RepDarkNet: A Multi-Branched Detector for Small-Target Detection in Remote Sensing Images

Aiming at the problem of low target detection accuracy caused by the close arrangement of targets, complex feature information, and numerous small targets in remote sensing images, this paper proposes a target detection algorithm based on the improved Yolov5 algorithm for remote sensing images. Firstly, this paper designs a remote sensing image target detection algorithm C²-YOLO based on multi-level feature fusion. Combined with the attention mechanism module to focus on the spatial location information of the target, which improves the extraction effect of the features of the target of interest. The network uses the CARAFE upsampling operator with a larger sensing field to better utilize the surrounding information. The proposal is to incorporate a predictive header hl that is generated from a low-level, high-resolution feature map to improve the recognition of tiny objects. The problem of large target scales in remote sensing images and poor detection of small targets has been improved. Angular loss is also added to the loss function to improve the detection of rotating targets. Finally, the accuracy comparison and experimental effect visualization are carried out on the dataset, and the experimental results show that the improvement work in this paper achieves better results no matter comparing the original YOLOv5 detection algorithm or other types of detection algorithms. Specifically, the accuracy of C²-YOLO on the NWPU dataset 1-shot recognition task is 78.45%, which is much higher than that of other detection algorithms, and is able to reach a superior performance in the remote sensing image target detection task.

Small and dim target detection is a critical challenge in hyperspectral remote sensing, particularly in complex, large-scale scenes where spectral variability across diverse land cover types complicates the detection process. In this paper, we propose a novel target reasoning algorithm named Adaptive Global Dense Nested Reasoning Network (AGDNR). This algorithm integrates spatial, spectral, and domain knowledge to enhance the detection accuracy of small and dim targets in large-scale environments and simultaneously enables reasoning about target categories. The proposed method involves three key innovations. Firstly, we develop a high-dimensional, multi-layer nested U-Net that facilitates cross-layer feature transfer, preserving high-level features of small and dim targets throughout the network. Secondly, we present a novel approach for computing physicochemical parameters, which enhances the spectral characteristics of targets while minimizing environmental interference. Thirdly, we construct a geographic knowledge graph that incorporates both target and environmental information, enabling global target reasoning and more effective detection of small targets across large-scale scenes. Experimental results on three challenging datasets show that our method outperforms state-of-the-art approaches in detection accuracy and achieves successful classification of different small targets. Consequently, the proposed method offers a robust solution for the precise detection of hyperspectral small targets in large-scale scenarios.

The presence of shadows in high-resolution (HR) remote-sensing images reduces object detection accuracy. To address this problem, in this paper, we proposed a deep neural network algorithm for shadow detection by using the AISD and SSAD remote-sensing shadow image datasets. To improve the ability to extract spatial information from feature maps, we developed a cross-spatial attention module that focuses on semantic information in the horizontal and vertical directions at each position point on the remote-sensing image. This module overcomes the limitations of existing technologies in accurately judging small areas and suspected shadow areas and in missing or incorrectly detected shadow areas. In addition, to improve the ability to extract shadow features and the accuracy of shadow detection in remote-sensing images, we developed a channel attention module that assigns more attention to channels that conform to the shadow color characteristics. The network architecture comprises an encoder – decoder structure, with ResNeXt50 used as the backbone for the encoder and a multi resolution parallel fusion (MRPF) designed for the decoder; cross-spatial and channel attention were incorporated into the decoder unit. Experimental results demonstrated the superior performance of the proposed algorithm, with an F1 score of 92.6% for the shadow category on the test set, thus, outperforming other algorithms and making the proposed method an effective solution for shadow detection in HR remote-sensing images.

With the continuous growth of maritime activities and the shipping trade, the application of maritime target detection in remote sensing images has become increasingly important. However, existing detection methods face numerous challenges, such as small target localization, recognition of targets with large aspect ratios, and high computational demands. In this paper, we propose an improved target detection model, named YOLOv5-ASC, to address the challenges in maritime target detection. The proposed YOLOv5-ASC integrates three core components: an Attention-based Receptive Field Enhancement Module (ARFEM), an optimized SIoU loss function, and a Deformable Convolution Module (C3DCN). These components work together to enhance the model’s performance in detecting complex maritime targets by improving its ability to capture multi-scale features, optimize the localization process, and adapt to the large aspect ratios typical of maritime objects. Experimental results show that, compared to the original YOLOv5 model, YOLOv5-ASC achieves a 4.36 percentage point increase in mAP@0.5 and a 9.87 percentage point improvement in precision, while maintaining computational complexity within a reasonable range. The proposed method not only achieves significant performance improvements on the ShipRSImageNet dataset but also demonstrates strong potential for application in complex maritime remote sensing scenarios.

With the rapid advancement of technology, satellite and drone technologies have had significant impacts on various fields, creating both opportunities and challenges. In areas like the military, urban planning, and environmental monitoring, the application of remote sensing technology is paramount. However, due to the unique characteristics of remote sensing images, such as high resolution, large-scale scenes, and small, densely packed targets, remote sensing object detection faces numerous technical challenges. Traditional detection methods are inadequate for effectively detecting small targets, rendering the accurate and efficient detection of objects in complex remote sensing images a pressing issue. Current detection techniques fall short in accurately detecting small targets compared to medium and large ones, primarily due to limited feature information, insufficient contextual data, and poor localization capabilities for small targets. In response, we propose an innovative detection method. Unlike previous approaches that often focused solely on either local or contextual information, we introduce a novel Global and Local Attention Mechanism (GAL), providing an in-depth modeling method for input images. Our method integrates fine-grained local feature analysis with global contextual information processing. The local attention concentrates on details and spatial relationships within local windows, enabling the model to recognize intricate details in complex images. Meanwhile, the global attention addresses the entire image’s global information, capturing overarching patterns and structures, thus enhancing the model’s high-level semantic understanding. Ultimately, a specific mechanism fuses local details with global context, allowing the model to consider both aspects for a more precise and comprehensive interpretation of images. Furthermore, we have developed a multi-head prediction module that leverages semantic information at various scales to capture the multi-scale characteristics of remote sensing targets. Adding decoupled prediction heads aims to improve the accuracy and robustness of target detection. Additionally, we have innovatively designed the Ziou loss function, an advanced loss calculation, to enhance the model’s precision in small target localization, thereby boosting its overall performance in small target detection. Experimental results on the Visdrone2019 and DOTA datasets demonstrate that our method significantly surpasses traditional methods in detecting small targets in remote sensing imagery.

An improved algorithm is proposed to solve the problems of inaccurate recognition and low recall of Faster-Regions with Convolutional Neural Network (Faster-RCNN) algorithm for the detection of ship targets in remote sensing images. The algorithm is based on the Faster-RCNN network framework. Aiming at the small size and dense distribution of ship targets in remote sensing images, the feature extraction network is improved to enhance the detection ability of small targets. ResNet50 is used as the basic feature extraction network of the algorithm,and the hole residual block is introduced for multi-layer feature fusion to construct a new feature extraction network,which improves the feature extraction capability of the algorithm. The experimental results show that compared with the Faster-RCNN algorithm, this algorithm can learn more abundant target features in smaller pixel areas, thereby effectively improving the detection accuracy of ship targets.

This study focuses on advancing the field of remote sensing image target detection, addressing challenges such as small target detection, complex background handling, and dense target distribution. We propose solutions based on enhancing the YOLOv7 algorithm. Firstly, we improve the multi-scale feature enhancement (MFE) method of YOLOv7, enhancing its adaptability and precision in detecting small targets and complex backgrounds. Secondly, we design a modified YOLOv7 global information DP-MLP module to effectively capture and integrate global information, thereby improving target detection accuracy and robustness, especially in handling large-scale variations and complex scenes. Lastly, we explore a semi-supervised learning model (SSLM) target detection algorithm incorporating unlabeled data, leveraging information from unlabeled data to enhance the model’s generalization ability and performance. Experimental results demonstrate that despite the outstanding performance of YOLOv7, the mean average precision (MAP) can still be improved by 1.9%. Specifically, under testing on the TGRS-HRRSD-Dataset, the MFE and DP-MLP models achieve MAP values of 93.4% and 93.1%, respectively. Across the NWPU VHR-10 dataset, the three models achieve MAP values of 93.1%, 92.1%, and 92.2%, respectively. Significant improvements are observed across various metrics compared to the original model. This study enhances the adaptability, accuracy, and generalization of remote sensing image object detection.

With the rapid development of UAV and infrared imaging technology, the cost of UAV infrared imaging technology has decreased steadily. Small target detection technology in aerial infrared images has great potential for applications in many fields, especially in the field of traffic surveillance. Because of the low contrast and relatively limited feature information in infrared images compared to visible images, the difficulty involved in small road target detection in infrared aerial images has increased. To solve this problem, this study proposes a feature-enhanced attention and dual-GELAN net (FEADG-net) model. In this network model, the reliability and effectiveness of small target feature extraction is enhanced by a backbone network combined with low-frequency enhancement and a swin transformer. The multi-scale features of the target are fused using a dual-GELAN neck structure, and a detection head with the parameters of the auto-adjusted InnerIoU is constructed to improve the detection accuracy for small infrared targets. The viability of the method was proved using the HIT-UAV dataset and IRTS-AG dataset. According to a comparative experiment, the mAP50 of FEADG-net reached more than 90 percent, which was higher than that of any previous method and it met the real-time requirements. Finally, an ablation experiment was conducted to demonstrate that all three of the modules proposed in the method contributed to the improvement in the detection accuracy. This study not only designs a new algorithm for small road object detection in infrared remote sensing images from UAVs but also provides new ideas for small target detection in remote sensing images for other fields.

Research on Infrared Dim and Small Target Detection Algorithm Based on Low-Rank Tensor Recovery

The earthquake-induced landslide targets in remote sensing images vary greatly in size and are unevenly distributed with many small targets. Achieving a balance between high accuracy, computational capability, and small sample size remains challenging. This study proposes to enhance earthquake-induced landslide detection by developing a new algorithm for remote sensing images based on the C3-Swin-Transformer and Multiscale Feature Fusion-YOLO (TLSTMF-YOLO). Utilizing a feature extraction layer and Swin-Transformer structure captures dependencies and preserves spatial information. Introducing the Convolutional Block Attention Module (CBAM) enhances feature representation. Incorporating a Bidirectional Feature Pyramid Network (BiFPN) optimizes bidirectional cross-scale feature fusion, improving landslide detection accuracy across scales. The training utilizes an AdamW optimizer and cosine learning rate strategy for accelerated convergence and improved speed. Transfer learning applies to Jiuzhaigou and Luding landslide datasets. Experimental results show that the TLSTMF-YOLO model outperforms YOLOv5 and other detection models in terms of precision, recall, and mAP@0.5. Specifically, on the Jiuzhaigou dataset, it achieves a precision of 95.7%, a recall of 89.9%, and a mAP@0.5 of 90.5%. On the Luding dataset, it achieves a precision of 96.0%, a recall of 90.9%, and a mAP@0.5 of 94.5%. Additionally, the frame processing times for the TLSTMF-YOLO model are 6.61 ms and 12.2 ms on the Jiuzhaigou and Luding datasets, respectively, demonstrating superior efficiency and confirming its effective feature extraction and fusion capabilities.

Dim and small infrared target fast detection guided by visual saliency

Images captured using unmanned aerial vehicles (UAVs) often exhibit dense target distribution and indistinct features, which leads to the issues of missed detection and false detection in target detection tasks. To address these problems, an improved method for small target detection called YOLOv5s is proposed to enhance the detection accuracy for small targets such as apple fruits. By applying improvements to the RFA module, DFP module, and Soft-NMS algorithm, as well as integrating these three modules together, accurate detection of small targets in images can be achieved. Experimental results demonstrate that the integrated, improved model achieved a significant improvement in detection accuracy, with precision, recall, and mAP increasing by 3.6%, 6.8%, and 6.1%, respectively. Furthermore, the improved method shows a faster convergence speed and lower loss value during the training process, resulting in higher recognition accuracy. The results of this study indicate that the proposed improved method exhibits a good performance in apple fruit detection tasks involving UAV imagery, which is of great significance for fruit yield estimation. The research findings demonstrate the effectiveness and feasibility of the improved method in addressing small target detection tasks, such as apple fruit detection.

To address the challenges of multi-scale objects, dense distributions, occlusions, and numerous small targets in UAV image detection, we present CMS-YOLOv7, a real-time target detection method based on an enhanced YOLOv7 model. Firstly, the detection layer P2 for small targets was added to YOLOv7 to enhance the detection ability of small and medium-sized targets, and the deep detection head P5 was taken out to mitigate the influence of excessive downsampling on small target images. The anchor frame was calculated by the K-means++ method. Using the concept of Inner-IoU, the Inner-MPDIoU loss function was constructed to control the range of the auxiliary border and improve detection performance. Furthermore, the CARAFE module was introduced to replace traditional upsampling methods, offering improved integration of semantic information during the image upsampling process and enhancing feature mapping accuracy. Simultaneously, during the feature extraction stage, a non-strided convolutional SPD-Conv module was constructed using space-to-depth techniques. This module replaced certain convolutional operations to minimize the loss of fine-grained information and improve the model’s ability to extract features from small targets. Experiments on the UAV aerial photo dataset VisDrone2019 demonstrated that compared with the baseline YOLOv7 object detection algorithm, CMS-YOLOv7 achieved an improvement of 3.5% mAP@0.5, 3.0% mAP@0.5:0.95, and the number of parameters decreased by 18.54 M. The ability of small target detection was significantly enhanced.

To detect small infrared targets under the condition of dense clutters, we propose a single-frame target detection algorithm based on a small bounding-box filter, which is characterized by good adaptability to the position and size of a small target. During the small target detection process, the proposed algorithm first searches for the local maximum gray pixel and then, a set of concentric bounding boxes whose center is the pixel found in the first step is constructed, and the detection thresholds of a neighboring region of this pixel are calculated based on the bounding boxes. Finally, the minimum threshold is used to detect small target pixels in the neighboring region. A fast version of the proposed algorithm is a minimum bounding-box filter, which can be implemented by dividing an image into blocks and using the mid-range and range to assess the concentration trend and dispersion of the background. Simulation and analysis results show that the proposed algorithm can achieve high detection probability and low false alarm rate when detecting small targets in the complex background; while its fast version has high computational efficiency. The proposed algorithm can be used in infrared searching and tracking systems.

ABSTRACTIn this article, we propose a novel difference image (DI) creation method for unsupervised change detection in multi-temporal multi-spectral remote-sensing images based on deep learning theory. First, we apply deep belief network to learn local and high-level features from the local neighbour of a given pixel in an unsupervised manner. Second, a back propagation algorithm is improved to build a DI based on selected training samples, which can highlight the difference on changed regions and suppress the false changes on unchanged regions. Finally, we get the change trajectory map using simple clustering analysis. The proposed scheme is tested on three remote-sensing data sets. Qualitative and quantitative evaluations show its superior performance compared to the traditional pixel-level and texture-level-based approaches.