Recent advancements in neural radiance fields (NeRF) have demonstrated significant potential in 3D reconstruction and novel view synthesis. However, their application in photogrammetry remains limited due to challenges in geometric accuracy for unbounded scenes and training inefficiency. We propose Bidirectional Depth-Space Optimized NeRF (BiDS-NeRF), a framework that integrates depth estimation with space warping optimization to balance high-precision geometric reconstruction and high-fidelity rendering. A lightweight architecture leveraging multi-resolution hash encoding is first established, achieving significant training acceleration while preserving rendering quality. Building upon this foundation, an implicit surface segmentation mechanism with dynamic space warping is proposed, effectively resolving distant object representation challenges in unbounded scenes. Finally, geometric consistency is enhanced through the integration of a monocular depth estimation pre-training model, coupled with a novel bidirectional optimization strategy that synchronizes photometric and geometric constraints. Complementing our technical contributions, an open-source UAV benchmark dataset with multiple categories of unbounded scenes is established, providing quantifiable metrics for cross-scene applicability assessment. Experiments on the OMMO dataset and our UAV dataset demonstrate improvements in relevant metrics, particularly in complex geometric regions. Our method provides a practical solution for UAV-based 3D reconstruction, advancing NeRF’s applicability in large-scale scene modeling.

Land surface temperature (LST) serves as a key indicator for studying the thermal characteristics of the land-atmosphere interface. Nevertheless, cloud cover and atmospheric particulates frequently obstruct the thermal infrared (TIR) spectral range in satellite observations, leading to significant data missing in derived LST products. Multi-source data fusion methods are the most effective means for all-weather TIR LST reconstruction. However, most of the existing multi-source data fusion reconstruction methods only consider the spatial and temporal differences individually, which makes it challenging to ensure spatio-temporal consistency in LST gap-filling. In this study, a novel framework for LST reconstruction is proposed by constructing a spatio-temporal consistency-guided global-local fusion network (STCGL-Net) to fuse TIR and reanalysis data. Utilizing a deep convolutional network architecture, the STCGL-Net efficiently captures the local spatial features and texture details of multi-source observations. At the same time, spatio-temporal consistency Transformer and coordinate attention mechanisms are cleverly embedded to capture the global spatiotemporal dependencies that drive changes in LST. Using spatio-temporal consistency as the core physical constraint to guide the STCGL-Net's training process ensures that the reconstructed LST is not only numerically accurate, but also evolves in accordance with physical laws. As a key input parameter, ERA5 surface solar radiation data quantitatively characterize solar forcing on land surface thermal dynamics and enhance LST reconstruction accuracy. Evaluation demonstrates STCGL-Net achieves cloud-contaminated LST reconstruction with R²=0.87 and MAE=0.54K. Comparative evaluations with three established classic methods confirm the STCGL-Net’s consistent performance across various spatial and temporal scales. Validation against Surface Radiation Budget station LST, confirms reconstruction robustness (R²:0.8-0.9, MAE:3-5K), establishing reliable all-weather LST generation capability.

Shadows significantly reduce the available information in remote sensing images, obstructing downstream tasks such as object detection, scene classification, and localization. However, shadow removal from remote sensing images is still an open issue, for the following reasons. Firstly, deep neural networks are difficult to train since the corresponding ground truths of shadows are almost always unavailable in practice. Secondly, the existing shadow removal methods still suffer from blurry details and boundary artifacts. In this paper, we describe how a generative shadow synthesis and removal framework that couples data-driven methods with illumination models was developed to address the above challenges effectively. Various shadows were synthesized in shadow-free regions of remote sensing images by GSS-Net, which is a generative shadow synthesis network that considers the physical process of shadow illumination attenuation. In this way, a large-scale, diverse, and realistic shadow dataset (RS-SynShadow) was built. A generative shadow removal network—GSR-Net—embedding a histogram-enhanced illumination model, was then developed for high-fidelity shadow removal without artifacts. Extensive experiments conducted on synthetic and real data demonstrate that the proposed shadow synthesis and removal framework significantly outperforms the state-of-the-art methods, both visually and quantitatively. The dataset and code will be made available at https://github.com/fzzfRS/RS-GSSR.

Data quantity and quality are both critical for information extraction and analyzation in remote sensing. The current remote sensing datasets, however, often fail to meet these two requirements, for which the cloud is a primary factor degrading the data quantity and quality. This limitation affects the precision of results in remote sensing applications, particularly those derived from data-driven techniques. In this article, a physical law embedded generative cloud synthesis (PGCS) method is proposed to generate diverse,ealistic cloud images to enhance real data and promote the development of algorithms for subsequent tasks, such as cloud correction, cloud detection, and data augmentation for classification, recognition, and segmentation. The PGCS method involves two key phases: spatial synthesis and spectral synthesis. In the spatial synthesis phase, a style-based generative adversarial network is used to simulate the spatial characteristics, generating an infinite number of single-channel clouds. In the spectral synthesis phase, the atmospheric scattering law is embedded through a local statistics and global fitting method, converting the single-channel clouds into multispectral clouds. The experimental results demonstrate that PGCS achieves a high accuracy in both phases and performs better than three other existing cloud synthesis methods. Two cloud correction methods are developed from PGCS and exhibits a superior performance compared to state-of-the-art methods in the cloud correction task. The application of PGCS with data from various sensors was, furthermore, investigated and successfully extended. Code will be provided athttps://github.com/Liying-Xu/PGCS.

The task of instance segmentation in remote sensing images, aiming at performing per-pixel labeling of objects at the instance level, is of great importance for various civil applications. Despite previous successes, most existing instance segmentation methods designed for natural images encounter sharp performance degradations when they are directly applied to top-view remote sensing images. Through careful analysis, we observe that the challenges mainly come from the lack of discriminative object features due to severe scale variations, low contrasts, and clustered distributions. In order to address these problems, a novel context aggregation network (CATNet) is proposed to improve the feature extraction process. The proposed model exploits three lightweight plug-and-play modules, namely, dense feature pyramid network (DenseFPN), spatial context pyramid (SCP), and hierarchical region of interest extractor (HRoIE), to aggregate global visual context at feature, spatial, and instance domains, respectively. DenseFPN is a multi-scale feature propagation module that establishes more flexible information flows by adopting interlevel residual connections, cross-level dense connections, and feature reweighting strategy. Leveraging the attention mechanism, SCP further augments the features by aggregating global spatial context into local regions. For each instance, HRoIE adaptively generates RoI features for different downstream tasks. Extensive evaluations of the proposed scheme on iSAID, DIOR, NWPU VHR-10, and HRSID datasets demonstrate that the proposed approach outperforms state-of-the-arts under similar computational costs. Source code and pretrained models are available at https://github.com/yeliudev/CATNet.

Shadow removal is beneficial for various remote sensing applications, such as semantic segmentation and object detection. However, traditional shadow removal methods perform poorly when applied to high-resolution remote sensing images. Some deep learning-based models exhibit high-level accuracy of removal, but it is difficult to be expanded to multi-scenarios due to insufficient pairs of shadow/shadow-free data in reality. In this paper, a novel framework is proposed from shadow synthesis to shadow removal, aiming to improve the usability of deep learning-based models in the task of shadow removal in high-resolution remote sensing images. First, we combine the physical shadow illumination model and a domain alignment network to synthesize a large variety of realistic shadows. Then, a shadow removal network considering global-local features is built to restore the ground surface details finely. Numerous experiments have shown that the proposed framework can effectively remove various shadows and is superior to existing methods.

There is no "cloudy & cloud-free" paired images with totally identical surface information under the same spatial and temporal condition in reality, which limits the development of supervised deep learning methods in the field of cloud removal and detection. In this regard, a high-fidelity thin cloud synthetic method is proposed, which is more challenging than the synthesis of thick clouds. This method combines physical model and data-driven methods. The expression of scattering laws at the pixel level is extended to the channel level to synthesize multi-channel cloud from cirrus band with controlled cloud thickness. Besides, spatial and spectral features are learned from real data using generative adversarial networks and transformed to synthetic data. Based on it, a dataset containing infinite number of "cloudy & cloud-free" pairs can be constructed. Experimental results show that the proposed method has the best visual effect with highest quantitative evaluations compared with current methods.

Transportation network extraction based on deep learning has become a hotspot. However, the existing models all aim to distinguish between background and transportation network, while ignoring the class attributes within the transportation networks. In this letter, we propose a multi-class transportation network extraction network (MCTN-Net) to simultaneously extract railways, roadways, trails and bridges. Inspired by multi-task learning, the network first extracts the semantic and information together by the use of a dense feature shared encoder (DFSE). The orientation and semantic features are then fused in the orientation-guided stacking module (OGSM) to enhance the connection between transportation network pixels. Furthermore, a semantic refinement branch (SRB) is designed to improve the ability of classifying different transportation network types through deep supervised fusion and class attention. A multi-class transportation network dataset was constructed and used in the experiments. The experiential results indicate that the proposed method achieves an MIoU of 64.29% and an FWIoU of 71.20% without the background, which is significantly better than the other road extraction models and semantic segmentation methods. The code and dataset are available at https://github.com/fzzfRS/MCTN-Net.

Local climate zone (LCZ) mapping can explore the variability of the impact of urban form on the thermal environment in different urban contexts, and large-scale LCZ mapping can help us to better understand the spatial and temporal dynamics of the climate in urban areas around the world. Studies have indicated that deep learning-based methods can effectively perform LCZ classification. However, the accuracy of LCZ classification on large-scale datasets is still unsatisfactory, mainly due to the fact that the traditional convolutional neural networks are not good at mining contextual information, which is crucial for fully understanding remote sensing scenes. In this paper, to solve this problem, we propose an LCZ mapping method based on remote sensing images by coupling multi-level features mined from global and local ranges with prior knowledge, named LCZ-MFKNet. The global and local features are extracted through Swin Transformer and space-maintained ResNet (SM-ResNet) model branches, respectively, and then fused through an improved squeeze-and-excitation (iSE) module. The prior knowledge studied from the theoretical definition and experimental tests is that two typical sets of LCZ categories are easily confounded in multi-class classification but separable in two-class classification. Experiments are conducted on the large publicly available So2Sat LCZ42 dataset, where the proposed LCZ-MFKNet method achieved the highest LCZ mapping accuracy. Moreover, six megacities were selected globally for LCZ mapping, and the results verified the accuracy and the general applicability of the proposed LCZ-MFKNet method in large-scale LCZ mapping.

Shadow correction is an important task in the analysis of high-resolution remote sensing images, as the existence of shadows reduces radiometric information and causes changes in the energy distribution. This is especially the case when the spatial resolution is very high, as the shadows disturb the subsequent processing and applications, such as image mosaicking, classification, segmentation, etc. Traditional shadow correction methods are limited by the shadow detection accuracy and the available non-shaded samples in the imagery. In this paper, we propose an evolutionary shadow correction network (ESCNet) and describe how we built a benchmark unmanned aerial vehicle (UAV) image dataset to achieve shadow correction directly, without shadow detection. The proposed ESCNet is made up of two sub-networks with an evolutionary relationship: a shadow removal network (SRNet) followed by a radiation adjustment network (RANet). The shadows are first removed by SRNet trained on the UAV image dataset to achieve the primary shadow-corrected image, and the global radiation is then adjusted to a sunlit-like status by RANet. Shadow detection is not required in the proposed method, which effectively overcomes the error accumulation and shadow edge artifact problem of the traditional methods. Experiments were carried out and the results were compared with those of both traditional and deep learning-based shadow correction methods, for which both qualitative and quantitative evaluations were performed. The results suggest that the proposed method shows obvious advantages in information recovery for shadow regions, and the global brightness of the corrected imagery is consistent with that of sunlit conditions.

Cloud contamination is inevitable in optical satellite images, especially for those in visible and near-infrared (VNIR) spectra. A general thin cloud correction method for satellite VNIR images is proposed in this study by coupling statistical information with a scattering model to solve the abovementioned problem. A thin cloud map (TCM) is created by utilizing the characteristics of land surface and thin clouds to depict the thin cloud spatial distribution and relative intensity. Furthermore, different cloud reflectance estimation (CRE) algorithms are proposed for different VNIR bands by considering the scattering properties of thin clouds. For coastal and blue bands with short wavelengths, the images are divided into multiple layers to search for dark pixels based on TCM so that thin clouds can be estimated via robust regression. The CRE of the green, red, and near-infrared bands with long wavelengths is realized via the scattering model by taking the thin clouds of coastal or blue bands as a reference. Experiments are performed on cloud-covered VNIR images captured by different satellites to validate the universality of the proposed method. Two traditional methods and one deep learning method are utilized for a comparison. Compared with the benchmark methods, the proposed method yields totally cloud-free images and more credible color. The quantitative measures obtained by the proposed method are the closest to the ideal values among the four methods. Discussions of the novelties of the proposed method, the extended applications of TCM, and the parallax problem in experiments is also performed to complete the evaluation.

In the task of change detection (CD), high-resolution remote sensing images (HRSIs) can provide rich ground object information. However, the interference from noise and complex background information can also bring some challenges to CD. In recent years, deep learning methods represented by convolutional neural networks (CNNs) have achieved good CD results. However, the existing methods have difficulty in detecting the detailed change information of the ground objects effectively. The imbalance of positive and negative samples can also seriously affect the CD results. In this letter, to solve the above problems, we propose a method based on a multiscale fully convolutional neural network (MFCN), which uses multiscale convolution kernels to extract the detailed features of the ground object features. A loss function combining weighted binary cross-entropy (WBCE) loss and dice coefficient loss is also proposed, so that the model can be trained from unbalanced samples. The proposed method was compared with six state-of-the-art CD methods on the DigitalGlobe dataset. The experiments showed that the proposed method can achieve a higherF1-score, and the detection effect of the detailed changes was better than that of the other methods.

Visible remotely sensed images usually suffer from the haze, which contaminates the surface radiation and degrades the data quality in both spatial and spectral dimensions. This study proposes a spatial-spectral adaptive haze removal method for visible remote sensing images to resolve spatial and spectral problems. Spatial adaptation is considered from global and local aspects. A globally nonuniform atmospheric light model is constructed to depict spatially varied atmospheric light. Moreover, a bright pixel index is built to extract local bright surfaces for transmission correction. Spectral adaptation is performed by exploring the relationships between image gradients and transmissions among bands to estimate spectrally varied transmission. Visible remote sensing images featuring different land covers and haze distributions were collected for synthetic and real experiments. Accordingly, four haze removal methods were selected for comparison. Visually, the results of the proposed method are completely free from haze and colored naturally in all experiments. These outcomes are nearly the same as the ground truth in the synthetic experiments. Quantitatively, the mean-absolute-error, root-mean-square-error, and spectral angle are the smallest, and the coefficient-of-determination (R2) is the largest among the five methods in the synthetic experiments. R2, structural similarity index measure, and the correlation coefficient between the result of the proposed method and the reference image are closest to 1 in the real data experiments. All experimental analyses demonstrate that the proposed method is effective in removing haze and recovering ground information faithfully under different scenes.

Due to the rapid urbanization, the impact of urban spatial form on surface temperature cannot be ignored and is becoming increasingly significant. This study investigates the impact of the urban spatial form which includes building form and urban land surface moisture (ULSM) on the land surface temperature (LST) of communities in Wuhan, China. The LST was retrieved in the summer and winter by Landsat 8 OLI/TIRS data. Thirty typical communities in the three-ring area were selected to represent general residential area in Wuhan. By extracting the building height (BH), building density (BD), floor area ratio (FAR), sky view factor (SVF), frontal area index (FAI) and ULSM of each community, the result shows that BH, BD, FAR, SVF and ULSM are critical variables in lowering LST in both winter and summer.

The optical remote sensing images not only have to make a fundamental tradeoff between the spatial and spectral resolutions, but also are inevitable to be polluted by the clouds; however, the existing pansharpening methods mainly focus on the resolution enhancement of the optical remote sensing images without cloud contamination. How to fuse the cloud-contaminated images to achieve the joint resolution enhancement and cloud removal is a promising and challenging work. In this paper, a pansharpening method for the challenging cloud-contaminated very high-resolution remote sensing images is proposed. Furthermore, the cloud-contaminated conditions for the practical observations with all the thick clouds, the thin clouds, the haze, and the cloud shadows are comprehensively considered. In the proposed methods, a two-step fusion framework based on multisource and multitemporal observations is presented: 1) the thin clouds, the haze, and the light cloud shadows are proposed to be first jointly removed and 2) a variational-based integrated fusion model is then proposed to achieve the joint resolution enhancement and missing information reconstruction for the thick clouds and dark cloud shadows. Through the proposed fusion method, a promising cloud-free fused image with both high spatial and high spectral resolutions can be obtained. To comprehensively test and verify the proposed method, the experiments were implemented based on both the cloud-free and cloud-contaminated images, and a number of different remote sensing satellites including the IKONOS, the QuickBird, the Jilin (JL)-1, and the Deimos-2 images were utilized. The experimental results confirm the effectiveness of the proposed method.

The presence of shadows in urban aerial images can degrade the quality of the images and cause problems in image interpretation. In this paper, a novel shadow removal method based on the clustering correction of illumination field is proposed. We construct a spatially adaptive weighted total variation model to achieve the optimized illumination field. The land surface types are considered to correct the distribution of the illumination field based on the clustering moment matching method. Then the shadows are recovered in the illumination field with the texture well preserved and the shadow boundary smoothed. Experiments and comparisons are presented to verify the effectiveness of the proposed method.

Remote sensing images are widely used in various fields. However, they are usually degraded by haze. In this paper, a spatial-spectral adaptive haze removal method for remote sensing images is proposed. The proposed method is based on haze imaging model and the dark channel prior. Our method is able to diminish the phenomenon of color distortion and haze removed not completely in different bands. The experimental results verify that the proposed method can remove the haze completely and yield visually haze free images, even the haze distribution is uneven.

The cloud and cloud shadow are difficult to capture accurately in optical imagery because of insufficient spectral information. In this paper, an automatic multiple features combined (MFC) method is proposed for cloud and cloud shadow detection in GF-1 WFV imagery which includes three visible and one near-infrared bands. The local optimization strategy with guided filtering, and the proposed object-based filter combining geometry and texture features are used in the proposed method to refine cloud detection results and exclude non-cloud bright objects. The experimental results indicate that MFC performs well under different conditions.

Impervious surface has been recognized as a key indicator in assessing urban environments. Referring to the previous research, linear spectral mixture analysis has been widely used to extract impervious surface. In this paper, a material-based endmember selection is applied to support linear spectral unmixing, which suggests that the impervious surface should be classified by their essential impervious materials. Taking Landsat images of Wuhan city for experiment, the results show that the classification accuracy is around 95%. Besides, the extracted impervious surface distribution is highly similar to the ground truth and its variation possesses a similar tendency with Urban Heat Island Intensity.

Most of the empirical topographic correction methods are based on the universal assumptions of the relationship between radiations and solar incident angles. The correction accuracy is hardly to be accessed quantitatively. This paper introduces a land cover adaptive C (LCAC) method for topographic correction, and verifies its advantage quantitatively. Experiments on synthetic and real remote sensing data are performed. The synthetic data derives from the SMARTS2 model, which simulates the surface reflectance and the atmospheric conditions. The global land cover map produced by the National Geomatics Center of China is taken as the auxiliary data. The LCAC outperforms the traditional C method in experiments on both synthetic and real remote sensing data by visual and quantitative assessments.

In this paper, a unified framework for the spatio-temporal-spectral fusion of remote sensing images is proposed. The relationships between the observed images and the desired image are first established based on general image observation models. Maximum a posteriori (MAP) theory is then employed to formulate the unified fusion framework. The proposed method is able to fuse images from an arbitrary number of optical sensors with different spatial, temporal, and spectral resolutions. The experimental results verify the effectiveness of the proposed method.

In the past decade, the synthesis-based methods have drawn people's attention more and more in the sparse representation community. The synthesis model decomposes the data into a combination of a few atoms of the overcomplete dictionary. However, the dual analysis-based methods have not been studied deeply. The analysis model results in a sparse outcome by multiplying an analysis dictionary. This work proposes an analysis-based recovery of the missing information of remote sensing data, by extracting supplementary information from another term of data at a different period. Our method is verified by the qualitative and quantitative assessments in the experiments.

Land surfaces are commonly obstructed by haze in remote sensing images, which reduces the available land cover information. Haze detection is therefore important for locating, avoiding, or restoring hazy regions. In this letter, a principal component (PC)-based haze masking (PCHM) method is developed for the masking of haze in visible remote sensing images covering land surfaces at middle latitudes. Owing to the evidence of haze in the second PC, the PCHM method results in accurate haze masks. The complete procedure comprises two steps: haze construction and spatial optimization. The validity of the PCHM method is demonstrated through its application to several hazy visible images clipped from Landsat Enhanced Thematic Mapper Plus scenes. The quantitative assessments verify the superiority of the proposed method over the haze optimized transformation method for the production of binary haze masks. In addition, the resulting haze masks are compared with a MODIS cloud product, which further proves the necessity and validity of the proposed method.

Shadows are evident in most aerial images with high resolutions, particularly in urban scenes, and their existence obstructs the image interpretation and the following application, such as classification and target detection. Most current shadow removal methods were proposed for natural images, whereas shadows in remote sensing images show distinct characteristics. We have therefore analyzed the characteristics of shadows in aerial images, and in this paper, we propose a new shadow removal method for aerial images, using nonlocal (NL) operators. In the proposed method, the soft shadow is introduced to replace the traditional binary hard shadow. NL operators are used to regularize the shadow scale and the updated shadow-free image. Furthermore, a spatially adaptive NL regularization is introduced to handle compound shadows. The combination of the soft shadow and NL operators yields satisfying shadow-free results, preserving textures and holding regular color. Different types of shadowed aerial images are employed to verify the proposed method, and the results are compared with two other methods. The experimental results confirm the validity of the proposed method and the advantage of the soft-shadow approach.

Perceptually inspired color correction methods are characterized by human visual system properties. In this paper, we propose a perceptually inspired variational method for uneven intensity correction of remote sensing images. The proposed method shares the same intrinsic scheme as the Retinex theory, but the reflectance in this method is solved directly within the limited dynamic range and is supposed to comply with the gray world assumption. Considering the smoothness of illumination and the complexity of reflectance, the proposed method integrates L2 norm and total variation prior to inflict varying constraints to different components and regions. The minimum of this variational model is calculated using the steepest descent approach. Experimental results are presented to validate the effectiveness of the proposed method.