detection method for complex scenarios based on parallel line cluster positioning will not be repeated here.

[0115] Those skilled in the art will understand that all or some of the steps, systems, or apparatuses in the methods, systems, and apparatuses described above can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0116] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0117] Those skilled in the art will understand that although some embodiments described herein include certain features that are included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this embodiment and form different embodiments.

[0118] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of this application, and this application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of this application, and these modifications and improvements are also considered to be within the scope of protection of this application.

Claims

1. An airport detection method in complex scenarios based on parallel line cluster localization, characterized in that, The method includes: Preprocess the acquired synthetic aperture radar images; Extract the clusters of parallel lines present in the preprocessed synthetic aperture radar image; Determine the region of interest corresponding to each cluster of parallel lines; The region of interest that meets the homogeneity requirement is defined as the airport area; Clustering is performed on parallel line segments within the airport area to determine all airports included in the airport area.

2. The method according to claim 1, characterized in that, The preprocessing of the acquired synthetic aperture radar image includes: The synthetic aperture radar image is divided into blocks to obtain multiple image blocks; For each image patch, a three-level wavelet decomposition is performed using the Haar wavelet basis to obtain the decomposition result corresponding to the image patch; the decomposition result includes high-frequency components and low-frequency components; Based on a preset component threshold, threshold truncation is performed on the high-frequency components in the decomposition results of each image block. For each image patch, an inverse wavelet transform is performed based on the low-frequency components and the high-frequency components after threshold truncation to reconstruct the image patch. The reconstructed image blocks are stitched together in their original positions to form a complete image, thus obtaining a preprocessed synthetic aperture radar image.

3. The method according to claim 2, characterized in that, The extraction of parallel line clusters in the preprocessed synthetic aperture radar image includes: Based on the line segment detector, the line segment information of the line segments in the preprocessed synthetic aperture radar image is extracted; the line segment information includes at least the positions of the first and last endpoints, the position of the centroid, the slope, and the length of the line segment; From all extracted line segments, retain those with a length greater than or equal to a length threshold; The retained line segments are divided into multiple groups of parallel line segments based on the slope; Based on the group of parallel line segments, a cluster of parallel lines is determined.

4. The method according to claim 3, characterized in that, Determining the region of interest corresponding to each cluster of parallel lines includes: For each cluster of parallel lines, select the maximum and minimum values ​​of the abscissa, and the maximum and minimum values ​​of the ordinate from the coordinates of the first and last endpoints of each segment of the cluster. The region of interest is determined based on the maximum and minimum values ​​of the horizontal coordinate and the maximum and minimum values ​​of the vertical coordinate.

5. The method according to claim 4, characterized in that, The step of identifying the region of interest that meets the homogeneity requirement as the airport region includes: Determine the two-dimensional entropy of each region of interest; Determine the entropy interval of a single ground feature image from the same sensor as the synthetic aperture radar image; The region of interest whose two-dimensional entropy is greater than the maximum value of the entropy interval is defined as the airport region.

6. The method according to claim 5, characterized in that, The clustering operation on parallel line segments within the airport area to determine all airports included in the airport area includes: Clustering of parallel line segments within the airport area is performed based on the Mean-Shift method; When the number of clusters formed is equal to 1, the airport region is determined to be the region where an airport is located; When the number of clusters formed is greater than 1, the airport region is determined to be the region where the number of airports in each cluster is located, and an airport region is determined based on the parallel line segments contained in each cluster.

7. The method according to claim 3, characterized in that, The step of determining the parallel line cluster based on the parallel line segment group includes: For each group of parallel line segments, a line segment clustering process is performed to obtain a cluster of parallel lines within the group; the line segment clustering process includes: Traverse the parallel line segments in the group of parallel line segments, and take each of the parallel line segments as the center parallel line segment in turn; For each central parallel line segment, determine the centroid distance between the other parallel line segments in the parallel line segment group and the central parallel line segment, and determine the set of parallel line segments whose centroid distance is less than the distance threshold, along with the central parallel line segment, as a parallel line cluster.

8. An airport detection device for complex scenarios based on parallel line cluster positioning, characterized in that, The device includes: The preprocessing module is used to preprocess the acquired synthetic aperture radar images; The extraction module is used to extract parallel line clusters present in the preprocessed synthetic aperture radar image; The partitioning module is used to determine the region of interest corresponding to each cluster of parallel lines; The identification module is used to identify the region of interest that meets the homogeneity requirement as the airport area; The clustering module is used to perform clustering operations on parallel line segments within the airport area to determine all airports included in the airport area.

9. An electronic device, characterized in that, include: One or more processors; A storage device having stored one or more programs thereon, which, when executed by the one or more processors, cause the one or more processors to implement the method according to any one of claims 1-7; One or more I / O interfaces are connected between the processor and the memory and configured to enable information interaction between the processor and the memory.

10. A computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method according to any one of claims 1-7.

Images