Remote sensing image target detection plays a pivotal role in earth observation, offering substantial value for applications such as urban planning and environmental monitoring. Due to the significant scale variations among targets, complex backgrounds with dense small object distributions, and strong inter-target scene correlations, existing target detection methods usually fails to effectively model target relationships and contextual information for remote sensing imagery. To address these limitations, we proposed YOLO-ALS, a novel remote sensing target detection network that integrates adaptive local scene context. The proposed framework introduces three key points: First, a full-dimensional dynamic convolution reconstruction C2f module enhances target feature representation by overcoming local context extraction limitations and target co-occurrence prior deficiencies. Second, an adaptive local scene context module dynamically integrates multi-scale receptive field features through spatial attention, enabling background window adaptive selection and cross-scale feature alignment. Finally, a co-occurrence matrix-integrated classification auxiliary module mines target association rules through data-driven learning, correcting classification probabilities in low-confidence areas by combining high-confidence areas co-occurrence information with optimal threshold, which can significantly reduce missed detection rates. Comprehensive experiments on multiple public remote sensing datasets demonstrate the superiority of the proposed method through extensive ablation studies and comparative analyses. The proposed method has achieved state-of-the-art performance while addressing the unique challenges of remote sensing target detection.

Urban Informal Settlements (UIS) denote densely populated locales characterized by inadequate urban infrastructure standards, often exhibiting an amalgamation of rural and urban attributes, primarily situated within the confines of major cities or metropolitan regions. UIS mapping is a typical task that aims to identify pixels corresponding to urban informal settlements in remote sensing images. The extremely similar visual characteristics, sample uncertainty, and highly manual costs bring large-scale UIS mapping noteworthy challenges. In this paper, we propose a sample uncertainty-aware semi-supervised learning method guided by background-knowledge from crowdsourced geospatial data (namely SemiUIS) for UIS mapping and produces a 1-meter resolution UIS map in the Urban Agglomeration in the Middle Reaches of the Yangtze River, China (UAMRYR). Our proposed SemiUIS method integrates semisupervised learning with the guidance of background knowledge from crowdsourcing data to reduce the uncertainty of the generated pseudo-semantic segmentation samples to improve the quality of the constructed data set by jointly optimizing the background (Not UIS areas) and the foreground (UIS areas). In contrast, traditional semi-supervised learning methods only focus on explicitly optimizing the foreground (UIS areas) in the pseudo-sample generation process. Experiments were conducted in the 31 prefecture-level cities of UAMRYR. Visual interpretation of labeled remote sensing samples, street view images and two novel knowledge validation indicators, Global Knowledge Precision (GKP) and Local Knowledge Precision (LKP) are proposed to verify the UIS mapping results. The proposed method reached an overall accuracy (OA) of 90.58 %, mean intersection over union (mIoU) of 75.88 %, Accuracy (Acc) of 78.77 %, Intersection over Union (IoU) of 62.97 %, GKP of 90.33 %, and LKP of 85.01 %, respectively with only 36 (5%) training samples. This work will be available at https://github.com/RunyuFan/UisYangtze.

The ground object distribution in remote sensing images exhibits strong regularities. However, existing deep learning-based object detection models often focus solely on instance-level information within sample labels, neglecting the modeling of relationships between instances. Additionally, these models fail to effectively utilize the substantial background information present in remote sensing images. To address these issues, we propose a remote sensing object detection network named SCENE-YOLO, building upon the YOLOv8 architecture and introducing scene supervision. First, we introduce a scene information gathering and distribute network (SGD), based on transformer, to inject high-level semantic information into the feature pyramid. A slice-and-distribute mechanism is employed to prevent information loss during feature fusion across layers. Second, the backbone network is redesigned, incorporating the attention mechanism of omni-dimensional dynamic convolution (ODConv) to dynamically redistribute weights for target features. Subsequently, a scene label generation algorithm (SLGA) based on prototype learning is proposed to supervise the model by generating scene-level labels, modeling instance-instance relationships through the introduction of artificial knowledge and multilevel classification. Finally, a scene-assisted detection head (SADHead) is introduced to enhance detection performance in complex backgrounds by leveraging scene features with global contextual information to assist the model in target classification. Experimental validation on the publicly available DOTA and DIOR datasets demonstrates the effectiveness and superiority of the proposed algorithm.

Spatial-regularized spectral unmixing has achieved great progress and attracted widespread attention for addressing homogeneous regions with identical spectral characteristics. In this paper, a new hyperspectral unmixing algorithm with graph Laplacian regularization and local collaborative sparse regression is proposed, based on superpixel segmentation, namely GLCGSU. Considering mixed pixels in homogeneous areas have similar endmembers and abundances, we utilize superpixel image clustering (SLIC) to cluster similar pixels, leveraging boundary information for uniform area extraction. Spatial similarity is investigated via graph Laplacian regularization. Meanwhile, we apply local collaborative weighted sparse regression to achieve abundance matrix sparsity. Experimental results demonstrates the effectiveness of the proposed method both on simulated and real data, proving its superiority for hyperspectral unmixing.

Spatial correlations between different ground objects are an important feature of mining land cover research. Graph Convolutional Networks (GCNs) can effectively capture such spatial feature representations and have demonstrated promising results in performing hyperspectral imagery (HSI) classification tasks of complex land. However, the existing GCN-based HSI classification methods are prone to interference from redundant information when extracting complex features. To classify complex scenes more effectively, this study proposes a novel spatial-spectral reliable contrastive graph convolutional classification framework named S2RC-GCN. Specifically, we fused the spectral and spatial features extracted by the 1D- and 2D-encoder, and the 2D-encoder includes an attention model to automatically extract important information. We then leveraged the fused high-level features to construct graphs and fed the resulting graphs into the GCNs to determine more effective graph representations. Furthermore, a novel reliable contrastive graph convolution was proposed for reliable contrastive learning to learn and fuse robust features. Finally, to test the performance of the model on complex object classification, we used imagery taken by Gaofen-5 in the Jiang Xia and Xin Jiang area to construct complex land cover datasets. The test results show that compared with other models, our model achieved the best results and effectively improved the classification performance of complex remote sensing imagery.

High-dimensional and complex spectral structures make the clustering of hyperspectral images (HSI) a challenging task. Subspace clustering is an effective approach for addressing this problem. However, current subspace clustering algorithms are primarily designed for a single view and do not fully exploit the spatial or textural feature information in HSI. In this study, contrastive multi-view subspace clustering of HSI was proposed based on graph convolutional networks. Pixel neighbor textural and spatial-spectral information were sent to construct two graph convolutional subspaces to learn their affinity matrices. To maximize the interaction between different views, a contrastive learning algorithm was introduced to promote the consistency of positive samples and assist the model in extracting robust features. An attention-based fusion module was used to adaptively integrate these affinity matrices, constructing a more discriminative affinity matrix. The model was evaluated using four popular HSI datasets: Indian Pines, Pavia University, Houston, and Xu Zhou. It achieved overall accuracies of 97.61%, 96.69%, 87.21%, and 97.65%, respectively, and significantly outperformed state-of-the-art clustering methods. In conclusion, the proposed model effectively improves the clustering accuracy of HSI. Our implementation is available at https://github.com/GuanRX/CMSCGC.

Ensuring the quality of trajectories is of utmost importance in traffic flow analysis. Traditional approaches rely on reconstructing nearly complete trajectories and subsequently denoising them. However, low detection rates often pose challenges and result in failed trajectory construction. To overcome this issue, this paper presents a trajectory concatenation method that combines NS Transformer prediction and Siamese-VGG16 similarity confirmation, specifically designed to address low detection rates. The employed transformer model can withstand missing values, efficiently extracting internal associations among multiple traffic parameters in conditions of sparse data. Furthermore, a lightweight image feature similarity verification step is integrated after trajectory prediction to find the most similar target to the image in the predicted spatiotemporal domain. Additionally, a lightweight image feature similarity verification step is integrated after trajectory prediction to identify the most similar targets within the predicted spatiotemporal domain. Experimental results demonstrate the efficacy of the proposed method, successfully connecting over 80% of fragmented tracks and yielding significant maintenance of MOTA above 0.74 under low detection accuracy.

Finding correspondences between a pair of images is the key ingredient for many applications such as localization and panorama. However, due to a variety of challenges between multi-view images in practice, the results of using a single kind of descriptor may vary significantly across different scenes. In this paper, we treat the assignment task as a clustering problem and propose an image matching method that fuses multiple descriptors to tackle the above difficulties. First, we extract multiple descriptors at the keypoints on two images. Then, we compute a pairwise similarity matrix for each kind of descriptor. Afterwards, we compute a weighted combination of these similarity matrices, and use it to build correspondences via a modified multi-kernel clustering module. The proposed method is tested on three public image datasets: two ground image sets and an Unmanned Aerial Vehicle (UAV) image set. Experiments show that the proposed method can adapt to different number of descriptors. It significantly improves the matching accuracy in a variety of scenarios and downstream tasks.

The active population is an indicator of urban vitality, representing socioeconomic vitality. An accurate mapping for the active population is the foundation for supporting regional sustainable development. Current research mainly relies on human mobile location data, which are challenging to access due to privacy and data sharing concerns. Thus, it is essential to map a wide area of the active population cost-effectively. This study demonstrates how an end-to-end deep learning approach can be used to reliably estimate urban active population distribution from remote sensing imagery. Learning based on the object and regional feature may provide insights into active population distribution by coupling ResNet-50 and Fully Convolutional Networks to achieve economic features that fuse Google surface features and night-light remote sensing data. Our results demonstrate the feasibility of using multiple remote sensing data to quantify the distribution of active populations. The proposed model$(\text{ResNet}-50+\text{FCN},\ R^{2}=0.82)$can better explain the differences in active population estimation than the ResNet-50 model ($R^{2}=0.61$). The study provides new insights into estimating active populations in cities that lack location-based service records.

Recently, deep-learning-based methods, especially deep convolutional neural networks (DCNNs), have effectively shown state-of-the-art performance in road extraction from high resolution remote sensing images (HRSI). However, due to the loss of location information and global context information, most existing DCNNs are inadequate for extracting tiny roads or roads which are severely occluded, leading to incomplete and discontinuous results. To address this problem, this paper proposes a graph-based dual convolutional network (GDCNet), which combines graph convolutional network (GCN) and convolutional neural network (CNN). In this model, GCN and CNN branches perform feature learning on large-scale irregular regions and small-scale regular regions, and generate complementary spatial-spectral features at superpixel and pixel levels, respectively. Then, a graph decoder is utilized to propagate features between graph nodes and image pixels, enabling the GCN and CNN to collaborate in a single network. Extensive experiments on two benchmark datasets demonstrate that the proposed GDCNet is competitive compared with other state-of-the-art methods both qualitatively and quantitatively, and is effective against the incomplete and discontinuous problems of the extracted roads.

Sparse unmixing aims at finding the optimal subset of endmembers in a spectral library to approximate the observed data, and has received increasing attention as it can circumvent the estimation of the endmember. In this paper, a graph Laplacian regularized spectral-spatial-sparse unmixing algorithm is proposed, namely, gLapS3U, incorporating the graph Laplacian regularization to consider the similarity between pixels of the whole image, and enforcing the spectral-spatial-sparse constraints to enhance the local spatial information as well as the sparsity of the abundance solution jointly. Experimental results on simulated and real data show the superiority of the proposed algorithm compared with state-of-the-art existing methods.

Due to the limitations of the imaging sensor, the spatial resolution of satellite imagery is often insufficient, namely, low resolution (LR). Therefore, super-resolution (SR) is proposed, which strives to improve image resolution, perfectly to compensate for the shortcomings of satellite sensor imaging. In this study, we develop a unique dilated convolution network with gradient prior (DCNG) for remote sensing SR, aiming to extract powerful low-level features with gradient prior and efficitive network and then reconstruct the high-level feature details. The DCNG is built of two components: the Multi-Scale Feature Extraction Network and the Feature Reconstruction Network. In the Multi-Scale Feature Extraction Network, the Double-Path Dilated Residual Block (DPDRB) is designed with the dilation convolution operation to obtain the multi-scale features and increase the receptive field, the Global Self-attention Module (GSA) to catch the long-range dependency among picture patches, and a Gradient Propagation Network (GPN) is proposed to extract high-level gradient information. In the Feature Reconstruction Network, the Pixel Shuffle is introduced to reconstruct the feature by combining characteristics of different frequency bands. Experiments using Massachusetts_Roads and 3K VEHICLE_SR data sets indicate that our DCNG surpasses state-of-the-art algorithms in terms of quantitative and qualitative evaluations.

Sea ice concentration (SIC) is an indispensable parameter for the study of polar sea ice. The existing methods can obtain accurate SICs for most situations, but they usually perform poorly in low SIC regions because of the spatial differences in the neighboring pixels induced by the discontinuity of the sea ice cover. In this letter, to cope with the difficulty of this problem, an improved SIC estimation method is proposed to retrieve SIC, focusing on low SIC regions. The proposed method introduces the spatial relationships into SIC estimation by employing a total variation (TV) regularizer. Moreover, nonnegative constrained least squares (NCLS) is used to derive the optimal solutions from the SIC estimation equation. Verification was conducted in low SIC regions (0%–50%) of the Antarctic utilizing ship-based in situ data and the Moderate Resolution Imaging Spectroradiometer (MODIS), and the results were compared with those of some of the mature methods. The results indicated that the proposed method can obtain a superior accuracy with a smaller root-mean-square error (RMSE) (6.0%–14.61%) than the other algorithms in low SIC regions. Furthermore, the proposed method can accurately estimate the SIC of both first-year ice and multiyear ice. The findings of this study confirm the need to consider the spatial relationships in the processing of SIC estimation.

Hyperspectral anomaly detection, which is aimed at locating anomaly, has received widespread attention. In this article, a new anomaly detector, named local spatial constraint and total variation (LSC-TV), is proposed for hyperspectral imagery. In anomaly detection methods based on low-rank representation, background pixels are usually considered to have a global low-dimensional structure. However, the complex background distribution in hyperspectral images (HSIs) means that this global low-dimensional structure rarely occurs. In LSC-TV, the effective local spatial information is extracted by superpixel segmentation, and the regularization based on the F-norm is used to force the background within the same superpixel to show uniform spectral features. Moreover, each pixel is given a penalty based on the degree of anomaly determined during model iteration, while the anomaly is not considered by the background constraint. In addition, the background pixels in the neighborhood often show a high correlation, whereas the anomaly does not possess this feature. Nonisotropic TV is introduced into the proposed LSC model using the correlation of first-order neighborhoods to make it easier for anomalies to be separated. The proposed LSC-TV method and current state-of-the-art methods are tested on a set of simulated data and four sets of real data. The experimental results demonstrate that the proposed method is superior to the comparative method in terms of both color map detection and quantitative evaluation.

High-resolution (HR) remote sensing imagery plays a critical role in remote sensing image interpretation, and single image super-resolution (SISR) reconstruction technology is becoming increasingly valuable and significant. The state-of-the-art deep-learning-based SISR methods have demonstrated remarkable advantages, while reconstructing complex texture details still remains a big challenge. Besides, as a typical ill-posed inverse problem, how to determine the optimal solution is another important topic. To address these problems, in this work, a dual learning-based graph neural network (DLGNN) is proposed, in which the GNN is utilized to consider the self-similarity patches in remote sensing imagery by aggregating cross-scale neighboring feature patches, and dual learning strategy is adopted to refine the reconstruction results by constraining the mapping process in terms of the loss function, transferring the typical ill-posed problem to a well-posed one. Abundant experiments on 3K VEHICLE_SR datasets and Massachusetts Roads demonstrate the validity and outstanding performance for remote sensing image super-resolution tasks compared with other state-of-the-art super-resolution construction methods. Code is available at https://github.com/CUG-RS/DLGNN.

Hyperspectral unmixing is an essential task in hyperspectral imagery applications. Because of the strong feature extract ability and satisfying performance, deep learning methods have been used for hyperspectral unmixing. However, there are still several problems in existing deep learning based spectral unmixing methods. Supervised learning methods can only accomplish a single task and lack a large amount of data for supervised learning. While the unsupervised learning unmixing methods are easily misled by the traditional way of initialization. In this paper, a weakly supervised deep convolutional neural network is proposed for hyperspectral unmixing. The experimental results show that competitive results can also be obtained by pretraining with a small number of samples, and weakly supervised learning still has potential for hyperspectral unmixing.

Extracting roads from remote sensing images has both civilian and military value, such as GIS data update, road navigation, military command and so on. The existing road extraction methods are mainly based on fully convolutional neural networks, and have achieved the state-of-the-art results. However, the convolutional and deconvolutional forms of these methods destroy the completeness of the extracted road. In this paper, we present a novel road extraction method for extracting complete roads from high-resolution remote sensing imagery based on joint superpixel segmentation and Graph Convolutional Network(GCN). The proposed method retains more spatial detail information as well as effectively improves the integrity of the extracted roads. Experiments were conducted on the Massachusetts Road dataset to compare our proposed method to other commonly used full convolutional techniques for road extraction. The results demonstrated the validity and better performance of the proposed method.

Recently, hyperspectral anomaly detection methods based on low-rank representation(LRR) have been widely studied. However, the assumption of global low dimension of background may ignore the local structure information of hyperspectral image. In this paper, a novel LRR incorporating local spatial constraint method is proposed for hyperspectral anomaly detection. Different from LRR detector, the proposed method considers the spatial information based on the supe pixel in the background part. The proposed method and current state-of-the-art methods are tested on two sets of real data. The experimental results demonstrate that the proposed method is superior to the comparative method in terms of both colour map detection and quantitative evaluation.

Timely and effective quantitative measurement of enterprises' offline resumption of work after public emergencies is conducive to the formulation and implementation of relevant policies. In this paper, we analyze the level of work resumption after the coronavirus disease 2019 (COVID-19)-influenced Chinese Spring Festival in 2020 with National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) Day/Night Band (DNB) daily data. The results demonstrate that COVID-19 has seriously affected the resumption of work after the Spring Festival holiday. Since February 10th, work has been resuming in localities. By late March, the work resumption indexes of most cities exceeded 50%, and Shanghai and Nanjing even had achieved complete resumption of work. Our method effectively estimates the resumption of work, which provides a scientific basis for local governments to formulate subsequent resumption policies.

Due to the trade-off of temporal resolution and spatial resolution, spatiotemporal image-fusion uses existing high-spatial-low-temporal (HSLT) and high-temporal-low-spatial (HTLS) images as prior knowledge to reconstruct high-temporal-high-spatial (HTHS) images. However, some existing spatiotemporal image-fusion algorithms ignore the issue that the spatial information of HTLS images is insufficient to support the acquisition of spatial information, which leads to the unsatisfactory accuracy of the fusion result. To introduce more spatial information, the algorithm in this article uses Cycle-generative adversarial networks (GANs) to simulate the change process of two HSLT images at k-1 and k+1, and to generate some simulated images between k-1 and k+1. Then, the generated images are selected under the help of HTLS images, and the selected ones are then enhanced with wavelet transform. Finally, the image with spatial information is introduced into the Flexible Spatiotemporal DAta Fusion (FSDAF) framework to improve the performance of spatiotemporal image-fusion. Extensive experiments on two real data sets demonstrate that our proposed method outperforms current state-of-the-art spatiotemporal image-fusion methods.

Object detection, aiming to recognize and locate objects of interest in aerial images, has historically played a significant role in the remote sensing community. Following remarkable improvements in Earth observation technologies, high-resolution remote sensing (HRRS) images with a bird’s eye view perspective have revealed many categories of objects with sufficient variations in appearance and on complex backgrounds that make HRRS object detection an active but challenging task. The selection of positive samples and negative training instances is an essential factor in influencing detectors’ performance. Related studies have found that many low-quality negative samples in the detectors’ training process have caused training instability and low detection accuracy. In this work, a novel sampling-balance-based multistage network (SB-MSN) is presented to adaptively mine high-quality positive and negative instances for training an accurate detector. It has a series of components to ensure the selection and generation of high-quality examples for training an accurate detector, including a multiscale information retention module, an intersection over union balance sampling strategy, a balance L1 loss, and a multistage network. The proposed detector has been evaluated on three representative HRRS data sets. The extensive experimental results show that our detector can solve the problem of low-quality samples and significantly improve the detection performance of the mAP by 1.4% with the NWPU VHR-10 data set, 3.5% with the high-resolution remote sensing detection (HRRSD) data set, and 4.2% with the detection in the optical remote (DIOR) data set.1

Sparse unmixing, as a semisupervised unmixing method, has attracted extensive attention. The process of sparse unmixing involves treating the mixed pixels of hyperspectral imagery as a linear combination of a small number of spectral signatures (endmembers) in a standard spectral library, associated with fractional abundances. Over the past ten years, to achieve a better performance, sparse unmixing algorithms have begun to focus on the spatial information of hyperspectral images. However, less accurate spatial information greatly limits the performance of the spatial-regularization-based sparse unmixing algorithms. In this article, to overcome this limitation and obtain more reliable spatial information, a novel sparse unmixing algorithm named superpixel-based reweighted low-rank and total variation (SUSRLR-TV) is proposed to enhance the performance of the traditional spatial-regularization-based sparse unmixing approaches. In the proposed approach, superpixel segmentation is adopted to consider both the spatial proximity and the spectral similarity. In addition, a low-rank constraint is enforced on the objective function as pixels within each superpixel have the same endmembers and similar abundance values, and they naturally satisfy the low-rank constraint. Differing from the traditional nuclear norm, a reweighted nuclear norm is used to achieve a more efficient and accurate low-rank constraint. Meanwhile, low-rank consideration is also used to enhance the spatial continuity and suppress the effects of random noise. Furthermore, TV regularization is introduced to promote the smoothness of the abundance maps. Experiments on three simulated data sets, as well as a well-known real hyperspectral imagery data set, confirm the superior performance of the proposed method in both the qualitative assessment and the quantitative evaluation, compared with the state-of-the-art sparse unmixing methods.

Hyperspectral unmixing is an essential task in hyperspectral imagery applications. Deep learning methods have been taken into hyperspectral unmixing because of its great feature extraction ability and better performance. However, there are several problems in existing deep learning based spectral unmixing methods. The networks are not deep enough to exploit their feature extraction capabilities in these unsupervised autoencoders based methods, and their effects are not stable. The main reason may be the limited prior information limited the ability of conducting the supervised method. In this manuscript, a semi-supervised deep learning based unmixing method is proposed. Unlike the existing methods, our model uses deeper neural networks without pooling layers, and the endmember spectrum are selected supervised from the original data, which uses nature and nurture cooperatively. The experimental results show that the proposed method achieves better performance and produces more accurate abundance maps, as well as higher quantitative results, compared with the current state-of-the-art deep learning unmixing algorithms.

Spatiotemporal fusion utilizes the complementarity of high-temporal-low-spatial (HTLS) and high-spatial-low-temporal (HSLT) resolution data to obtain high temporal and spatial (HTHS) resolution fusion data, which can effectively satisfy the demand for HTHS data. However, due to the difference of spatial resolution, it is difficult to obtain precise spatial information in spatiotemporal fusion. To solve this problem, a multi-level strategy-based spatial domain prediction algorithm is proposed to enhance the spatial information extraction in spatiotemporal remote sensing imagery fusion, which can reduce the noise superposition in the process of multiple reconstruction. By learning-based first and then interpolation-based Super resolution reconstruction, the proposed method can obtain better prediction of spatial information and improve the accuracy of spatiotemporal fusion.

Urban land use and land cover (LULC) change is the result of urban population economic activities and national policy. Determining the spatial pattern of land cover types in cities is of particular significance for regional sustainable development. To achieve a better understand the spatiotemporal patterns of land use types in Shenzhen, the fractal dimension of spatial distributions is adopted as an index of the complex evolution of urban land-use. In addition, a long-term sequences LULC datasets are collected to do analysis, which covers the period 1988-2015 by employing Landsat TM/ETM+/OLI of 1988, 1993, 1999, 2001, 2005, 2008, 2011, 2013 and 2015. Last but not least, a granularity analysis is adopted to study the structural changes of each land cover. After analysis, it can be observed that a significant self-similarity law exists in the Shenzhen city. From 1988 to 2015, the fractal dimension of grassland, waterbody and bare land exhibits a bi-fractals dimension. However, grassland and bare land structure show a bi-fractals trend which increases every year, and the water-body bi-fractals trend is weakening. The development of urban land in this region experiences a process of a multiscale differential development with a hierarchical spatial system. These findings will provide some scientific references for the regional planning decisions on evolution of urban land use of Shenzhen city.

Focusing on the problems of insufficient high-quality training samples to conduct an ideal detector for high-resolution remote sensing (HRRS) image object, we applied a multi-stage based detector to apply a resampling progressively strategy, which guarantees the amount of the positive training set and minimizing overfitting. The method has a sequence of regression heads training on the samples chosen by different Intersection over Union (IoU) thresholds. The first head with a low IoU threshold trained by a large number of positive samples and can prepare more high-quality samples for the remaining branches. The subsequent heads with the increasing IoU thresholds would train on more abundant positive samples and to conduct an accurate detector and avoid the problem of overfitting. The proposed method reached the best mAP value and outperformed the comparison methods by about 10%. The experimental results show that our method can significantly improve detection performance and solve the problem of lacking high-quality samples.

The deep learning (DL)-based scene classification methods have been obtained the remarkable attention for the high spatial resolution remote sensing (HRS) imagery. However, from one aspect, the existing DL methods in HRS image scene classification are usually the variations of the natural image processing methods and often the inherent network structures; from another aspect, the strenuous and significant efforts have been devoted to the design of relevant network structures by human experts. In this paper, learning from the natural evolution, the deep neural network is expected to be globally evolved by the machine for automatically adapting the structure of the HRS imagery, a multi-objective neural architecture search based HRS image scene classification method is proposed (RSSM-Net). The two objectives of minimizing a classification error and the computational complexity have been simultaneously optimized through the evolutionary multi-objective method, the competitive neural architectures in a Pareto solution set are then obtained. The effectiveness is proved by the experiment of the UC Merced dataset with several networks designed by human experts.

High-resolution remote sensing (HRRS) image scene classification takes an important role in many applications and has attracted much attention. Recently, notable efforts have been made to present massive methods for HRRS scene classification, wherein deep-learning-based methods demonstrate remarkable performance compared with state-of-the-art methods. However, HRRS images contain complex contextual relationships and large differences of object scale, which are significantly different from natural images. The existing deep-learning-based scene classification methods are originally designed for natural image processing and have not been optimized to adapt to the characteristics of HRRS images, which significantly affects the efficiency of the feature extraction and recognition accuracy. In addition, when designing a model for remote sensing tasks, the pretraining of the model is time-consuming. The enormous amount of pretraining time and computation resources necessarily increase the difficulty of producing an excellent model. In this letter, focusing on the problems above, we proposed a new convolutional neural network (CNN)-based scene classification method. The CNN-based scene classification method is constructed by spatial-scale-aware blocks and is efficient in extracting the abundant spatial features, but can also adaptively adjust feature responses to maximize the function of informative features in the classification results. In addition, an HRRS imagery-based learning strategy is utilized to obtain an initial model for fine-tuning the model parameters, which drastically reduces the pretraining time. The proposed method has been demonstrated using two HRRS data sets, and experimental results have proven the superiority of the proposed method.

High-resolution remote sensing (HRRS) image scene classification has attracted an enormous amount of attention due to its wide application in a range of tasks. Due to the rapid development of deep learning (DL), models based on convolutional neural network (CNN) have made competitive achievements on HRRS image scene classification because of the excellent representation capacity of DL. The scene labels of HRRS images extremely depend on the combination of global information and information from key regions or locations. However, most existing models based on CNN tend only to represent the global features of images or overstate local information capturing from key regions or locations, which may confuse different categories. To address this issue, a key region or location capturing method called key filter bank (KFB) is proposed in this article, and KFB can retain global information at the same time. This method can combine with different CNN models to improve the performance of HRRS imagery scene classification. Moreover, for the convenience of practical tasks, an end-to-end model called KFBNet where KFB combined with DenseNet-121 is proposed to compare the performance with existing models. This model is evaluated on public benchmark data sets, and the proposed model makes better performance on benchmarks than the state-of-the-art methods.

High-resolution remote sensing (HRRS) image scene classification has been a critical task and greatly important for many applications, wherein convolutional neural network (CNN)-based methods have achieved considerable improvements. However, the CNN-based methods have countered a severe problem that massive annotation samples are required to obtain ideal model for scene classification. There is no dataset with a comparative scale to ImageNet to meet the sample requirement and labelling samples is labor-intensive and time-consuming. To solve the problem of insufficient annotation samples, a new generative adversarial network (GAN)-based sample generation method for scene classification is implemented. The proposed method is able to generate HRRS images with specific label and improve scene classification performance for the CNN-based methods.

Sea-land segmentation is a key step in inshore ship detection and coast monitoring. Among the state-of-art segmentation approaches, semantic segmentation networks show great potential on this task, but there is still room for improvement. In this paper, we propose a method based on UNet for sea-land segmentation. We replace its contraction part with ResNet which specializes in handling complicated scenes, and construct a new network structure Res-UNet. After preliminary segmentation results are obtained, the fully connected Conditional Random Field (CRF) model and morphological operation are then used as post-processing to obtain more precise coastlines and intact regions. We test our model on a dataset collected from Google Earth and the inspiring results validate the effectiveness of our method.

Remote sensing image scene classification, which aims to identify the types of land cover, is a fundamental task in remote sensing image analysis. Remote sensing images contain a variety of land-cover objects. These land-cover objects form a complex and diverse scene through spatial combination and correlation, which makes remote sensing imagery scenes classification difficult. In addition, remote sensing images contain redundant information that has a negative impact on remote sensing imagery scene classification, which makes remote sensing imagery scenes classification rather challenging. Recently, there are many deep learning based methods, which have achieved remarkable performance through an end-to-end supervised training process. Existing advances in remote sensing imagery scene classification mainly focus on training multi-layer convolutional neural networks (CNNs). These CNNs do not explicitly distinguish between key information and redundant information of the image. Therefore, the ability to extract features is limited. How to focus on key information and ignore redundant information in remote sensing imagery scene classification is a valuable problem. Inspired by the attention mechanism, we propose a CNN-based network that combines residual units and attention mechanism. It automatically assigns large weights to key areas of the image and thus has the ability to adaptively ignore redundant information. We evaluated the proposed approach with some state-of-the-art methods on the UC Merced Land-Use dataset and the NWPU-RESISC45 dataset. Experimental results show that the proposed attention model has achieved the best classification performance.

Spatial regularization sparse unmixing (SRSU) has been widely studied and proved to be far better than the traditional spectral unmixing methods. These spatial sparse unmixing algorithms have obtained many competitive results except for the negative influences of inaccurate estimated unmixing abundances or outliers in abundances. In this paper, to obtain a more accurate SRSU results, a local block grouping with noise-adjusted principal component analysis method is used to do spatial preprocessing in sparse unmixing process. Here, local blocks are treated as a series of vector variables, and these variables are selected by grouping the pixels with similar local spatial structures to the underlying one in the local window. Then noise-adjusted principal component analysis (NAPCA) is taken to transform the original datasets into PCA domain and maintain only the most significant principal component as well as wipe off the inaccurate estimated fractional abundances. Compared with total variation-based and nonlocal means-based SRSU algorithms, the proposed joint local block grouping with NAPCA sparse unmixing method can yield competitive results with state-of-the-art spatial sparse unmixing algorithms using both simulated dataset and real hyperspectral imagery.

Reliably extracting information from satellite imagery is a difficult problem with many practical applications. One specific case of this problem is the task of automatically detecting roads. Road extraction from satellite images has been a hot research topic in the past decade. In this paper, we propose a semantic segmentation neural network, named D-ResUnet, which adopts U-Net structure, residual learning, and dilated convolutions for road area extraction. The network is built with ResUnet architecture and has dilated convolution layers in its center part. ResUnet architecture combines the strengths of residual units and feature concatenate, which help to ease training of networks and facilitate information propagation. Dilation convolution is a powerful tool that can enlarge the receptive field of feature points without reducing the resolution of the feature maps. We test our network and compare it with U-Net and ResUnet based road extraction methods. The proposed approach outperforms all the comparing methods, which demonstrates its superiority over recently developed state of the arts.

Subpixel mapping (SPM) of remote sensing imagery is aimed at generating a classification map with a finer spatial resolution based on the abundance maps. The sparse subpixel mapping (SSM) method reformulates the SPM problem into a spatial pattern linear regression problem based on the preconstructed subpixel patch dictionary. However, in the SSM model, the optimization of the L0-norm is a nonconvex NP-hard problem, so the L1-norm is used to replace the L0-norm to obtain an approximate solution, and the selection of the optimal weight parameter between multiple terms is difficult. Thus, in this paper, a novel multiobjective SSM (MOSSM) framework for remote sensing imagery is proposed, which transforms the SSM problem into a multiobjective optimization problem. In MOSSM, first, the sparsity term is accurately modeled using the L0-norm instead of the L1-norm to avoid the potential errors caused by the L1-norm, and an evolutionary algorithm is used to directly optimize the L0-norm. Second, a subfitness-based multiobjective evolutionary algorithm is employed to simultaneously optimize the fidelity term, the sparsity term, and the spatial prior term, and to generate a set of optimal sparse coefficients to balance these three terms. Thus, there is no need to determine sensitive weight parameters. Finally, two spatial prior terms, which can be applied to the overcomplete dictionary, are presented in the proposed MOSSM-TV and MOSSM-L algorithms to incorporate the spatial correlation of subpixels. Experiments were conducted with two synthetic images and two real data sets, and the results were compared with those of ten other SPM algorithms to demonstrate the effectiveness of the proposed method.

Spatial regularization based sparse unmixing has been attracted much attention and has achieved improved fractional abundance results. However, the traditional approach to spatial consideration can only suppress discrete wrong unmixing points and smooth an abundance map with low-contrast changes, and it has no concept of scale difference. As the different levels of structures and edges in remote sensing have different meanings and importance, to better extract the different levels of spatial details, rolling guidance based scale-aware spatial sparse unmixing (RGSU), is proposed in this paper to extract and recover the different levels important structures and details in the hyperspectral remote sensing image unmixing procedure. Differing from the existing spatial regularization based sparse unmixing approaches, the proposed method considers the different levels of edges by combining a Gaussian filter-like method to realize small-scale structure removal with a joint bilateral filtering process to account for the spatial domain and range domain correlations. The experimental results obtained with both simulated and real hyperspectral images show that the proposed method achieves a better performance and produces more accurate abundance maps, as well as higher quantitative results, when compared to the current state-of-the-art sparse unmixing algorithms

High-resolution remote sensing (HRRS) scene classification plays an important role in numerous applications. During the past few decades, a lot of remarkable efforts have been made to develop various methods for HRRS scene classification. In this paper, focusing on the problems of complex context relationship and large differences of object scale in HRRS scene images, we propose a deep CNN-based scene classification method, which not only enables to enhance the ability of spatial representation, but adaptively recalibrates channel-wise feature responses to suppress useless feature channels. We evaluated the proposed method on a publicly large-scale dataset with several state-of-the-art convolutional neural network (CNN) models. The experimental results demonstrate that the proposed method is effective to extract high-level category features for HRRS scene classification.

Sparse unmixing has been successfully applied to hyperspectral remote sensing imagery based on the assumption that the observed image signatures can be expressed in a linear sparse regression with a large standard spectral library. Prior work for sparse unmixing usually utilizes L1norm or Laplacian distribution to promote sparsity. Unfortunately, the L1norm is not differentiable, which may lead to unstable results. In this paper, we adopt Bregman divergence for sparse unmixing, which is a differentiable, smoother prior. Based on the Maximum A Posterior (MAP) estimation, the proposed method has achieved sparse, stable and precise fractional abundances. The experimental results both simulated dataset and the real hyperspectral image demonstrate the effectiveness of the proposed differentiable sparse unmixing algorithm.

Geospatial object detection from high spatial resolution (HSR) imagery is significant and challenging for further analyzing the object-related information in various civil and military applications. Traditional object detection methods based on the handcrafted features are limited by their efficiency in describing the multi-class objects from large-swath and complex-context HSR imagery. Although convolutional neural network (CNN) can extract the features automatically, the feature extraction and detection stages are still separate and time-consuming. In addition, manual labelling information is limited and an efficient real-time one-stage detection framework for HSR imagery is scare. In this paper, a robust pre-trained efficient multi-class geospatial object detection framework - pre-trained Faster R-CNN sharing the convolutional features between region proposal stage and detection stage is proposed for HSR imagery. Extensive experiments and evaluations on a ten-class object detection dataset are conducted for the proposed method.

Sparse representation theory has become a powerful tool since it can obtain the sparsest or the unique solution for the underdetermined problem with the development of linear algebra, optimization, scientific computing and more. As subpixel information extraction encountered in hyperspectral remote sensing, which contains many mixed pixels, are famous under-determined ill-posed problem. In addition, there is no unified model to conquer the problems with the subpixel analysis techniques, i.e., spectral unmixing and subpixel mapping. To cope with this under-determined problem, a unified sparse subpixel information extraction framework was proposed in this paper, which connects sparse unmixing and sparse subpixel mapping methods in a unified theoretical system as a serious of sparse regression problem. The experimental results with hyperspectral images indicate that the proposed sparse representation framework outperforms the previous subpixel analysis approaches, hence, provides an effective option for subpixel information extraction idea for hyperspectral remote sensing imagery.

Sparse unmixing has been successfully applied to hyperspectral remote sensing imagery, based on an available standard spectral library. However, as the number of hyperspectral remote sensors increases, more and more hyperspectral remote sensing images are requiring analysis without the use of a corresponding standard spectral library. To address this problem, sparse unmixing with a complete dictionary online self-learning technique is proposed in this paper. This paper focuses on complete dictionary, which can tackle the unmixing problem with exactly atoms needed in the dataset and online learning means to process the specific data, or the current single hyperspectral remote sensing imagery, at real time. The proposed method addresses the sparse unmixing problem by considering the physical meaning of atoms in the complete dictionary, as well as a non-negative constraint for the abundance. Compared with the classical dictionary learning approaches in sparse representation theory, the experiments with two simulated hyperspectral datasets and a real dataset confirmed the effectiveness of the proposed method.

Subpixel mapping, which is a promising technique based on the assumption of spatial dependence, enhances the spatial resolution of images by dividing a mixed pixel into several subpixels and assigning each subpixel to a single land-cover class. The traditional subpixel mapping methods usually utilize the fractional abundance images obtained by a spectral unmixing technique as input and consider the spatial correlation information among pixels and subpixels. However, most of these algorithms treat subpixels separately and locally while ignoring the rationality of global patterns. In this paper, a novel subpixel mapping model based on sparse representation theory, namely, adaptive sparse subpixel mapping with a total variation model (ASSM-TV), is proposed to explore the possible spatial distribution patterns of subpixels by considering these subpixels as an integral patch. In this way, the proposed method can obtain the optimal subpixel mapping result by determining the most appropriate subpixel spatial pattern. However, the number of possible spatial configurations of subpixels can increase sharply with large-scale factors, and therefore, in ASSM-TV, the subpixel mapping is considered as a sparse representation problem. A preconstructed discrete cosine transform dictionary, which consists of piecewise smooth subpixel patches and textured patches, is utilized to express the original subpixel mapping observation in a sparse representation pattern. The total variation prior model is designed as a spatial regularization constraint to characterize the relationship between a subpixel and its neighboring subpixels. In addition, a joint maximum a posteriori model is proposed to adaptively select the regularization parameters. Compared with the other traditional and state-of-the-art subpixel mapping approaches, the experimental results using a simulated image, three synthetic hyperspectral remote sensing images, and two real remote sensing images demonstrate that the proposed algorithm can obtain better results, in both visual and quantitative evaluations.

As a result of the spatial consideration of the imagery, spatial sparse unmixing (SU) can improve the unmixing accuracy for hyperspectral imagery, based on the application of a spectral library and sparse representation. To better utilize the spatial information, spatial SU methods such as SU via variable splitting augmented Lagrangian and total variation (SUnSAL-TV) and nonlocal SU (NLSU) have been proposed. However, the spatial information considered in these algorithms comes from the estimated abundance maps, which will change along with the iterations. As the spatial correlations of the imagery are fixed and certain, the spatial relationships obtained from the variable abundances are not reliable during the process of optimization. To obtain more precise and fixed spatial relationships, an improved weight calculation NLSU (I-NLSU) algorithm is proposed in this letter by changing the spatial information acquisition source from the variable estimated abundances to the original hyperspectral imagery. A noise-adjusted principal component analysis strategy is also applied for the feature extraction in the proposed algorithm, and the obtained principal components are the foundation of the spatial relationships. The experimental results of both simulated and real hyperspectral data sets indicate that the proposed I-NLSU algorithm outperforms the previous spatial SU methods.

Sparse unmixing based on sparse representation theory has been successfully applied to hyperspectral remote sensing imagery. To better utilize the abundant spatial information and improve the unmixing accuracy, spatial sparse unmixing methods such as non-local sparse unmixing (NLSU) have been proposed. Although the NLSU method utilizes the nonlocal spatial information as its spatial regularization term, and obtains a satisfactory unmixing accuracy, the final abundances are affected by the non-local neighborhoods and drift away from the true abundance values when the hyperspectral images are contaminated by strong noise. To solve this problem, a non-local Euclidean medians sparse unmixing (NLEMSU) method is proposed to improve NLSU by replacing the non-local means total variation spatial consideration with non-local Euclidean medians filtering approach. The experimental results using simulated and real hyperspectral images indicate that NLEMSU outperforms the previous sparse unmixing algorithms and, hence, provides an effective option for the unmixing of hyperspectral remote sensing imagery.