Aircraft target detection method in remote sensing images based on region importance-aware attention

Abstract

The present invention discloses a method for detecting aircraft targets in remote sensing images based on regional importance-aware attention, comprising the following steps: S1: performing image preprocessing on the remote sensing image to obtain a preprocessed remote sensing image; S2: inputting the preprocessed remote sensing image into a feature extraction network to obtain a multi-scale fusion feature map; S3: preprocessing the multi-scale fusion feature map to obtain an initial feature map; S4: inputting the initial feature map into a regional importance-aware attention module to obtain an enhanced feature map; S5: inputting the enhanced feature map into a target detection network to obtain an aircraft target detection result. By introducing a regional importance-aware attention mechanism, the present invention can intelligently identify and enhance key features closely related to aircraft targets in remote sensing images. Furthermore, the regional importance-aware attention mechanism introduced by the present invention greatly enhances the robustness of the model.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, and more particularly to a method for detecting aircraft targets in remote sensing images based on regional importance perception attention. Background Art

[0002] With the rapid development of remote sensing technology, remote sensing imagery has been widely used in fields such as geographic information monitoring, environmental protection, resource management, disaster assessment, and military reconnaissance. Among these applications, target detection in remote sensing images, especially aircraft targets, is of great practical significance. Aircraft targets in remote sensing images are typically small, highly deformed, have low contrast, and are located against complex backgrounds. Consequently, traditional target detection methods face numerous challenges, such as blurred target features, significant background interference, and complex target scale variations.

[0003] Traditional target detection methods in remote sensing images rely heavily on manual feature extraction and classic machine learning models. Early detection methods used techniques such as edge detection, region growing, and morphological processing to extract target features from images. These methods then identified targets using classification algorithms such as support vector machines (SVMs) and K-nearest neighbor (K-NN). However, these methods often rely on manually designed features and have poor adaptability to image quality and environmental conditions, performing poorly in complex backgrounds and with targets of varying scales.

[0004] With the rise of deep learning technology, convolutional neural networks (CNNs) have been widely used in remote sensing image target detection, especially for the detection of aircraft targets. Deep learning methods can automatically learn high-level features from raw images, avoiding the tedious manual feature extraction. However, existing deep learning-based target detection methods still have some problems, such as: 1) Background interference problem. In remote sensing images, aircraft targets are often mixed with complex backgrounds (such as clouds, ground buildings, etc.). Existing detection methods are difficult to effectively suppress background information, resulting in serious false detection and missed detection. 2) Scale change problem. The sizes of targets in remote sensing images vary greatly. Aircraft targets from the ground to the air may appear at different sizes in the image. When dealing with targets of different scales, existing methods may have poor detection effects on small targets due to the limitations of the receptive field.

[0005] To address these issues, researchers have recently proposed a variety of optimization methods, such as multi-scale feature fusion, attention mechanisms, and region proposal networks, aiming to improve the accuracy and efficiency of object detection. The introduction of attention mechanisms, in particular, has become a key technology for enhancing object detection. By simulating the selective attention of human vision, attention mechanisms empower the network to focus more on key information and suppress less important information, thereby improving model performance.

[0006] However, existing attention mechanisms primarily focus on weighting global or local information, while assigning relatively simple weights to the importance of different regions. In remote sensing images, the complexity of the background and the diversity of objects make it difficult for traditional attention mechanisms to accurately identify aircraft targets and distinguish them from the background. Therefore, how to better process objects in remote sensing images through regional importance perception to improve object detection performance is a pressing issue for those skilled in the art. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a method for detecting aircraft targets in remote sensing images based on region importance-aware attention.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] In a first aspect, a method for detecting aircraft targets in remote sensing images based on region importance-aware attention is provided, comprising the following steps:

[0010] S1: performing image preprocessing on the remote sensing image to obtain a preprocessed remote sensing image;

[0011] S2: Inputting the preprocessed remote sensing image into a feature extraction network to obtain a multi-scale fusion feature map;

[0012] S3: Preprocessing the multi-scale fusion feature map to obtain an initial feature map;

[0013] S4: Inputting the initial feature map into a region importance-aware attention module to obtain an enhanced feature map;

[0014] S5: Input the enhanced feature map into the target detection network to obtain the aircraft target detection result.

[0015] Preferably, the image preprocessing in S1 includes normalization, contrast enhancement, brightness adjustment, denoising, cropping or scaling.

[0016] Preferably, S2 specifically includes:

[0017] S21: Inputting the pre-processed remote sensing image into a pre-trained improved ResNet-50 network;

[0018] S22: Extract the feature maps output by the pre-trained improved ResNet-50 network at different stages, upsample the high-resolution feature maps, and then splice or weighted fuse them with the low-resolution feature maps to obtain the multi-scale fused feature maps.

[0019] Preferably, the pre-trained improved ResNet-50 network removes the fully connected layer, and uses the second stage, the third stage, the fourth stage and the fifth stage as the output layer for feature extraction.

[0020] Preferably, the preprocessing in S3 includes adjusting the feature map size and normalizing the feature map.

[0021] Preferably, the region importance perception attention module includes a 1*1 convolution layer, a first matrix multiplication unit, an exponential operation unit, a global average pooling layer, a second matrix multiplication unit, an inverse operation unit, a first 3*3 convolution layer, a second 3*3 convolution layer, a first activation function layer, a bilinear transformation layer, a second activation function layer and a third matrix multiplication unit;

[0022] The input end of the 1*1 convolutional layer is used to input the initial feature map;

[0023] The output end of the 1*1 convolutional layer is connected to the input end of the first matrix multiplication unit and the input end of the exponential operation unit respectively;

[0024] The output end of the exponential operation unit is connected to the input end of the first matrix multiplication unit and the input end of the global average pooling layer respectively;

[0025] An output end of the first matrix multiplication unit is connected to an input end of the global average pooling layer;

[0026] The output end of the global average pooling layer is connected to the input end of the second matrix multiplication unit and the input end of the inverse operation unit respectively;

[0027] An output terminal of the inverse operation unit is connected to an input terminal of the second matrix multiplication unit;

[0028] The output end of the second matrix multiplication unit is connected to the input end of the third matrix multiplication unit through the first 3*3 convolution layer, the second 3*3 convolution layer, the first activation function layer, and the bilinear transformation layer in sequence;

[0029] The input end of the second activation function layer is used to input the first channel feature map in the initial feature map;

[0030] The output end of the second activation function layer is connected to the input end of the third matrix multiplication unit;

[0031] The input end of the third matrix multiplication unit is also used to input the initial feature map.

[0032] Preferably, S5 specifically includes: performing target frame regression and classification on the enhanced feature map through the detection head of the target detection network, then removing redundant frames in combination with non-maximum suppression, and finally outputting an aircraft target detection result with category and confidence.

[0033] In a second aspect, an electronic device is provided, 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 remote sensing images based on regional importance perception attention as described in any one of the above items is implemented.

[0034] In a third aspect, a non-transitory computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for detecting aircraft targets in remote sensing images based on regional importance perception attention as described in any one of the above items is implemented.

[0035] In a fourth aspect, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the method for detecting aircraft targets in remote sensing images based on regional importance perception attention as described in any one of the above items.

[0036] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method for detecting aircraft targets in remote sensing images based on regional importance perception attention, which can achieve the following beneficial effects:

[0037] 1) This invention introduces a region-importance-aware attention mechanism to intelligently identify and enhance key features in images closely related to aircraft targets. This mechanism is particularly suitable for complex scenes, including remote sensing images with low contrast, low resolution, or large variations in target scale. By adaptively adjusting the focus on features, this invention effectively highlights target features while suppressing irrelevant background information, significantly improving target detection accuracy in these challenging environments.

[0038] 2) The region-importance-aware attention mechanism also significantly enhances the model's robustness. It maintains stable performance across varying environmental conditions and image quality, reducing the likelihood of false and missed detections. This robustness is particularly important for practical applications, as it ensures that the system can reliably perform object detection tasks even under less-than-ideal observation conditions, providing users with accurate and reliable results. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0040] Figure 1 A flowchart of a method for detecting aircraft targets in remote sensing images based on region importance perception attention provided in an embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of the structure of a region importance perception attention module provided in an embodiment of the present invention;

[0042] Figure 3 The figure is a schematic structural diagram of an electronic device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] like Figure 1 As shown, in a first aspect, an embodiment of the present invention discloses a method for detecting aircraft targets in remote sensing images based on region importance perception attention, comprising the following steps:

[0045] S1: performing image preprocessing on the remote sensing image to obtain a preprocessed remote sensing image;

[0046] Furthermore, the image preprocessing in S1 includes normalization, contrast enhancement, brightness adjustment, denoising, cropping or scaling.

[0047] It is understood that before extracting the initial feature map, the input remote sensing image is preprocessed to ensure that the image quality is suitable for subsequent feature extraction. First, the pixel values ​​of the input remote sensing image are normalized to the range [0, 1] to eliminate the impact of factors such as uneven lighting on image quality. Then, based on the characteristics of remote sensing images, contrast enhancement, brightness adjustment, or denoising can be used to enhance key information in the image and improve feature extraction. Finally, the image is cropped or scaled to a size suitable for the input size of the feature extraction network to enter the subsequent processing stages.

[0048] S2: Inputting the preprocessed remote sensing image into a feature extraction network to obtain a multi-scale fusion feature map;

[0049] Furthermore, S2 specifically includes:

[0050] S21: Inputting the pre-processed remote sensing image into a pre-trained improved ResNet-50 network;

[0051] S22: Extract the feature maps output by the pre-trained improved ResNet-50 network at different stages, upsample the high-resolution feature maps, and then splice or weighted fuse them with the low-resolution feature maps to obtain the multi-scale fused feature maps.

[0052] Furthermore, the pre-trained improved ResNet-50 network removes the fully connected layer, and uses the second stage, the third stage, the fourth stage and the fifth stage as the output layer for feature extraction.

[0053] It is understandable that: in the feature extraction stage: first, you need to select a suitable deep learning network. Usually, a convolutional neural network (CNN) pre-trained on a large-scale dataset (such as ImageNet) is selected as the feature extraction network, such as ResNet, VGG or DenseNet.

[0054] To ensure that the extracted features contain rich semantic information, this paper selects ResNet-50 as the feature extraction network and uses the second, third, fourth, and fifth stages of ResNet-50 as the output layers of feature extraction. The second stage (conv2_x) can extract relatively simple texture and edge information; the third stage (conv3_) can extract higher-level semantic information while still retaining certain spatial details, which is helpful for identifying aircraft targets in remote sensing images; the fourth and fifth stages (conv4_x and conv5_x) can extract more complex semantic information and abstract features.

[0055] That is, in order to adapt to the target detection task, the present invention removes the fully connected layer in ResNet-50, retains the feature maps output by the convolutional layers of multiple stages (second stage, third stage, fourth stage and fifth stage), and uses them as the basic features for subsequent processing.

[0056] Multi-scale feature fusion: Taking into account the different scales of aircraft targets in remote sensing images, a multi-scale feature fusion strategy is adopted to enhance the model's ability to detect targets of different scales. The present invention can capture multi-level feature information from low-level to high-level by extracting feature maps from multiple different layers of ResNet-50 (such as conv2_x, conv3_x, conv4_x, and conv5_x). On this basis, the high-resolution feature map is upsampled and spliced ​​or weighted fused with the low-resolution feature map to obtain a multi-scale feature map containing information of different scales, ensuring that the model can handle aircraft targets of various sizes in the image.

[0057] The present invention specifically upsamples the feature map output by the second stage (conv2_x) and concatenates or weightedly fuses it with the feature map output by the third stage (conv3_x);

[0058] The feature map output by the second stage (conv2_x) is upsampled and concatenated or weighted fused with the feature map output by the fourth stage (conv4_x);

[0059] The feature map output by the second stage (conv2_x) is upsampled and concatenated or weighted fused with the feature map output by the fifth stage (conv5_x);

[0060] The feature map output by the third stage (conv3_x) is upsampled and concatenated or weighted fused with the feature map output by the fourth stage (conv4_x);

[0061] The feature map output by the third stage (conv3_x) is upsampled and concatenated or weighted fused with the feature map output by the fifth stage (conv5_x);

[0062] S3: Preprocess the multi-scale fusion feature map to obtain the initial feature map φ o ;

[0063] Furthermore, the preprocessing in S3 includes adjusting the feature map size and feature map normalization.

[0064] It is understandable that after completing the multi-scale feature fusion, the present invention adjusts the size of the multi-scale fusion feature map to ensure that it meets the input requirements of the subsequent network. The present invention appropriately adjusts the multi-scale fusion feature map through upsampling or downsampling operations to ensure that its size and resolution are adapted. Then, these multi-scale fusion feature maps are normalized to stabilize the network training process and ensure that the information of each channel of the multi-scale fusion feature map is balanced. Finally, these feature maps that have undergone multi-scale fusion and normalization are defined as the initial feature map φ o, providing basic features for subsequent region importance perception and target detection tasks.

[0065] S4: The initial feature map φ o Input to the region importance-aware attention module to obtain the enhanced feature map φ′;

[0066] Further, such as Figure 2 As shown, the region importance perception attention module includes a 1*1 convolution layer, a first matrix multiplication unit, an exponential operation unit, a global average pooling layer, a second matrix multiplication unit, an inverse operation unit, a first 3*3 convolution layer, a second 3*3 convolution layer, a first activation function layer, a bilinear transformation layer, a second activation function layer and a third matrix multiplication unit;

[0067] The input end of the 1*1 convolutional layer is used to input the initial feature map;

[0068] The output end of the 1*1 convolutional layer is connected to the input end of the first matrix multiplication unit and the input end of the exponential operation unit respectively;

[0069] The output end of the exponential operation unit is connected to the input end of the first matrix multiplication unit and the input end of the global average pooling layer respectively;

[0070] An output end of the first matrix multiplication unit is connected to an input end of the global average pooling layer;

[0071] The output end of the global average pooling layer is connected to the input end of the second matrix multiplication unit and the input end of the inverse operation unit respectively;

[0072] An output terminal of the inverse operation unit is connected to an input terminal of the second matrix multiplication unit;

[0073] The output end of the second matrix multiplication unit is connected to the input end of the third matrix multiplication unit through the first 3*3 convolution layer, the second 3*3 convolution layer, the first activation function layer, and the bilinear transformation layer in sequence;

[0074] The input end of the second activation function layer is used to input the first channel feature map in the initial feature map;

[0075] The output end of the second activation function layer is connected to the input end of the third matrix multiplication unit;

[0076] The input end of the third matrix multiplication unit is also used to input the initial feature map.

[0077] It is understandable that:

[0078] For gating mechanism initialization and feature mapping:

[0079] The initial feature map φ obtained o , different from the existing gate units that apply additional networks, for simplicity, the present invention selects the initial feature map φ o The first channel feature map φ[0] in is mapped as the gate, namely Sigmoid(φ[0]). In addition, φ o Through 1×1 convolution operation, the input initial feature map φ is realized o Adjust the channel dimension to obtain new features

[0080] feature Instantiate computation by applying exponential operations, global average pooling, and inverse operations (adjusting the width and compressed convolutions in these operations to reduce computation and expand the receptive field) Importance values ​​within surround R:

[0081]

[0082] in, Indicated in the feature The weight or importance measure at . R represents the set of neighborhood regions. R k is a specific area in R, k represents the index in the area R, which is used to traverse all areas in the neighborhood centered on pixel x. i and j are R k The pixel index within is used to iterate over all pixels in the area. and Represents region R k The eigenvalues ​​of pixels i and j. k is with area R k The associated learnable weight is used to adjust the importance of the region. This formula is used to calculate the feature The local importance of each region in its neighborhood is calculated and integrated through weighted summation. In the context of aircraft target detection in remote sensing images, this metric can help the model identify and emphasize features related to aircraft targets in the image, thereby improving detection accuracy.

[0083] Next, two 3*3 convolutional layers, sigmoid activation functions, and bilinear transformation layers are used for activation and rescaling to obtain the regional importance-aware attention A:

[0084]

[0085] Among them, Bilinear represents the bilinear transformation layer, and sigmoid represents the sigmoid activation function. This step ensures that the output of the attention mechanism has both nonlinear characteristics and can adapt to different feature scales.

[0086] Apply region importance-aware attention to features φ[0] and φ o , and get the enhanced feature map φ′:

[0087]

[0088] in, represents a matrix multiplication operation. This process not only enhances important features related to target detection but also suppresses unimportant background information. The enhanced feature map φ′ obtained through this operation enables the model to focus more on aircraft targets in remote sensing images, thereby improving the accuracy and robustness of target detection.

[0089] S5: Input the enhanced feature map into the target detection network to obtain the aircraft target detection result.

[0090] Furthermore, S5 specifically includes: performing target box regression and classification on the enhanced feature map through the detection head of the target detection network, and then removing redundant boxes in combination with non-maximum suppression, and finally outputting the aircraft target detection result with category and confidence.

[0091] It is understandable that after obtaining the enhanced feature map φ′, a dedicated object detection network is required to generate the object detection output for object detection. This process typically includes the following steps: object bounding box regression and classification by the detection head, object localization and classification using the enhanced feature map φ′, and non-maximum suppression to remove redundant bounding boxes, ultimately obtaining the detection result.

[0092] Specifically:

[0093] 1) Object Detection Head for Bounding Box Regression and Classification: After obtaining the enhanced feature map φ′, the detection head of the object detection network is used to perform bounding box regression and object classification. The detection head typically consists of convolutional layers and fully connected layers. Its primary task is to generate candidate object bounding boxes and corresponding class probabilities for each position in the enhanced feature map φ′.

[0094] In the target box regression task, the target detection network will generate the coordinates of the candidate box (such as the coordinates of the upper left corner and lower right corner of the bounding box). Usually, the target detection network will output 4 regression values ​​(x min ,y min , x max ,y max ), where (xmin ,y min , x max ,y max ) represents the coordinates of the candidate box). Input enhanced feature map (Where H and W are the height and width of the enhanced feature map, and C is the number of channels) Through a series of convolution operations, the target detection network outputs 4 channels as the result of target box regression:

[0095]

[0096] The classification task is to assign a category label to each candidate box. The number of categories is N. Each candidate box will output an N-dimensional probability distribution through the fully connected layer, and the Softmax function will convert it into a category probability:

[0097]

[0098] in, It is the category score output by the target detection network. The Softmax function can convert the output into probability so that the sum of the probabilities of all categories is 1.

[0099] 2) Generate object detection results: Based on the regressed bounding box and classification probability, the object detection network assigns a category to each candidate box and calculates a confidence score. Each candidate box will have a confidence score, usually expressed as:

[0100]

[0101] in, is the confidence of the candidate box, is the predicted bounding box, is the intersection over union (IoU) value with the true target, P class is the class probability of the box.

[0102] 3) After obtaining multiple candidate frames, it is usually necessary to use non-maximum suppression (NMS) to remove redundant detection frames. First, calculate the intersection over union (IoU) between each pair of candidate frames. and Calculate their IoU values. Then remove the overlapping boxes, and for the candidate boxes and If their IoU exceeds a certain threshold (such as 0.5), the box with higher confidence is retained and the box with lower confidence is suppressed (removed). The commonly used strategy is to sort by confidence from high to low, select the box with the highest confidence, and remove the box with the highest overlap. Finally, after the detection box output is processed by NMS, a set of non-redundant target boxes is retained, each box has a corresponding category and confidence score.

[0103] 4) Finally, the target detection results are output. After regression, classification, and NMS, the network outputs the target's bounding box and its category. Each bounding box contains the following information:

[0104]

[0105] in, are the coordinates of the detection box, is the probability of the category to which the box belongs. The final output is the set of these bounding boxes and their categories, representing all aircraft targets detected in the remote sensing image.

[0106] In a second aspect, an embodiment of the present invention provides an electronic device, such as Figure 3 As shown, the electronic device may include: a processor 301, a communications interface 302, a memory 303, and a communication bus 304. The processor 301, the communications interface 302, and the memory 303 communicate with each other via the communication bus 304. The processor 301 may call logic instructions in the memory 303 to execute a method for detecting aircraft targets in remote sensing images based on regional importance perception attention.

[0107] In addition, the logic instructions in the above-mentioned memory 303 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0108] In a third aspect, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a remote sensing image aircraft target detection method based on regional importance perception attention provided by the above methods.

[0109] In a fourth aspect, 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 a remote sensing image aircraft target detection method based on regional importance perception attention provided by the above methods.

[0110] 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.

[0111] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0113] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein.

[0114] Rather, it is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting aircraft targets in remote sensing images based on region importance perception attention, characterized in that: The following steps are involved: S1: performing image preprocessing on the remote sensing image to obtain a preprocessed remote sensing image; S2: Inputting the preprocessed remote sensing image into a feature extraction network to obtain a multi-scale fusion feature map; S3: Preprocessing the multi-scale fusion feature map to obtain an initial feature map; S4: Inputting the initial feature map into a region importance-aware attention module to obtain an enhanced feature map; The region importance perception attention module includes a 1*1 convolution layer, a first matrix multiplication unit, an exponential operation unit, a global average pooling layer, a second matrix multiplication unit, an inverse operation unit, a first 3*3 convolution layer, a second 3*3 convolution layer, a first activation function layer, a bilinear transformation layer, a second activation function layer and a third matrix multiplication unit; The input end of the 1*1 convolutional layer is used to input the initial feature map; The output end of the 1*1 convolutional layer is connected to the input end of the first matrix multiplication unit and the input end of the exponential operation unit respectively; The output end of the exponential operation unit is connected to the input end of the first matrix multiplication unit and the input end of the global average pooling layer respectively; An output end of the first matrix multiplication unit is connected to an input end of the global average pooling layer; The output end of the global average pooling layer is connected to the input end of the second matrix multiplication unit and the input end of the inverse operation unit respectively; An output terminal of the inverse operation unit is connected to an input terminal of the second matrix multiplication unit; The output end of the second matrix multiplication unit is connected to the input end of the third matrix multiplication unit through the first 3*3 convolution layer, the second 3*3 convolution layer, the first activation function layer, and the bilinear transformation layer in sequence; The input end of the second activation function layer is used to input the first channel feature map in the initial feature map; The output end of the second activation function layer is connected to the input end of the third matrix multiplication unit; The input end of the third matrix multiplication unit is also used to input the initial feature map; S5: Input the enhanced feature map into the target detection network to obtain the aircraft target detection result.

2. The method for detecting aircraft targets in remote sensing images based on region importance perception attention according to claim 1, characterized in that: Image preprocessing in S1 includes normalization, contrast enhancement, brightness adjustment, denoising, cropping or scaling.

3. The method for detecting aircraft targets in remote sensing images based on region importance perception attention according to claim 1, characterized in that: S2 specifically includes: S21: input the preprocessed remote sensing image into the pre-trained improved ResNet-50 network; S22: extract the feature maps output by the pre-trained improved ResNet-50 network at different stages, and upsample the high-resolution feature maps and splice or weighted fuse them with the low-resolution feature maps to obtain the multi-scale fused feature maps.

4. The method for detecting aircraft targets in remote sensing images based on region importance perception attention according to claim 3, characterized in that: The pre-trained improved ResNet-50 network removes the fully connected layer and uses the second, third, fourth and fifth stages as output layers for feature extraction.

5. The method for detecting aircraft targets in remote sensing images based on region importance perception attention according to claim 1, characterized in that: The preprocessing in S3 includes feature map resizing and feature map normalization.

6. The method for detecting aircraft targets in remote sensing images based on region importance perception attention according to claim 1, characterized in that: S5 specifically includes: performing target frame regression and classification on the enhanced feature map through the detection head of the target detection network, then removing redundant frames in combination with non-maximum suppression, and finally outputting the aircraft target detection result with category and confidence.

7. 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 remote sensing images based on region importance-aware attention is implemented as described in any one of claims 1 to 6.

8. 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 remote sensing images based on region importance-aware attention is implemented as claimed in any one of claims 1 to 6.

9. 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 remote sensing images based on region importance-aware attention is implemented as claimed in any one of claims 1 to 6.

Images