Remote sensing image target detection method and system

Abstract

The present invention relates to the technical field of remote sensing image processing, and specifically to a remote sensing image target detection method and system. After acquiring the remote sensing image to be detected, the method inputs it into a trained target detection model. In the model, the data processing module performs feature processing on the remote sensing image to obtain a multi-scale feature map. The adaptive occlusion perception module performs occlusion processing on the multi-scale feature map to obtain an image occlusion mask map, and then optimizes the features based on the occlusion mask map. The mapping module performs feature mapping to obtain mapping features, and finally obtains the occlusion area detection result based on the mapping features. This model can effectively perform target detection on remote sensing images through the collaboration of multiple modules, improve the accuracy and reliability of detection, and provide strong support for remote sensing image analysis and application. This method solves the problem that the existing target detection method has a fast detection speed and can handle target occlusion well.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image processing, and in particular to a remote sensing image target detection method and system. Background Art

[0002] Current object detection methods can be roughly divided into two categories: anchor-based detection methods and anchor-free detection methods. Anchor-based detection methods primarily include Faster R-CNN. This method primarily involves extracting feature maps from the input image using a convolutional neural network (CNN). Next, the feature maps are fed into a Region Proposal Network (RPN) to generate region proposals. The RPN uses a sliding window to generate multiple anchor boxes and uses classification and regression branches to predict the class and location of each anchor box. The generated region proposals undergo non-maximum suppression (NMS) to identify the optimal regions. These region proposals are then passed to the Fast R-CNN network for accurate classification and bounding box regression to determine the final object class and location. This method uses the RPN to generate region proposals. The RPN uses a sliding window to generate multi-scale and multi-ratio anchor boxes at each pixel in the feature map. However, most of these anchor boxes have little impact on the detection results, significantly increasing the computational load of the entire network and affecting detection speed. FCOS (Fully Convolutional One-Stage Object Detection) is an anchor-free object detection algorithm. It makes predictions directly on feature maps, without relying on predefined anchor boxes. The algorithm first extracts feature maps using a convolutional neural network. It then predicts the target category, center, and bounding box regression parameters at each pixel. This information is then used to generate the final target detection box. While this method does not rely on predefined anchor boxes, significantly reducing computational complexity and accelerating training and inference, in some cases, particularly when dealing with complex backgrounds or scenes with densely packed targets, target localization accuracy may be inferior to methods based on anchor boxes.

[0003] In summary, anchor-based methods require the generation and processing of a large number of anchor boxes, which may lead to high training and inference costs on high-resolution images and large-scale datasets. Anchor-free methods cannot effectively handle object occlusion. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a remote sensing image target detection method and system in response to the above-mentioned deficiencies in the prior art, so as to solve the technical problems that the existing target detection methods have slow detection speed and cannot handle target occlusion well.

[0005] The purpose of the present invention is achieved by the following technical solutions:

[0006] In a first aspect, the present invention provides a method for detecting targets in remote sensing images, comprising:

[0007] Acquire remote sensing images to be detected;

[0008] Inputting the remote sensing image to be detected into the trained target detection model, performing target detection on the remote sensing image, and obtaining a detection result;

[0009] The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; the features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain the detection result of the occluded area according to the mapping features, obtain the detection result of the non-occluded area according to the features in the multi-scale feature map, and finally obtain the detection result according to the occluded area detection result and the non-occluded area detection result.

[0010] As a further improvement of the present invention, the backbone network module adopts a ResNet network; the backbone network module includes a backbone sub-module and a feature pyramid sub-module, the backbone sub-module performs feature extraction on the input remote sensing image to obtain a first feature map; the feature pyramid sub-module is used to perform convolution processing on the features in the first feature map to obtain a multi-scale feature map.

[0011] As a further improvement of the present invention, the adaptive occlusion perception module is used to perform convolution processing on the input multi-scale feature map, and then process it using a softmax function to obtain a first result map, in which the pixel positions greater than a first set threshold in the first result map are set to 1, and the remaining pixel positions are set to 0, thereby obtaining an image occlusion mask map;

[0012] As a further improvement of the present invention, the mapping module is used to perform dot product processing on the feature map and the image occlusion mask map, hide the area where there is no occlusion, add the corresponding dot product result and the feature map, and input it into the feature mapping layer to obtain the mapping feature;

[0013] The mapping module is further provided with an occlusion detection module and a non-occlusion detection module; the occlusion detection module is used to convolve the mapping features to obtain a classification and detection box offset prediction result, perform positive sample determination based on the prediction result, and obtain a detection result of the occlusion area; the detection box is a preset anchor box, which is used to sum with the offset result to obtain the corresponding detection result;

[0014] The non-occlusion detection module is used to obtain the prediction results of classification, detection frame and center, perform positive sample determination based on the prediction results, and obtain the detection results of the non-occlusion area.

[0015] As a further improvement of the present invention, a detection module is further connected to the occlusion detection module and the non-occlusion detection module; the detection module respectively accepts positive samples of all occlusion detection modules and non-occlusion detection modules, and after scale transformation, performs non-maximum suppression processing on the positive samples to obtain occlusion detection results and non-occlusion detection results; for the occlusion detection results, the IoU is calculated with the non-occlusion detection results, and the occlusion detection results with IoU greater than 0.7 are retained. Finally, the retained occlusion detection results and the non-occlusion detection results are integrated as the final detection results.

[0016] As a further improvement of the present invention, the remote sensing image data is preprocessed, and the preprocessing step includes standardizing and normalizing the remote sensing image data.

[0017] As a further improvement of the present invention, after the remote sensing image data is pre-processed, the remote sensing image data is further enhanced, specifically including horizontal flipping and random cropping:

[0018] The horizontal flipping specifically includes: flipping the remote sensing image along the vertical center, making the left and right mirror symmetrical, and mirror-reversing the annotation box corresponding to the remote sensing image;

[0019] The random cropping specifically includes: randomly cropping a sub-region in the original remote sensing image as a training sample, detecting whether the corresponding annotation box exceeds the image boundary through the training sample, and if so, aligning the annotation box to the image edge.

[0020] As a further improvement of the present invention, the target detection model sets different learning rates and weight decay parameters during the training process, sets the number of training rounds and hyperparameters;

[0021] Adamw is used as the model optimizer and poly is used as the learning rate strategy; the DIOR remote sensing dataset is used to iteratively train the target inspection model;

[0022] According to the value of the loss function, the corresponding parameters of the target detection model are updated through back propagation until the optimal model parameters are obtained.

[0023] As a further improvement of the present invention, the loss function is:

[0024]

[0025] Where L is the loss function, represents the classification loss, represents the regression loss.

[0026] In a second aspect, the present invention provides a remote sensing image target detection system for implementing the above-mentioned remote sensing image target detection method, comprising:

[0027] A data acquisition unit, used for acquiring remote sensing images to be detected;

[0028] The target detection unit is used to input the remote sensing image to be detected into the trained target detection model, perform target detection on the remote sensing image, and obtain a detection result;

[0029] The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; the features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain the detection result of the occluded area according to the mapping features, obtain the detection result of the non-occluded area according to the features in the multi-scale feature map, and finally obtain the detection result according to the occluded area detection result and the non-occluded area detection result.

[0030] In a third aspect, the present invention provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions that, when executed by a computing device, enable the computing device to perform the above-mentioned remote sensing image target detection method.

[0031] In a fourth aspect, the present invention provides a computing device, comprising:

[0032] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing the above-mentioned remote sensing image target detection method.

[0033] The present invention provides a method for remote sensing image target detection that achieves adaptive perception of target occlusion through an adaptive occlusion perception module. By performing occlusion processing on multi-scale feature maps to generate an image occlusion mask, the method effectively perceives occlusion conditions in the image, thereby specifically addressing occlusion issues during feature optimization, reducing false detections and missed detections due to occlusion, and improving target detection accuracy. Furthermore, the adaptive occlusion perception module automatically adjusts occlusion processing and alignment mark processing based on different remote sensing images, adapting to occlusion conditions in different scenarios and improving the model's adaptability. By performing feature mapping and rapidly obtaining detection results through a mapping module, the model reduces computation time, improves target detection efficiency, and meets the requirements of real-time or large-scale remote sensing image detection. The model's multi-module structure enables it to comprehensively leverage the advantages of different modules when processing complex remote sensing image data, enhancing the model's robustness and enabling stable target detection under various environmental conditions. The target detection model of the present invention effectively balances computational complexity with occlusion issues, improving the efficiency and accuracy of model detection.

[0034] Furthermore, the backbone network module utilizes a relatively efficient ResNet network structure, reducing computational effort and the number of parameters while ensuring effective feature extraction. This allows the target detection model to perform computations more quickly when processing remote sensing images, improving detection efficiency. This is particularly true for large-scale remote sensing image datasets, saving significant computational time and resources. The convolution processing within the feature pyramid submodule typically utilizes an efficient convolution algorithm, further improving computational efficiency and enabling the processing of more remote sensing images within a limited timeframe, thus meeting the real-time requirements of practical applications.

[0035] Furthermore, in the adaptive occlusion perception module, the multi-scale feature map is convolved and then the softmax function is used to obtain the first result map, which can effectively classify the features and estimate the probability. By setting a first set threshold and binarizing the pixel position to obtain the image occlusion mask map, the model can accurately identify the occluded area in the image, providing clear guidance for subsequent feature optimization processing. This precise occlusion mask generation helps to reduce target detection errors caused by occlusion and improve detection accuracy. The processing method using convolution and softmax functions can automatically learn the occlusion patterns in different images, adapt to various complex occlusion situations, and improve the generalization ability of the model. Processing the paired boxes whose median value in the IoU matrix is ​​greater than the set value can effectively screen out target pairs that may have occlusion relationships, reducing unnecessary calculations and misjudgments.

[0036] Furthermore, by preprocessing the remote sensing image data input to the target detection model, we can prevent certain features from dominating the model training due to excessively large or small values. This allows the model to learn the importance of each feature more fairly, improving the accuracy and stability of target detection.

[0037] Furthermore, horizontal flipping allows remote sensing images to be presented from different angles, increasing data diversity. This also enables the model to learn more universal features, improving its ability to detect objects from different orientations. After horizontal flipping and random cropping, the model is exposed to a variety of image variations during training, thus becoming more robust to these variations. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] 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 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.

[0039] Figure 1 1 is a flow chart of a method for detecting a target in a remote sensing image according to an embodiment of the present invention;

[0040] Figure 2 Schematic diagram of the structure of the target detection model in an embodiment of the present invention;

[0041] Figure 3 1 is a schematic structural diagram of a remote sensing image target detection system according to an embodiment of the present invention;

[0042] Figure 4 Schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose and technical solution of the present invention clearer and easier to understand, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0044] Explanation of relevant terms:

[0045] Anchor-free: A method in the field of object detection that does not rely on predefined anchor boxes (AnchorBox). It predicts the position and shape of the target directly on the image or feature map instead of the offset relative to the anchor box.

[0046] FPN (Feature Pyramid Network): A feature pyramid network; a network structure used for object detection. It aims to solve the multi-scale problem in object detection, that is, objects of different sizes need to be detected on feature maps of different scales.

[0047] ResNet (Residual Network): Residual network; a deep neural network structure, the core of which is to solve the gradient vanishing and gradient exploding problems in deep network training by introducing residual connections, so that the network can be trained deeper and have stronger performance.

[0048] ASS: Adaptive Occlusion Perception Module;

[0049] DIOR: A large-scale benchmark dataset for object detection in optical remote sensing images.

[0050] Threshold: threshold.

[0051] GIoU (Generalized Intersection over Union): Generalized intersection over union, an indicator used to measure the degree of overlap between two shapes (mainly the predicted box and the real box in target detection).

[0052] GTbox (GroundTruthbox): real box.

[0053] IoU: Intersection over Union function.

[0054] The technical solutions of the present invention will be described clearly and completely below with reference to the accompanying drawings and specific embodiments. The described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0055] Example 1

[0056] like Figure 1 、 Figure 2 As shown, this embodiment provides a remote sensing image target detection method, and the following is a specific implementation method.

[0057] Obtain a remote sensing image to be detected; input the remote sensing image to be detected into a trained target detection model, perform target detection on the remote sensing image, and obtain a detection result.

[0058] The dataset used to train the target detection model in this embodiment is a remote sensing image dataset with three RGB channels. In this embodiment, the DIOR dataset is used as the basic dataset. The DIOR dataset contains 23,463 images and 192,472 instances, covering 20 object classes. The dataset construction process specifically includes:

[0059] The dataset has an image size of 800×800 pixels and a spatial resolution of 0.5m to 30m. The dataset is divided into a training set (11,725 ​​images) and a test set (11,738 images).

[0060] The remote sensing images in the dataset are labeled images. Image labels are classified into 20 categories, and the box annotation type uses horizontal boxes, which are the coordinates of the upper left and lower right corners of the box. The images and annotation files are stored in jpg and xml formats, respectively.

[0061] The remote sensing images in the dataset are preprocessed images. The specific preprocessing steps include standardization and normalization.

[0062] The standardization process specifically includes: scaling the input image data to a range according to the calculated mean and standard deviation so that it satisfies the normal distribution. The standardization formula is expressed as:

[0063]

[0064] Here, μ and σ represent the mean and variance of the dataset, respectively.

[0065] Normalization specifically involves scaling the data so that it falls into a small, specific range. By default, the maximum and minimum normalization is used to scale it to [0, 1]. The normalization formula is:

[0066]

[0067] Where x is the original data, min(x) and max(x) represent the minimum and maximum values ​​of the graph, respectively.

[0068] In addition, in order to highlight local feature learning and increase data diversity, this embodiment also performs data enhancement processing on remote sensing images. The specific processing methods include horizontal flipping and random cropping:

[0069] The horizontal flipping specifically includes: flipping the remote sensing image along the vertical center, making it mirror-symmetrical on the left and right sides, and at the same time, mirror-reversing the annotation box while processing the image symmetrically.

[0070] Random cropping specifically includes: randomly cropping a sub-region in the original remote sensing image as a training sample, detecting whether the corresponding annotation box exceeds the image boundary, and if so, aligning the annotation box to the edge of the image.

[0071] Among them, the target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module.

[0072] The backbone network module processes the input remote sensing imagery to generate a multi-scale feature map. The backbone network module uses the ResNet network as its architecture. It comprises a backbone submodule and a feature pyramid submodule. The backbone submodule extracts features from the input remote sensing imagery to generate a first feature map. The feature pyramid submodule performs convolution processing on the features in the first feature map to generate a multi-scale feature map.

[0073] The backbone submodule is constructed based on the ResNet network. The enhanced image is input into the backbone submodule, and the first feature map is obtained after four stages. The first feature map contains four layers of feature vectors: C1, C2, C3, and C4. The first feature map is a multi-scale feature map.

[0074] Select the last three layers of features C2, C3, and C4 in the first feature map and input them into the pyramid submodule (FPN submodule) to obtain a multi-scale feature map, where the multi-scale feature map includes features P2, P3, P4, and P5. The specific formula is as follows:

[0075]

[0076]

[0077]

[0078] in, is the network convolution layer.

[0079] The adaptive occlusion perception module is used to perform convolution processing on the input multi-scale feature map, and then process it using the softmax function to obtain a first result map. In the first result map, the pixel positions greater than the first set threshold are set to 1, and the remaining pixel positions are set to 0 to obtain an image occlusion mask map;

[0080] The adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; and optimize the features according to the image occlusion mask map;

[0081] Specifically, the multi-scale feature maps P2, P3, P4, and P5 output by the backbone network module are input into the parallel adaptive occlusion perception module (ASS) to obtain the image occlusion mask maps M2, M3, M4, and M5.

[0082] Specifically, the parallel adaptive occlusion perception module uses parameter-independent 1×1 convolution on feature maps of different scales, and then processes them using the softmax function. Finally, the positions in the result map that are greater than the threshold are set to 1, and the other positions are set to 0, thus obtaining the required image occlusion mask map.

[0083] The forward mapping module (FMM) is used to perform dot product processing on the feature map and the image occlusion mask map, hide the area where there is no occlusion, add the corresponding dot product result with the feature map and input it into the feature mapping layer to obtain the mapping feature, and then input the mapping feature into the subsequent module.

[0084] An occlusion detection module and a non-occlusion detection module are also provided after the mapping module. The occlusion detection module is used to convolve the mapping features output by the mapping module to obtain the prediction results of the classification and detection frame offset, and perform positive sample judgment based on the prediction results to obtain the detection results of the occluded area. The non-occlusion detection module is used to obtain the prediction results of the classification, detection frame and center according to the features in the multi-scale feature map, and perform positive sample judgment based on the prediction results to obtain the detection results of the non-occlusion area. Among them, the weights of the occlusion detection module and the non-occlusion detection module are independent. Among them, the detection frame of the occlusion module is a preset anchor frame, which is used to add the offset result to obtain the corresponding detection result; and the non-occlusion module directly obtains the detection result.

[0085] In addition, the occlusion detection module and the non-occlusion detection module are connected to a detection module. The detection module receives all positive samples from the occlusion detection module and the non-occlusion detection module respectively. After scaling, these results are subjected to non-maximum suppression processing to obtain occlusion detection results and non-occlusion detection results. For the occlusion detection results, the Intersection over Union (IoU) is calculated with the non-occlusion detection results. The occlusion results with an IoU greater than 0.7 are retained, and finally all results are integrated as the final detection result.

[0086] Specifically, the feature map layer is composed of two stacked 3×3 convolution and group normalization operations. In this embodiment, the stride of the 3×3 convolution is set to 1, and the element padding is set to 1.

[0087] For the occlusion detection module, this embodiment uses 1×1 convolution to obtain prediction results for classification and detection box offset, where the number of classification result channels is 20, which is the number of categories in the dataset. The number of detection box channels is 4×3, corresponding to three preset anchor boxes with aspect ratios. The anchor boxes are preset to an area of ​​256 pixels and use aspect ratios of [1:2, 1:1, 2:1]. The preset anchor boxes are resized to the corresponding scales, and the prediction results of the anchor boxes and box offsets are added to obtain the prediction results of the occluded object detection box. Finally, the prediction results of the occlusion module are judged as positive samples: if there is a real box (GTbox) with an IoU greater than a set value (for example, 0.7) with the detection box, it is considered a positive sample; otherwise, it is considered a negative sample.

[0088] For the non-occlusion detection module, this embodiment uses parameter-independent 3×3 convolution to obtain the prediction results of classification, box, and center respectively. The stride of the 3×3 convolution is set to 1, and the pad is set to 1. The number of classification result channels is the predefined number of categories; the number of detection box channels is 4; and the number of center channels is 1. Finally, the prediction results of the non-occlusion detection module are judged as positive samples: if the point (x, y) is within the (c x -rs,c y -rs,c x +rs,c y +rs) range, it is considered a positive sample, otherwise it is considered a negative sample. x ,c y ) is the center point of the real box, s is the step size of the feature map relative to the original image, and r is a hyperparameter controlling the distance from the GT center.

[0089] In the process of training the target detection function, this embodiment sets different learning rates and weight decay parameters according to the training results, sets the number of training rounds and hyperparameters, and the hyperparameters include parameters such as the number of categories and threshold.

[0090] Adamw is used as the model optimizer and poly is used as the learning rate strategy; the DIOR remote sensing dataset is used to iteratively train the target inspection model.

[0091] In this example, the number of training rounds is set to 50 epochs, and the formula is as follows:

[0092]

[0093] Among them, base_lr defaults to 6e-5 (i.e. 6×10 -5 ), num_epoch, power defaults to 0.9.

[0094] According to the value of the loss function, the corresponding parameters of the target detection model are updated through back propagation until the optimal model parameters are obtained.

[0095] Save the optimal model parameters. After each epoch, obtain the mIoU and mAP of each category on the validation set, calculate their mean, and select the model parameters with the highest mAP index to save.

[0096] Among them, the loss function of the target detection model is:

[0097]

[0098] Among them, L is the loss function, L cls represents the classification loss, L regRepresents the regression loss. The loss function calculation formula is as follows:

[0099]

[0100] Among them, L reg Represents the regression loss, using the GIoU loss function, T i is the predicted value of the i-th box, T * i is the true value of the corresponding box. cls Represents classification loss, using binary cross entropy loss function, S i is the category prediction of the i-th box, S * i is the true value of the category of the corresponding box. When judging the corresponding box, we use the annotated box with the largest IoU with the predicted box as the corresponding box to match it.

[0101] In summary, this embodiment performs feature processing on remote sensing images through the backbone network module to obtain a multi-scale feature map, which can extract image features from different levels and angles, provide richer information for subsequent target detection, and help identify targets more accurately. The adaptive occlusion perception module obtains an image occlusion mask map by performing occlusion processing on the multi-scale feature map, which can effectively perceive the occlusion situation in the image, thereby solving the occlusion problem in a targeted manner in the feature optimization processing, reducing false detection and missed detection caused by occlusion, and improving the accuracy of target detection. The adaptive occlusion perception module can automatically adjust the occlusion processing and alignment mark processing according to different remote sensing images, adapt to the occlusion situation in different scenes, and improve the adaptability of the model. The mapping module performs feature mapping and quickly obtains detection results, which can reduce calculation time, improve the efficiency of target detection, and meet the needs of real-time or large-scale remote sensing image detection.

[0102] Example 2

[0103] like Figure 3 As shown, this embodiment provides a remote sensing image target detection system for implementing the remote sensing image target detection method described in Example 1. The system includes a data acquisition unit and a target detection unit, wherein the target detection unit includes a backbone network module, an adaptive occlusion perception module, and a mapping module; specifically:

[0104] A data acquisition unit, used for acquiring remote sensing images to be detected;

[0105] The target detection unit is used to input the remote sensing image to be detected into the trained target detection model, perform target detection on the remote sensing image, and obtain a detection result;

[0106] The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; the features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain occlusion area detection results based on the mapping features, obtain non-occlusion area detection results based on the features in the multi-scale feature map, and finally obtain the detection results based on the occlusion area detection results and the non-occlusion area detection results.

[0107] The multi-module structure of the model enables it to comprehensively utilize the advantages of different modules for processing when faced with complex remote sensing image data, enhancing the robustness of the model and enabling stable target detection under various environmental conditions.

[0108] Example 3

[0109] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a terminal device, used to store programs and data. It is understood that the computer-readable storage medium herein may include both built-in storage media in the terminal device and, of course, extended storage media supported by the terminal device. It may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by a processor. These instructions may be one or more computer programs (including program code). It should be noted that more specific examples (a non-exhaustive list) of computer-readable storage media herein include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk-read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0110] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program code. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0111] The program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and the like, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0112] The processor may load and execute one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the remote sensing image target detection method in the above embodiment; the processor may load and execute the following steps:

[0113] Acquire remote sensing images to be detected;

[0114] Inputting the remote sensing image to be detected into the trained target detection model, performing target detection on the remote sensing image, and obtaining a detection result;

[0115] The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; the features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain occlusion area detection results according to the mapping features, obtain non-occlusion area detection results according to the features in the multi-scale feature map, and finally obtain the detection results according to the occlusion area detection results and the non-occlusion area detection results.

[0116] Figure 4 A schematic diagram of a computer device provided in one embodiment of the present invention.

[0117] See also Figure 4 The terminal device is a computer device. The computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable by the processor 61. When the computer program 63 is executed by the processor 61, the remote sensing image target detection method of the embodiment is implemented. To avoid repetition, the details are not described here. Alternatively, when the computer program 63 is executed by the processor 61, the functions of each model / unit in Example 1 are implemented. To avoid repetition, the details are not described here.

[0118] The computer device 60 may be a desktop computer, a notebook computer, a PDA, a cloud server, or other computing devices. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. It will be understood by those skilled in the art that Figure 4 This is merely an example of the computer device 60 and does not constitute a limitation of the computer device 60 . The computer device 60 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0119] The processor 61 may be a central processing unit (CPU), other general-purpose processors, central processing units (CPUs), graphics processors (GPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), other programmable logic devices, discrete gate or transistor logic devices, quantum computing-based data processing logic, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0120] The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device 60.

[0121] Furthermore, the memory 62 may include both an internal storage unit of the computer device 60 and an external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is about to be output.

[0122] Any reference to memory, database, or other media used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0123] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.

Claims

1. A remote sensing image target detection method, characterized in that: include: Acquire remote sensing images to be detected; Inputting the remote sensing image to be detected into the trained target detection model, performing target detection on the remote sensing image, and obtaining a detection result; The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; The features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain occlusion area detection results according to the mapping features, obtain non-occlusion area detection results according to the features in the multi-scale feature map, and finally obtain a detection result according to the occlusion area detection results and the non-occlusion area detection results; The mapping module is used to perform dot product processing on the feature map and the image occlusion mask map, hide the area where there is no occlusion, and add the corresponding dot product result and the feature map to input the feature mapping layer to obtain mapping features; The mapping module is further provided with an occlusion detection module and a non-occlusion detection module; the occlusion detection module is used to convolve the mapping features to obtain a classification and detection box offset prediction result, perform positive sample determination based on the prediction result, and obtain a detection result of the occlusion area; the detection box is a preset anchor box, which is used to sum with the offset result to obtain the corresponding detection result; The non-occlusion detection module is used to obtain the prediction results of classification, detection frame and center, perform positive sample determination based on the prediction results, and obtain the detection results of the non-occlusion area.

2. The remote sensing image target detection method according to claim 1, characterized in that: The backbone network module adopts the ResNet network; the backbone network module includes a backbone submodule and a feature pyramid submodule. The backbone submodule extracts features of the input remote sensing image to obtain a first feature map; the feature pyramid submodule is used to perform convolution processing on the features in the first feature map to obtain a multi-scale feature map.

3. The remote sensing image target detection method according to claim 1, characterized in that: The adaptive occlusion perception module is used to perform convolution processing on the input multi-scale feature map, and then process it using the softmax function to obtain a first result map. The pixel positions in the first result map that are greater than a first set threshold are set to 1, and the remaining pixel positions are set to 0 to obtain an image occlusion mask map.

4. The remote sensing image target detection method according to claim 1, characterized in that: A detection module is further connected to the occlusion detection module and the non-occlusion detection module; the detection module receives positive samples of all occlusion detection modules and non-occlusion detection modules respectively, and after scale transformation, performs non-maximum suppression processing on the positive samples to obtain occlusion detection results and non-occlusion detection results; For the occlusion detection results, the IoU is calculated with the non-occlusion detection results, and the occlusion detection results with IoU greater than 0.7 are retained. Finally, the retained occlusion detection results and the non-occlusion detection results are integrated as the final detection results.

5. The remote sensing image target detection method according to claim 1, characterized in that: The remote sensing image data is preprocessed, and the preprocessing step includes standardizing and normalizing the remote sensing image data.

6. The remote sensing image target detection method according to claim 5, characterized in that: After the remote sensing image data is pre-processed, the remote sensing image data is further enhanced, specifically including horizontal flipping and random cropping: The horizontal flipping specifically includes: flipping the remote sensing image along the vertical center, making the left and right mirror symmetrical, and mirror-reversing the annotation box corresponding to the remote sensing image; The random cropping specifically includes: randomly cropping a sub-region in the original remote sensing image as a training sample, detecting whether the corresponding annotation box exceeds the image boundary through the training sample, and if so, aligning the annotation box to the image edge.

7. The remote sensing image target detection method according to any one of claims 1 to 4, characterized in that: The target detection model sets different learning rates and weight decay parameters during training, sets the number of training rounds and hyperparameters; Adamw is used as the model optimizer and poly is used as the learning rate strategy; the DIOR remote sensing dataset is used to iteratively train the target detection model; According to the value of the loss function, the corresponding parameters of the target detection model are updated through back propagation until the optimal model parameters are obtained.

8. The remote sensing image target detection method according to claim 7, characterized in that: The loss function is: Where L is the loss function, represents the classification loss, represents the regression loss.

9. A remote sensing image target detection system, used to implement the remote sensing image target detection method according to any one of claims 1 to 8, characterized in that: include: A data acquisition unit, used for acquiring remote sensing images to be detected; The target detection unit is used to input the remote sensing image to be detected into the trained target detection model, perform target detection on the remote sensing image, and obtain a detection result; The target detection model includes a backbone network module, an adaptive occlusion perception module and a mapping module. The backbone network module is used to perform feature processing on the input remote sensing image to obtain a multi-scale feature map; the adaptive occlusion perception module is used to perform occlusion processing on the multi-scale feature map to obtain an image occlusion mask map; the features are optimized according to the image occlusion mask map; the mapping module is used to perform feature mapping to obtain mapping features, obtain the detection result of the occluded area according to the mapping features, obtain the detection result of the non-occluded area according to the features in the multi-scale feature map, and finally obtain the detection result according to the occluded area detection result and the non-occluded area detection result.

Images