Method and device for detecting aircraft targets in SAR remote sensing images

Abstract

The embodiments of the present invention relate to the technical field of SAR remote sensing aircraft detection, and provide a method and apparatus for detecting aircraft targets in SAR remote sensing images. The method comprises: obtaining a SAR remote sensing image of a target to be detected; inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, extracting multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; performing feature aggregation on the multi-scale local features of the target to obtain target interaction information between the local and global information of the target aircraft; and performing aircraft target detection on the target SAR remote sensing image based on the target interaction information. In this way, global perspective information is fully captured and the impact of aircraft of different sizes is reduced, thereby improving the detection accuracy and efficiency of different types of aircraft and other targets in SAR remote sensing images.

Description

Technical Field

[0001] The present invention relates to the technical field of SAR remote sensing aircraft detection, and in particular to a method and device for detecting aircraft targets in SAR remote sensing images. Background Art

[0002] Synthetic Aperture Radar (SAR) imagery is unrestricted by weather conditions and the time of day. It can capture data in a variety of complex environments and extract useful information about ground targets from complex SAR image data, enabling accurate target detection. This technology plays a vital role in both military and non-military fields. Early SAR aircraft target detection techniques primarily relied on manual feature extraction and classifier processing, making them difficult to apply on a large scale. Therefore, accurately detecting aircraft targets in SAR images has been a research hotspot.

[0003] Early aircraft target detection methods focused on simple machine learning approaches. For example, methods such as filtering, principal component analysis (PCA), and histogram of oriented gradients were used to extract features, and methods such as Adaboost, KNN, and SVM were used to classify the features. After feature extraction and classification of candidate regions, the presence of a target is determined. Similarly, methods such as CFAR use thresholds to distinguish targets from noise, clutter, and interference to search for targets. Since then, a series of improved CFAR methods have been widely proposed. For example, Ai J et al. employed an adaptive truncated clutter statistical model based on the lognormal distribution to improve the performance of TPCFAR detectors. Hou B et al. developed a novel multi-layer CFAR target detector to mitigate errors caused by loss of detail information. Although these traditional machine learning methods can effectively detect targets, they still suffer from poor robustness, low detection accuracy, and weak generalization. Furthermore, these target detection methods often rely on manual feature design and screening, resulting in limited detection accuracy.

[0004] In recent years, with the rapid development of deep learning technology, it has been widely applied to SAR remote sensing target detection and recognition tasks. However, in SAR images with complex backgrounds, it is difficult to extract features for small aircraft targets, and scattered points are more discrete, making detection more challenging. Therefore, many researchers have designed various solutions to address these limitations and further improve SAR target detection accuracy. For example, Wang Sy et al. designed a new LeNet-5 detection framework and used data augmentation methods to achieve coarse and rapid localization of candidate aircraft targets in large-scale SAR scenarios. Li M et al. proposed a lightweight detection model (LDM) that mainly includes a reuse module (RB) based on the Yolov3 framework and an information correction module (ICB). The RB module helps the neural network extract rich aircraft features by aggregating multi-layer information. By extracting grayscale features and enhancing spatial information, it helps suppress interference and redundant information in complex environments. Zhao S et al. proposed a Res-Clo network to denoise SAR images as a preprocessing step to improve detection accuracy. Subsequently, an improved network, DML-YoloV8, was designed based on the YoloV8 network. A specially designed MFB module was integrated into the feature extraction layer to effectively broaden the network's receptive field. Deformable convolution was also introduced into the feature fusion layer to enhance the network's multi-scale detection capabilities. Currently, SAR aircraft target detection algorithms, both domestically and internationally, have improved detection accuracy to a certain extent, but suffer from high model complexity and weak generalization capabilities.

[0005] The main problems in aircraft target detection in SAR remote sensing images are as follows:

[0006] SAR remote sensing images have complex content and large semantic ambiguity. Different aircraft targets in SAR remote sensing images also have similar appearances such as shape and color. This situation can easily lead to misclassification by simple aircraft target detection methods, that is, SAR aircraft targets cannot be directly located and detected based on their physical appearance. At the same time, a large amount of noise and false information poses a huge challenge to the accurate detection of aircraft targets.

[0007] SAR images contain a wide range of scales for different targets. Single-scale feature extractors cannot effectively capture global and contextual information from SAR images. Similarly, simple feature extractors reuse redundant information when capturing deep features of aircraft targets, reducing the representation of local details and ultimately affecting aircraft detection performance in SAR images. Summary of the Invention

[0008] The present invention provides a method and device for detecting aircraft targets in SAR remote sensing images, which are used to overcome the defects of low accuracy and great difficulty in detecting aircraft targets in SAR remote sensing images in the prior art, and to improve the detection accuracy and efficiency of different types of aircraft and other targets in SAR remote sensing images.

[0009] The present invention provides a method for detecting aircraft targets in SAR remote sensing images, comprising the following steps.

[0010] Acquire SAR remote sensing images of the target to be detected;

[0011] Inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting target multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model;

[0012] Performing feature aggregation on the multi-scale local features of the target to obtain target interaction information between local and global information of the target aircraft;

[0013] Aircraft target detection is performed on the target SAR remote sensing image based on the target interaction information, and the detection result includes target frame regression and aircraft classification.

[0014] In one possible implementation, the method further includes:

[0015] The SAR remote sensing aircraft detection model is trained based on the following method:

[0016] Collect SAR remote sensing images from different types of aircraft and perform image processing;

[0017] The SAR remote sensing images after image processing are divided into a model training set, a model verification set and a model test set;

[0018] The improved YoloV8 network is trained using the model training set and the model validation set;

[0019] The trained YoloV8 network is tested using the model test set. When the test result meets the preset conditions, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained.

[0020] In one possible implementation, the method further includes:

[0021] Acquire multi-scale local features of aircraft in the SAR remote sensing image in the model training set through the improved YoloV8 network;

[0022] Based on the SAR remote sensing images in the model training set, an improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model;

[0023] Aggregating the multi-scale local features layer by layer in the initial model to obtain interaction information between local and global information;

[0024] Based on the interactive information, the rotating detection head is used to regress the aircraft frame and classify the aircraft target;

[0025] When the accuracy of the regression results of the initial model on the aircraft frame and the classification results of the aircraft target reaches a first preset threshold, the initial model is verified using the model verification set.

[0026] In one possible implementation, the method further includes:

[0027] Squeezing the SAR remote sensing images in the model training set through a convolutional layer;

[0028] The squeezed feature map is passed through multiple maximum pooling operations and residual structures;

[0029] Using a global average pooling layer and a global maximum pooling layer to obtain global and contextual semantic information, and fusing the global and contextual semantic information with edge detail information;

[0030] Based on the information fusion results, global semantics and target edge details are modeled to obtain the initial model.

[0031] In one possible implementation, the method further includes:

[0032] When the initial model is verified, the SAR remote sensing images in the model test set are input into the verified initial model to test the model;

[0033] When the test result shows that the detection accuracy of the initial model for the model test set reaches a second preset threshold, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained.

[0034] In one possible implementation, the method further includes:

[0035] Performing image processing on the SAR remote sensing image to obtain an image-processed SAR remote sensing image, wherein the image processing includes random translation, rotation, color conversion, normalization and flipping processing operations.

[0036] The present invention also provides a device for detecting aircraft targets in SAR remote sensing images, comprising the following modules:

[0037] An acquisition module is used to acquire SAR remote sensing images of the target to be detected;

[0038] a feature extraction module, configured to input the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extract multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model;

[0039] A feature aggregation module is used to perform feature aggregation on the multi-scale local features of the target to obtain target interaction information between the local and global information of the target aircraft;

[0040] A detection module is used to perform aircraft target detection on the target SAR remote sensing image based on the target interaction information, and the detection result includes target frame regression and aircraft classification.

[0041] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for detecting aircraft targets in SAR remote sensing images as described above is implemented.

[0042] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for detecting aircraft targets in SAR remote sensing images as described above is implemented.

[0043] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above methods for detecting aircraft targets in SAR remote sensing images.

[0044] The present invention provides a method and device for detecting aircraft targets in SAR remote sensing images. The method and device obtain a SAR remote sensing image of a target to be detected; input the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model; extract the multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; perform feature aggregation on the multi-scale local features of the target to obtain target interaction information between the local and global information of the target aircraft; and perform aircraft target detection on the target SAR remote sensing image based on the target interaction information. The detection results include target box regression and aircraft classification. Compared with the existing technology that has the defects of low accuracy and high detection difficulty in SAR remote sensing image aircraft target detection, this solution fully captures global perspective information and reduces the impact of aircraft of different sizes, thereby improving the detection accuracy and efficiency of different types of aircraft and other targets in SAR remote sensing images. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 The present invention provides a flow chart of a method for detecting aircraft targets in SAR remote sensing images.

[0047] Figure 2 This is one of the flow charts of the training method for the SAR remote sensing aircraft detection model provided by the present invention.

[0048] Figure 3 This is the second flow chart of the training method for the SAR remote sensing aircraft detection model provided by the present invention.

[0049] Figure 4 The present invention provides a schematic structural diagram of an aircraft target detection device in SAR remote sensing images.

[0050] Figure 5 This is a schematic diagram of the network structure of the SAR remote sensing aircraft detection model provided by the present invention.

[0051] Figure 6 It is a schematic diagram of the detection results of the SAR remote sensing aircraft detection model provided by the present invention.

[0052] Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0053] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0054] To facilitate understanding of the embodiments of the present invention, specific embodiments will be further explained below with reference to the accompanying drawings. The embodiments do not limit the embodiments of the present invention.

[0055] Figure 1 FIG. 1 is a flow chart of a method for detecting aircraft targets in SAR remote sensing images provided by the present invention, such as Figure 1 As shown, the method specifically includes:

[0056] S11. Acquire a SAR remote sensing image of a target to be detected.

[0057] Synthetic Aperture Radar (SAR) remote sensing technology is a remote sensing technology that uses microwaves for imaging. It has all-weather and all-day observation capabilities and is particularly suitable for detecting targets such as aircraft.

[0058] In this embodiment of the present invention, the SAR remote sensing image to be inspected must first be acquired from a SAR remote sensing data source. These data sources may include satellites, unmanned aerial vehicles (UAVs), or ground-based SAR systems. The acquired image has high resolution and can clearly display targets such as aircraft. The acquired SAR remote sensing image may be in a specific image format, resolution, and size to ensure smooth subsequent processing.

[0059] S12: inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting target multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model.

[0060] The SAR remote sensing aircraft detection model is a deep learning-based model that can extract aircraft features from SAR remote sensing images for aircraft target detection. The specific training process is described in Figure 2 and Figure 3 The corresponding embodiments are described in detail and will not be described in detail here.

[0061] The acquired SAR remote sensing image of the target is then fed into a pre-trained SAR remote sensing aircraft detection model. This model is typically trained on a large dataset of SAR remote sensing images and boasts high detection accuracy. After the image is fed into the model, it automatically extracts features of the target aircraft, including its shape, size, and texture. These features are then used to extract multi-scale local features of the target.

[0062] S13. Perform feature aggregation on the target multi-scale local features to obtain target interaction information between the local and global information of the target aircraft.

[0063] Feature aggregation is a method of fusing multiple features to extract higher-level features of the target aircraft.

[0064] First, the extracted multi-scale local features of the target need to be screened and integrated to remove redundant and noisy features.

[0065] Then, these features are fused using feature aggregation methods to obtain the interactive information between the local and global information of the target aircraft. This interactive information may include the relative position relationship between the various parts of the aircraft, texture changes, etc.

[0066] Feature aggregation methods may include pooling layers and attention mechanisms in convolutional neural networks (CNNs).

[0067] S14. Performing aircraft target detection on the SAR remote sensing image of the target based on the target interaction information. The detection result includes target frame regression and aircraft classification.

[0068] Target box regression and classification are two core tasks in target detection, which are used to determine the location and category of the target respectively.

[0069] Based on the extracted target interaction information, the target box regression method is used to determine the position and size of the target aircraft. This is usually achieved by predicting the four boundary coordinates of the target box. At the same time, the classification method is used to determine the category of the target aircraft. The probability values ​​of different categories can be calculated, and the category with the highest probability value is selected as the final classification result. In practical applications, target box regression and classification are usually performed simultaneously, and the detection accuracy can be improved through joint optimization. Figure 6 Schematic diagram of the test results.

[0070] The above steps constitute the complete SAR remote sensing aircraft detection process. By acquiring SAR remote sensing images of the target to be detected, inputting them into a pre-trained model for feature extraction, performing feature aggregation, and then performing bounding box regression and classification based on interaction information, accurate aircraft detection can be achieved in SAR remote sensing images. This process has broad application prospects in military reconnaissance, civil aviation, and other fields.

[0071] The present invention provides a method for detecting aircraft targets in SAR remote sensing images. The method comprises the following steps: obtaining a SAR remote sensing image of a target to be detected; inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model; extracting multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; performing feature aggregation on the multi-scale local features of the target to obtain target interaction information between the local and global information of the target aircraft; and performing aircraft target detection on the target SAR remote sensing image based on the target interaction information. The detection results include target box regression and aircraft classification. Compared to the existing defects of low accuracy and high detection difficulty in SAR remote sensing image aircraft target detection, the present method fully captures global perspective information and reduces the impact of aircraft of different sizes, thereby improving the detection accuracy and efficiency of different types of aircraft and other targets in SAR remote sensing images.

[0072] Figure 2 This is one of the flow charts of the training method of the SAR remote sensing aircraft detection model provided by the present invention, such as Figure 2 As shown, the method specifically includes:

[0073] S21. Collect SAR remote sensing images from different types of aircraft and perform image processing.

[0074] A SAR remote sensing image dataset is collected of different types of aircraft and aircraft in different postures (takeoff, landing, cruising, etc.). For example, a total of 1,000 images include seven categories: Boeing 787, A220, ARJ21, A330, A320 / 321, Boeing 737-800, and others, and the data categories are unbalanced.

[0075] The collected raw SAR images may contain interference such as noise and clutter, requiring preprocessing to improve image quality. To enhance the accuracy of SAR aircraft target detection and increase the number of samples, multiple preprocessing operations are employed, including random translation, rotation, color transformation, normalization, and flipping. Secondly, to further improve detection performance, larger images are cropped to a size of 1024 × 1024 using a cropping step of r = 500. This processing method minimizes the effects of varying image sizes.

[0076] S22. Divide the processed SAR remote sensing images into a model training set, a model verification set, and a model test set.

[0077] In order to ensure the smooth progress of the experiments in the embodiments of the present invention and to demonstrate the superiority of the SAR aircraft target detection framework on small samples, all samples are divided into three parts: training, testing and validation sets. 400 and 100 SAR remote sensing images after image processing are randomly selected as training and validation samples, and the remaining 500 are test samples.

[0078] S23. Training the improved YoloV8 network using the model training set and the model verification set.

[0079] The SAR remote sensing aircraft detection model framework of the embodiment of the present invention mainly uses an improved YoloV8 network to obtain the multi-scale local details of the aircraft in the image, and uses the improved SPPF_AA component in the network to achieve the modeling of global semantics and target edge details. At the same time, these multi-scale features are aggregated layer by layer to achieve the interaction between local and global information. Then, the rotating detection head (OBB) is used to regress the target box and classify the target. The overall network structure of the framework is as follows Figure 5As shown in the figure, Conv(·) represents the convolution operation; SCDown represents the spatial and channel decoupled downsampling operation; Usample represents the upsampling operation; Concat represents the feature concatenation operation; C2F and C2FCIB represent the feature extraction layers; SPPF_AA represents the feature fusion module of the improved pyramid structure; PSA represents the local self-attention module; Concat represents the feature concatenation.

[0080] The improved YoloV8 network is trained using the training set data. During training, the model iteratively learns the mapping relationship between image features and labels, gradually optimizing model parameters. Simultaneously, the model is validated using the validation set data, and the training strategy (such as early stopping and adjusting the learning rate) is adjusted based on the validation results.

[0081] S24, testing the trained YoloV8 network through the model test set, when the test result meets the preset conditions, determining that the improved YoloV8 network training is completed, and obtaining the SAR remote sensing aircraft detection model.

[0082] The trained YoloV8 network is tested using the test set data. During the test, the model predicts each image in the test set and outputs information such as the location and category of the aircraft target.

[0083] Furthermore, based on the test results, the model's performance indicators (such as precision, recall, F1 score, etc.) are calculated. These indicators are used to evaluate the model's ability to detect aircraft in SAR remote sensing images.

[0084] Model training is considered complete when the test results meet pre-defined performance criteria (e.g., accuracy exceeding a certain threshold). The resulting improved YoloV8 network is now the SAR remote sensing aircraft detection model. This model can be used in subsequent aircraft detection tasks to automatically identify and locate aircraft in SAR remote sensing imagery.

[0085] The training method for a SAR remote sensing aircraft detection model provided by the present invention performs a series of enhancement processes on aircraft targets in SAR images, enhancing the distinction between aircraft and background and further reducing ambiguity between similar aircraft targets. Building on the SPPF module, the method utilizes global average pooling and global maximum pooling layers to incorporate global background information and target edge details, helping the network make better judgments. This method also fully captures global perspective information and mitigates the impact of aircraft of varying sizes.

[0086] Figure 3 This is the second flow chart of the training method of the SAR remote sensing aircraft detection model provided by the present invention, such as Figure 3As shown, the method specifically includes:

[0087] S31. Collect SAR remote sensing images from different types of aircraft and perform image processing.

[0088] A SAR remote sensing image dataset is collected of different types of aircraft and aircraft in different postures (takeoff, landing, cruising, etc.). For example, a total of 1,000 images include seven categories: Boeing 787, A220, ARJ21, A330, A320 / 321, Boeing 737-800, and others, and the data categories are unbalanced.

[0089] The collected raw SAR images may contain interference such as noise and clutter, requiring preprocessing to improve image quality. To enhance the accuracy of SAR aircraft target detection and increase the number of samples, multiple preprocessing operations are employed, including random translation, rotation, color transformation, normalization, and flipping. Secondly, to further improve detection performance, larger images are cropped to a size of 1024 × 1024 using a cropping step of r = 500. This processing method minimizes the effects of varying image sizes.

[0090] S32. Divide the processed SAR remote sensing images into a model training set, a model verification set, and a model test set.

[0091] The processed SAR remote sensing image dataset is divided into three parts according to a certain ratio (e.g., 70% training set, 15% validation set, and 15% test set). The training set is used to train the model, the validation set is used to adjust model parameters (such as learning rate and batch size) during training, and the test set is used to evaluate the final performance of the model.

[0092] Before or after dividing the dataset, it is necessary to annotate the aircraft targets in the image and generate label files containing information such as target location and category. These label files will serve as supervisory information for model training.

[0093] Specifically, in order to ensure the smooth progress of the experiment in the embodiment of the present invention and to demonstrate the superiority of the SAR aircraft target detection framework on small samples, all samples are divided into three parts: training, testing and validation sets. 400 and 100 SAR remote sensing images after image processing are randomly selected as training and validation samples, and the remaining 500 are test samples.

[0094] S33. Acquire multi-scale local features of aircraft in the SAR remote sensing image in the model training set through the improved YoloV8 network.

[0095] S34. Based on the SAR remote sensing image in the model training set, an improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model.

[0096] S35. Aggregate the multi-scale local features layer by layer in the initial model to obtain interactive information between local and global information.

[0097] S36 , performing aircraft frame regression and aircraft target classification using the rotating detection head based on the interactive information.

[0098] The following is a unified description of S33-S36:

[0099] Assume that the input feature map of SPPF_AA is F(X) ∈ RH×W×C, where H, W, and C represent the height, width, and channel dimensions of the feature map, respectively. To obtain effective local features, these input features are passed to a 1 × 1 convolutional layer for squeezing to obtain F1(X). This is shown in the following equation.

[0100] F1 (X) = Conv1×1 (F(X)) ∈ RH×W×C

[0101] Where Conv1×1 (·) represents a 1 × 1 convolution.

[0102] Furthermore, the squeezed feature map F1 (X) is propagated layer by layer. Considering that information flow is reused between layers during the propagation process, a residual structure is introduced, and multiple max pooling operations are used to reduce the use of irrelevant information. The three max pooling operations are shown in the following equations.

[0103] F2 (X) = MaxPool(F1 (X))

[0104] F3 (X) = MaxPool(F1 (X), F2 (X))

[0105] F4 (X) = MaxPool(F1 (X), F2 (X), F3 (X))

[0106] Among them, F2 (X), F3 (X), F4 (X) represent the output features of different maximum pooling layers.

[0107] MaxPool(·) represents the maximum pooling operation.

[0108] To capture global and contextual semantic information, global average pooling and global max pooling layers are used to incorporate global background information and object edge details, helping the network make better judgments. This also fully captures global perspective information and mitigates the impact of aircraft of varying sizes. The fused output feature F(X) is shown in the following equation.

[0109] F(X) = Concat[F2 (X), F3 (X), F4 (X), F1 (X), Adaptive_AvgPool(F1(X)), Adaptive_MaxPool(F1 (X))]

[0110] Where Adaptive_AvgPool(·) and Adaptive_MaxPool(·) represent the adaptive average pooling operation and the adaptive maximum pooling operation respectively. Then, the fused features are stimulated, that is, the input convolution layer is stimulated.

[0111] S37: When the accuracy of the initial model's regression of the aircraft frame and the classification result of the aircraft target reaches a first preset threshold, the initial model is verified using the model verification set.

[0112] The improved YoloV8 network is trained using the training set data. During training, the model iteratively learns the mapping relationship between image features and labels, gradually optimizing model parameters. When the initial model's accuracy in regressing the aircraft bounding box and classifying aircraft objects reaches a preset threshold (for example, 98%), the model is validated using the validation set data. Based on the validation results, the training strategy is adjusted (e.g., early stopping training, adjusting the learning rate, etc.).

[0113] S38. When the initial model is verified, the SAR remote sensing images in the model test set are input into the verified initial model to test the model.

[0114] The trained YoloV8 network is tested using the test set data. During the test, the model predicts each image in the test set and outputs information such as the location and category of the aircraft target.

[0115] Furthermore, based on the test results, the model's performance indicators (such as precision, recall, F1 score, etc.) are calculated. These indicators are used to evaluate the model's ability to detect aircraft in SAR remote sensing images.

[0116] S39. When the test result shows that the detection accuracy of the initial model for the model test set reaches a second preset threshold, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained.

[0117] Model training is considered complete when the test results meet the preset performance criteria (e.g., accuracy exceeding a second preset threshold of 99%). The resulting improved YoloV8 network is now the SAR remote sensing aircraft detection model. This model can be used in subsequent aircraft detection tasks to automatically identify and locate aircraft in SAR remote sensing imagery.

[0118] The experimental results of aircraft detection in SAR remote sensing images are shown in Table 1, and the visualization demonstration is shown in Figure 6 shown.

[0119] Table 1 Experimental results of aircraft detection in SAR remote sensing images

[0120]

[0121] The training method for a SAR remote sensing aircraft detection model provided by the present invention performs a series of enhancement processes on aircraft targets in SAR images, enhancing the distinction between aircraft and background and further reducing ambiguity between similar aircraft targets. Building on the SPPF module, the method utilizes global average pooling and global maximum pooling layers to incorporate global background information and target edge details, helping the network make better judgments. This method also fully captures global perspective information and mitigates the impact of aircraft of varying sizes.

[0122] The following describes an aircraft target detection device for SAR remote sensing images provided by the present invention. The aircraft target detection device for SAR remote sensing images described below and the aircraft target detection method for SAR remote sensing images described above can be referenced to each other.

[0123] Figure 4 The present invention provides a schematic structural diagram of an aircraft target detection device in SAR remote sensing images, which specifically includes:

[0124] The acquisition module 401 is used to acquire a SAR remote sensing image of a target to be detected. Detailed descriptions can be found in the corresponding descriptions of the above method embodiments, which will not be repeated here.

[0125] Feature extraction module 402 is configured to input the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model and extract multi-scale local features of the target aircraft in the target SAR remote sensing image using the SAR remote sensing aircraft detection model. For detailed descriptions, please refer to the corresponding descriptions of the above method embodiments and will not be repeated here.

[0126] The feature aggregation module 403 is used to perform feature aggregation on the target multi-scale local features to obtain target interaction information between the local and global information of the target aircraft. Detailed descriptions can be found in the corresponding descriptions of the above method embodiments, which will not be repeated here.

[0127] The detection module 404 is configured to detect an aircraft target in the SAR remote sensing image based on the target interaction information. The detection results include target bounding box regression and aircraft classification. For detailed descriptions, please refer to the corresponding descriptions of the above method embodiments and will not be repeated here.

[0128] Figure 5 An example of a physical structure diagram of an electronic device is shown below. Figure 5 As shown, the electronic device may include: a processor 810, a communications interface 820, a memory 830, and a communications bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other via the communications bus 840. The processor 810 may call logic instructions in the memory 830 to execute a method for detecting aircraft targets in SAR remote sensing images. The method includes: obtaining a SAR remote sensing image of a target to be detected; inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting multi-scale local features of the target aircraft in the target SAR remote sensing image using the SAR remote sensing aircraft detection model; performing feature aggregation on the multi-scale local features to obtain target interaction information between local and global information of the target aircraft; and performing aircraft target detection on the target SAR remote sensing image based on the target interaction information, wherein the detection result includes target box regression and aircraft classification.

[0129] Furthermore, the logic instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0130] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for detecting aircraft targets in SAR remote sensing images provided by the above methods, the method including: obtaining a SAR remote sensing image of a target to be detected; inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting target multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; performing feature aggregation on the target multi-scale local features to obtain target interaction information between the local and global information of the target aircraft; performing aircraft target detection on the target SAR remote sensing image based on the target interaction information, and the detection results include target box regression and aircraft classification.

[0131] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the method for detecting aircraft targets in SAR remote sensing images provided by the above-mentioned methods, the method comprising: obtaining a SAR remote sensing image of a target to be detected; inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting target multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; performing feature aggregation on the target multi-scale local features to obtain target interaction information between the local and global information of the target aircraft; and performing aircraft target detection on the target SAR remote sensing image based on the target interaction information, and the detection results include target box regression and aircraft classification.

[0132] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0133] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for detecting aircraft targets in SAR remote sensing images, characterized in that: include: Acquire SAR remote sensing images of the target to be detected; Inputting the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extracting target multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; Performing feature aggregation on the multi-scale local features of the target to obtain target interaction information between local and global information of the target aircraft; Performing aircraft target detection on the target SAR remote sensing image based on the target interaction information, wherein the detection result includes target frame regression and aircraft classification; The SAR remote sensing aircraft detection model is trained based on the following method: Collect SAR remote sensing images from different types of aircraft and perform image processing; The SAR remote sensing images after image processing are divided into a model training set, a model verification set and a model test set; The improved YoloV8 network is trained using the model training set and the model validation set; The trained YoloV8 network is tested using the model test set. When the test result meets the preset conditions, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained; The improved YoloV8 network is trained by the model training set and the model verification set, including: Acquire multi-scale local features of aircraft in the SAR remote sensing image in the model training set through the improved YoloV8 network; Based on the SAR remote sensing images in the model training set, an improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model; Aggregating the multi-scale local features layer by layer in the initial model to obtain interaction information between local and global information; Based on the interactive information, the rotating detection head is used to regress the aircraft frame and classify the aircraft target; When the accuracy of the initial model's regression of the aircraft frame and the classification result of the aircraft target reaches a first preset threshold, the initial model is verified using the model verification set; Based on the SAR remote sensing image in the model training set, the improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model, including: Squeezing the SAR remote sensing images in the model training set through a convolutional layer; The squeezed feature map is passed through multiple maximum pooling operations and residual structures; Using a global average pooling layer and a global maximum pooling layer to obtain global and contextual semantic information, and fusing the global and contextual semantic information with edge detail information; Based on the information fusion results, global semantics and target edge details are modeled to obtain the initial model.

2. The method according to claim 1, characterized in that The trained YoloV8 network is tested by the model test set. When the test result meets the preset conditions, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained, including: When the initial model is verified, the SAR remote sensing images in the model test set are input into the verified initial model to test the model; When the test result shows that the detection accuracy of the initial model for the model test set reaches a second preset threshold, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained.

3. The method according to claim 1, characterized in that The collecting of SAR remote sensing images of different types of aircraft and performing image processing include: Performing image processing on the SAR remote sensing image to obtain an image-processed SAR remote sensing image, wherein the image processing includes random translation, rotation, color conversion, normalization and flipping processing operations.

4. A device for detecting aircraft targets in SAR remote sensing images, characterized in that: include: An acquisition module is used to acquire SAR remote sensing images of the target to be detected; a feature extraction module, configured to input the target SAR remote sensing image into a pre-trained SAR remote sensing aircraft detection model, and extract multi-scale local features of the target aircraft in the target SAR remote sensing image through the SAR remote sensing aircraft detection model; A feature aggregation module is used to perform feature aggregation on the multi-scale local features of the target to obtain target interaction information between the local and global information of the target aircraft; A detection module, configured to perform aircraft target detection on the target SAR remote sensing image based on the target interaction information, wherein the detection result includes target frame regression and aircraft classification; The SAR remote sensing aircraft detection model is trained based on the following method: Collect SAR remote sensing images from different types of aircraft and perform image processing; The SAR remote sensing images after image processing are divided into a model training set, a model verification set and a model test set; The improved YoloV8 network is trained using the model training set and the model validation set; The trained YoloV8 network is tested using the model test set. When the test result meets the preset conditions, it is determined that the improved YoloV8 network training is completed, and the SAR remote sensing aircraft detection model is obtained; The improved YoloV8 network is trained by the model training set and the model verification set, including: Acquire multi-scale local features of aircraft in the SAR remote sensing image in the model training set through the improved YoloV8 network; Based on the SAR remote sensing images in the model training set, an improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model; Aggregating the multi-scale local features layer by layer in the initial model to obtain interaction information between local and global information; Based on the interactive information, the rotating detection head is used to regress the aircraft frame and classify the aircraft target; When the accuracy of the initial model's regression of the aircraft frame and the classification result of the aircraft target reaches a first preset threshold, the initial model is verified using the model verification set; Based on the SAR remote sensing image in the model training set, the improved SPPF_AA component is used in the improved YoloV8 network to model global semantics and target edge details to obtain an initial model, including: Squeezing the SAR remote sensing images in the model training set through a convolutional layer; The squeezed feature map is passed through multiple maximum pooling operations and residual structures; Using a global average pooling layer and a global maximum pooling layer to obtain global and contextual semantic information, and fusing the global and contextual semantic information with edge detail information; Based on the information fusion results, global semantics and target edge details are modeled to obtain the initial model.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for detecting aircraft targets in SAR remote sensing images according to any one of claims 1 to 3 is implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for detecting aircraft targets in SAR remote sensing images according to any one of claims 1 to 3 is implemented.

7. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for detecting aircraft targets in SAR remote sensing images according to any one of claims 1 to 3 is implemented.

Images