Spectral-spatial feature extraction widely serves as the foundation for hyperspectral image classification (HSIC). However, its effectiveness decreases when applied to land covers type with temporal variations. This limitation arises from the lack of the temporal dimension in existing HSIC methods, hindering their ability to model real-world surface dynamics. To tackle these problems, Bi-tEmporal HyperspectrAL image classiFication network (BehalfNet) employs the dual-branch stacked architecture to process bi-temporal images, learning spectral-spatial-temporal features. Within each stacked block, features undergo the sequential feature processing pipeline. Specifically, the progressive adaptive fusion (PAF) module first extracts foundational spectral-spatial features for each temporal phase through long-short term fusion. These features are then refined at an intra-temporal level by the gated spectral-spatial attention (GSSA) module. Subsequently, the bi-temporal self-cross attention (BTSCA) module effectively captures the complex dynamic changes between the bi-temporal features using a novel closed-loop attention mechanism. Furthermore, the Anji dataset is introduced as the first publicly available dataset for bi-temporal HSIC. Comprehensive experiments on the Anji dataset and public Viareggio dataset (originally used for anomaly change detection) demonstrate the competitiveness of the proposed BehalfNet over other state-of-the-art HSIC methods. The code and Anji dataset will be released at https://github.com/lixinghua5540/BehalfNet.

Optical remote sensing images are frequently contaminated by thin clouds, thus causing great challenges for subsequent applications. To address this issue, numerous methods guided by cloud features have been developed. However, the cloud features utilized in these methods are generally either unlearnable or lack cloud thickness data constraints, which may further mislead the cloud removal. In this paper, a THIcknEss Fused thin cloud removal network (ThiefCloud) with self-supervised learnable cloud prior is proposed. Firstly, in order to provide reliable cloud prior, a self-supervised cloud prior model (SCPM) is introduced. Secondly, an adaptive feature extraction (AFE) module efficiently extracts the cloud information of the original cloud image, and a physically guided feature fusion (PGFF) module, inspired by the atmospheric scattering model, accurately restores more realistic details. Finally, to enhance the generalizability of SCPM in real scenarios, a staged training strategy is adopted. SCPM is trained independently on the simulated thickness maps and cloud images in advance, then SCPM can guide ThiefCloud. During the training of ThiefCloud, SCPM is frozen initially and then tunable. The frozen SCPM provides effective cloud prior to the non-converged ThiefCloud. The tunable SCPM makes the cloud prior learnable, better aligning with real-world cloud removal. Experimental results demonstrate that compared with other 11 methods, ThiefCloud could achieve competitive results on three public datasets, namely T-CLOUD, RICE and SateHaze1k datasets. The implementation code and data will be available soon at: https://github.com/lixinghua5540/ThiefCloud.

In the past few decades, significant progress on object detection has been made. However, there are still limitations when it comes to the detection of super tiny objects. To address the issue of information loss in deep learning network, this study proposes a novel super tiny object detection network (STODNet) based on YOLOv7, which belongs to one-stage object detection method. A light-backbone with heavy-neck strategy is adopted to maintain the size of feature maps and alleviate information loss with the deepening of layers. Meanwhile, to ensure the computational efficiency, object mask-guided sparse convolution (OMSC) is designed, which only applies convolutions to regions of interest, thus the computational overhead of the network is greatly reduced. Compared to state-of-the-art methods, STODNet achieves best results on the AI-TOD dataset.

Cloud detection in remote sensing images is a challenging task that plays a crucial role in various applications. A novel residual dual U-shape network (RD-UNet) is proposed for cloud detection. The primary innovation lies in that it cascades two U-shaped networks and leverages a residual-like connection to enhance information flow between two networks, thus optimizing details and edge information. Moreover, another contribution is that the improved skip connections (ISCs) efficiently facilitate multiscale feature utilization, aiding in the identification of thin clouds and distinguishing other confounding land features. The effectiveness of RD-UNet was demonstrated through extensive experiments on two public datasets, outperforming state-of-the-art methods with a nearly 2% improvement in F1 score and superior visual effect for multispectral images. The code will be available athttps://github.com/lixinghua5540/RD-UNet.

Large-ratio cloud occlusion significantly hampers the utilization of high-resolution remote sensing imagery. The existing reconstruction methods (1) overlook the problem of reconstructed and composite images sharing high-and low-level semantic and visual attributes in non-reconstructed regions, exacerbating the pronounced boundary effects; (2) neglect appearance discrepancies between reconstructed and non-reconstructed regions, leading to spectral degradation, and texture loss; and (3) overlook the problem of reconstructing large-ratio missing information. To address these issues, a global and local dual fusion network is proposed in this study for large-ratio cloud occlusion removal in high-resolution remote sensing images. The global foreground–background aware attention module tackles shared high-level semantic features, whereas the local visual feature enhancement module addresses appearance differences. The global and local dual fusion network combines the Sobel and reconstruction loss functions for effective reconstruction by employing a two-stage fusion strategy. Compared to the classical recurrent feature reasoning network, spatiotemporal generator network, spatial-temporal-spectral convolutional neural network, and bishift network, the proposed model demonstrates superior quantitative and visual reconstruction outcomes for the 40%, 50%, and 70% missing ratios of Gaofen-1 (2 m).

The spatiotemporal fusion is a valid way to provide Earth observation applications with remote sensing images with both high temporal and high spatial resolution. Along with the advancement of remote sensing technology, many spatiotemporal fusion methods have been proposed in recent years. However, the existing public spatiotemporal fusion datasets are almost all composed of moderate-resolution imaging spectroradiometer (MODIS) and Landsat images, belonging to the medium resolution datasets. The spatial resolution difference between MODIS and Landsat is up to 16 times, resulting in some texture details indeed difficult to predict. Therefore, these datasets are inadequate for assessing the detail recovery ability of spatiotemporal fusion models accurately. To address this problem, a region with abundant features and significant changes at the junction of Hongshan District and Jiangxia District in Wuhan, Hubei Province, China was selected. For the first time, a high-resolution dataset consisting of Gaofen (GF) images and Landsat images was constructed, comprising eight pairs of images spanning more than seven years. Seven fusion methods were selected to be tested on the constructed Wuhan dataset, aiming to explore the detail recovery capacity of these methods and their adaptability to different data sources. The dataset will be available athttps://github.com/lixinghua5540/Wuhan-dataset.

Remote sensing (RS) images change detection (CD) plays a crucial role in monitoring surface dynamic. However, current deep learning (DL)-based CD methods still suffer from pseudo changes and scale variations due to inadequate exploration of temporal differences and under-utilization of multiscale features. Based on the aforementioned considerations, a spatiotemporal enhancement and interlevel fusion network (SEIFNet) is proposed to improve the ability of feature representation for changing objects. Firstly, the multilevel feature maps are acquired from Siamese hierarchical backbone. To highlight the disparity in the same location at different times, the spatiotemporal difference enhancement modules (ST-DEM) are introduced to capture global and local information from bitemporal feature maps at each level. Coordinate attention and cascaded convolutions are adopted in subtraction and connection branches, respectively. Then, an adaptive context fusion module (ACFM) is designed to integrate interlevel features under the guidance of different semantic information, constituting a progressive decoder. Additionally, a plain refinement module and a concise summation-based prediction head are employed to enhance the boundary details and internal integrity of CD results. The experimental results validate the superiority of our lightweight network over 8 state-of-the-art (SOTA) methods on LEVIR-CD, SYSU-CD and WHU-CD datasets, both in accuracy and efficiency. Also, the effects of different types of backbones and differential enhancement modules are discussed in the ablation experiments in details. The code will be available at https://github.com/lixinghua5540/SEIFNet.

Surface soil moisture plays a pivotal role in various hydrological processes. Precisely assessing soil moisture with high resolution is crucial for effective water resource management, informed agricultural decision-making, and in-depth climate change research. While passive microwave remote sensing is a primary technology for regional soil moisture monitoring, its practical application is hindered by data discontinuity and low resolution. To address these challenges, we propose an integrated learning framework to enhance the continuity and resolution of soil moisture data across China. Leveraging low-resolution passive microwave soil moisture data, moderate-resolution assimilated soil moisture data, and multiple high-resolution ancillary inputs, the network effectively captures the spatiotemporal dynamics of soil moisture through integrating gap-filling, multisource fusion, and spatial downscaling processes. Validation against in situ data demonstrates the significant enhancements achieved by the proposed method, with an average R value of 0.706 and an average unbiased root mean square error of 0.055 m3/m3. Comparative analysis further confirms its superior accuracy and robustness across diverse regions. These findings highlight the potential of this integrated learning framework to advance hydrological applications, enhance agricultural production, and support climate research.

Change detection (CD) has become an attractive research topic in the field of remote sensing imagery in recent years. Despite significant advancements driven by deep learning (DL) techniques, most current methods predominantly rely on fully supervised strategies. These methods require the collection of a large number of pixel-level labels, which is quite time consuming and label intensive. To address that, we propose a weakly supervised CD method with EfficientSAM (ESAM)-CD termed, which leverages multiscale class activation map (CAM) fusion and a fine-tuned EfficientSAM’s image encoder. First, we construct a classification model employing image-level labels with a deep supervision strategy to generate high-quality multiscale CAM. Subsequently, a multiscale CAM fusion module is proposed to refine the boundaries of change targets by harnessing information from various scales. Then, we utilize EfficientSAM with powerful generalization capabilities as the backbone and fine-tune it using a low-rank adaptation (LoRA) strategy to establish a CD network. In such a network, bitemporal images and the generated pseudolabels are fed into the network. In addition, to overcome the reliance of EfficientSAM’s decoder on prompts, we propose a prompt-free decoder based on the general convolutional layers to predict change maps. Finally, we validate the effectiveness of the proposed ESAM-CD using two publicly available CD datasets (i.e., WHU-CD and LEVIR-CD). Comprehensive experiments demonstrate that our method outperforms other weakly supervised CD methods, achieving outstanding performance on both datasets.

Because of the presence of clouds, the available information in optical remote sensing images is greatly reduced. These temporal‐based methods are widely used for cloud removal. However, the temporal differences in multitemporal images have consistently been a challenge for these types of methods. Towards this end, a bishift network (BSN) model is proposed to remove thick clouds from optical remote sensing images. As its name implies, BSN is combined of two dependent shifts. Moment matching (MM) and deep style transfer (DST) are the first shift to preliminarily eliminate temporal differences in multitemporal images. In the second shift, an improved shift net is proposed to reconstruct missing information under cloud covers. It introduces multiscale feature connectivity with shift connections and depthwise separable convolution (DSC), which can capture local details and global semantics effectively. Through experiments with Sentinel‐2 images, it has been demonstrated that the proposed BSN has great advantages over traditional methods and state‐of‐the‐art methods in cloud removal.

Seamline detection is one of the most important issues in mosaicking high-resolution remote sensing images (HRRSIs). However, it is difficult to make a balance between efficiency and accuracy. On that account, a shortest matrix path-based dynamic programming (SMP-DP) algorithm is proposed to find the optimal seamline for HRRSI mosaicking. First, a pixel cost matrix defined by intensity difference, gradient similarity, and geometric difference is constructed in the overlapping area. Second, the least average path cost from the starting pixel to each pixel is calculated and the SMP-DP algorithm is applied to find the optimal path. Experimental results on HRRSI prove that the proposed method detects high-quality seamline and crosses much fewer objects with the highest computational efficiency, compared with the state-of-the-art method and commercial software.

High-resolution nighttime light (HR-NTL) imagery is a reliable indicator of human activity in urban areas at night. However, HR-NTL images on Earth are blurred because of issues such as blooming, noise, and overexposure in the imaging system and transmission path. Toward this end, a nighttime light image deblooming network (NTL-Db-Net) is proposed, which aims to make the Earth clearer at night. The NTL-Db-Net employs a generative adversarial network (GAN) as its deblooming framework. As the lack of ideal clear NTL data for GAN training, our method integrates blur transfer algorithm to simulate degradation for generating training data. This method was tested using HR-NTL data on Wuhan, Macau, Changchun, Xiamen, Atlanta and Jerusalem from HR Jilin-1 satellite constellation. Compared with the results of state-of-the-art SEAM, Uformer, and DRBNet, those from NTL-Db-Net demonstrated superior visual features and achieved the highest scores across all selected non-reference evaluation metrics. NTL-Db-Net also exhibited a lower number of network parameters and a smaller time cost. In addition, this study conducted a spatial clustering analysis of nighttime light with road networks and permeable surfaces. The results indicated that nightlights and roads exhibited a distinct spatially aggregated distribution, while permeable surfaces displayed a clear spatially dispersed distribution. Code and data are available at https://github.com/lixinghua5540/NTL-Db-Net.

Spatio-temporal fusion can generate time-series images with high spatial resolution, and it is highly desirable in various applications, especially in monitoring fine dynamic changes of surface features on remote sensing images. Currently, most spatio-temporal fusion methods predict the target fine image by employing the auxiliary fine images on neighboring phases; however, they are generally limited in abrupt land cover changes between the target and the neighboring auxiliary images. In this paper, we propose a novel Detail Injection-based Spatio-Temporal Fusion (DISTF) model to alleviate this problem, by exploring the inherent relationship between the spatio-temporal fusion and spatio-spectral fusion. The proposed DISTF consists of three modules: a Three-branch Detail Injection (TDI) module, a Fine Detail Prediction (FDP) module, and a reconstruction module. The interpretable TDI module is inspired by spatio-spectral fusion, aiming to inject the non-changed detail information extracted from the neighboring fine images into the target coarse image, which can preserve the abrupt change information captured in the target coarse image. The FDP module is designed to further integrate the correlated information from the outputs of TDI and refine the spatial-spectral information to boost the fusion accuracy. Finally, the reconstruction module and the hybrid loss function are designed to more effective reconstruct the high-quality target fine image. The qualitative and quantitative experiment results on two datasets with different types of changes demonstrated that the proposed DISTF method achieves richer spatial detail and more accurate prediction than the eight existing methods.

Precipitation is a critical process in the terrestrial hydrological circulation, affecting climate change, water resource management, and agricultural production. Satellite-borne observations have prominent advantages in macro and mesoscopic quantitative precipitation estimation. Nevertheless, they are subject to low spatial resolution and inherent biases. Therefore, this study utilizes the surface-surface downscaling network and point-surface fusion network for fine-resolution and high-precision precipitation mapping over China. To deeply explore the complicated relationships between various ancillary factors, ground measurements and satellite precipitation, an attention mechanism based convolutional network (AMCN) is used for spatial downscaling and a geo-intelligent deep belief network (Geoi-DBN) is used for ground-satellite fusion. Experimental results indicate that cascaded networks toward two different objectives are superior to baseline methods, achieving R2 and RMSE of about 0.84 and 27.23 mm/month, respectively. Besides, the assistance of geo-intelligent items and ancillary factors contributes to fusion accuracy. This study provides an effective way for precipitation estimation over China.

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.

At the end of 2019, the very first COVID-19 coronavirus infection was reported and then it spread across the world just like wildfires. From late January to March 2020, most cities and villages in China were locked down, and consequently, human activities decreased dramatically. This letter presents an “offline learning and online inference” approach to explore the variation of PM2.5 pollution during this period. In the experiments, a deep regression model was trained to establish the complex relationship between remote sensing data andin situPM2.5 observations, and then the spatially continuous monthly PM2.5 distribution map was simulated using the Google Earth Engine platform. The results reveal that the COVID-19 lockdown truly decreased the PM2.5 pollution with certain hysteresis and the fine particle pollution begins to increase when advancing resumption of work and production gradually.

For radiometric normalization (RN) of multiple remote sensing (MRS) images within large-scale coverage, the traditional methods ignore the error accumulation and adaptive allocation of cumulative errors caused by the transfer paths in the classical one-after-another pipeline. To this end, a block adjustment-based RN method of MRS images is proposed by considering the global and local radiometric differences (RDs) in this letter. First, the block adjustment-based global RN is conducted to eliminate the global differences of MRS images. This step is independent of transfer paths so that it breaks through the corresponding error accumulation and uneven distribution in the one-after-another pipeline. Second, two local strategies based on block adjustment and edge optimization are further adopted to remove the local residual RDs. In the experiments, it demonstrates that the proposed method can obtain MRS images with a balanced and appealing visual effect, which outperforms the moment matching (MM) method and the popular ENVI software.

With the fast development of remote sensing platforms and sensors technology, change detection with heterogeneous remote sensing images (Hete-CD) has become an attractive topic in recent years and plays a vital role in land cover change detection for responding to natural disaster emergencies when homogeneous images are unavailable. Although Hete-CD has been developed for about three decades, and various related methods have been developed and applied successfully in practice, a systematic and comprehensive review of the current achievements regarding Hete-CD remains lacking. Therefore, in this article, we first present an overview of Hete-CD in terms of the related literature. Second, the major techniques of Hete-CD are reviewed in terms of publicly available datasets, the taxonomy of major techniques, results, performance, and quantitative evaluation. Then, some classical methods are selected for comparison and discussion. Finally, based on the discussion and literature review, challenges, opportunities, and future directions for Hete-CD are concluded. The review aims to provide a “one-stop-shop” understanding of the problems with the categories of existing approaches, open opportunities and challenges, and potential future directions for Hete-CD.

The data fusion technology aims to aggregate the characteristics of different data and to obtain products with multiple data advantages. To solve the problem of reduced resolution of polarimetric synthetic aperture radar (PolSAR) images due to system limitations, we propose a fully PolSAR images and single-polarization synthetic aperture radar (SinSAR) images fusion network to generate high-resolution PolSAR (HR-PolSAR) images. To take advantage of the polarimetric information of the low-resolution PolSAR (LR-PolSAR) images and the spatial information of the high-resolution single-polarization SAR (HR-SinSAR) images, we propose a fusion framework for joint LR-PolSAR images and HR-SinSAR images and design a cross-attention mechanism to extract features from the joint input data. Besides, based on the physical imaging mechanism, we designed the PolSAR polarimetric loss functions for constrained network training. The experimental results confirm the superiority of the fusion network over traditional algorithms. The average peak signal-to-noise ratio (PSNR) is increased by more than 3.6 dB, and the average mean absolute error (MAE) is reduced to less than 0.07. Experiments on polarimetric decomposition and polarimetric signature show that it maintains polarimetric information well.

Due to the tradeoff between spatial and temporal resolutions of remote sensing images, spatiotemporal fusion models were proposed to synthesize the high spatiotemporal image series. Currently, spatiotemporal fusion models usually employ one coarse-resolution image acquired on a prediction date and at least another pair of coarse–fine resolution images close to the prediction time as references to derive the fine-resolution image on the prediction date. After years of development, the model accuracy has gained a certain improvement, but nearly, all the models require at least three image inputs and rigid time constraints must be applied to the references to guarantee the fusion accuracy. However, it is not always that easy to collect adequate data pairs for fine-resolution image series simulation in practice because of the bad weather condition or the time inconsistency between the coarse–fine resolution data sources, which causes some difficulties in the actual application. This article introduces the conditional generative adversarial network (CGAN) and switchable normalization technique into the spatiotemporal fusion problem and proposes a flexible deep network named the GAN-based SpatioTemporal Fusion Model (GAN-STFM) to reduce the number of model inputs and broke the time restriction on reference image selection. The GAN-STFM just needs a coarse-resolution image on the prediction date and another fine-resolution reference image at an arbitrary time in the same area for model inputs. As far as we know, this is the first spatiotemporal fusion model that requires only two images as model inputs and puts no restriction on the acquisition time of references. Even so, the GAN-STFM performs on par or better than other classical fusion models in the experiments. With this improvement, the data preparation for spatiotemporal fusion tends to be much easier than before, showing a promising perspective for practical applications.

Clouds greatly reduce the available ground information in optical remote sensing images. This paper proposed a double shift network to remove the thick clouds from multitemporal remote sensing images. The proposed networks are divided into two shift steps. In the first shift, the moment match and style transfer play the role of multi-temporal image normalization to obtain more reliable training images. In the second shift, in order to improve the network architecture's ability of capturing global semantics and local details, the shift connection layer and depthwise separable convolutions are introduced into U-Net. These two shift steps can not only improve the visual effect of cloud removal, but also further improve the quantitative evaluation. Experiments prove that the double shift network shows great advantages in cloud removal.

Because of the internal malfunction of satellite sensors and poor atmospheric conditions such as thick cloud, the acquired remote sensing data often suffer from missing information, i.e., the data usability is greatly reduced. In this paper, a novel method of missing information reconstruction in remote sensing images is proposed. The unified spatial-temporal-spectral framework based on a deep convolutional neural network (CNN) employs a unified deep CNN combined with spatial-temporal-spectral supplementary information. In addition, to address the fact that most methods can only deal with a single missing information reconstruction task, the proposed approach can solve three typical missing information reconstruction tasks: (1) dead lines in Aqua Moderate Resolution Imaging Spectroradiometer band 6; (2) the Landsat Enhanced Thematic Mapper Plus scan line corrector-off problem; and (3) thick cloud removal. It should be noted that the proposed model can use multisource data (spatial, spectral, and temporal) as the input of the unified framework. The results of both simulated and real-data experiments demonstrate that the proposed model exhibits high effectiveness in the three missing information reconstruction tasks listed above.

Registration is a classical problem in the application of remote sensing images. The existing methods prefer to fit the relationship between target and source images with the same model on the whole. In fact, the geometrical relationship between two images is not always consistent, especially for the wide-field-viewed images of GaoFen-1 (GF-1) launched by the China Aerospace Science and Technology Corporation (CASC) in April 2013. Generally, The existing methods didn't take the local deformation into consideration. Towards this end, we solve the problem with three stages in this paper. Firstly, the coarse registration obtains the integral perspective transformation model of images. Secondly, the fine registration partitions image into many blocks and improves the relationship of every block with the inverse distance weighting (IDW) function. Finally, the coordinate transformation and resampling are the final step. Compared to other methods, the experiments demonstrate that the proposed algorithm is capable of generating satisfied results which are robust against deformation at local area.

Soil Moisture Active Passive (SMAP) satellite mission, launched on Jan. 31, 2015, can provide a 9 km soil moisture product globally by merging passive and active observations. However, the radar sensor of SMAP was failed since Jul. 7, 2015 and SMAP 9 km SM product (SMAP_AP) is only available for 85 days. To ameliorate the vacancy of SMAP_AP, a spatio-temporal fusion model STNLFFM combined with the SMAP 36 km soil moisture product (SMAP_P) is utilized to generate 9 km soil moisture SM product (SMAP_F). Generation of SMAP_F was implemented over one year from Apr. 13, 2015 to Apr. 12, 2016 in the paper. Then SMAP_F was evaluated by SMAP_AP and in-situ soil moisture from international soil moisture network. It is revealed that the STNLFFM can be taken as an effective method for SMAP 9 km soil moisture generation.

Remote sensing images often suffer from cloud cover. Cloud removal is required in many applications of remote sensing images. Multitemporal-based methods are popular and effective to cope with thick clouds. This paper contributes to a summarization and experimental comparison of the existing multitemporal-based methods. Furthermore, we propose a spatiotemporal-fusion with poisson-adjustment method to fuse multi-sensor and multitemporal images for cloud removal. The experimental results show that the proposed method is able to obtain more accurate results than the current multitemporal-based methods, especially when the multi-temporal images suffer from significant changes.

Mapping and monitoring urban area extents is a very relevant task for many applications related to risks, economic activities, population mapping and the interaction between people and the natural environment. This task can be accomplished using VHR and HR optical data. The main drawback of using the optical data is the irregular presence of the clouds, making mapping by these sensors completely useless in many cases. Even when the clouds are small and irregular, the extracted urban area extents may be erroneous and/or incomplete, and many uncertainties show up in thematic maps. In this paper, we prove that the joint use of algorithms for urban extent extraction, cloud detection and masking, and image reconstruction provide a clear processing chain able to overcome this issue and to increasing the reliability of the extracted urban area extents.

Global or regional environmental change is of wide concern. Extensive studies have indicated that long-term vegetation cover change is one of the most important factors reflecting environmental change, and normalized difference vegetation index (NDVI) time-series data sets have been widely used in vegetation dynamic change monitoring. However, the significant residual effects and noise levels impede the application of NDVI time-series data in environmental change research. This study develops a novel and robust filter method, i.e., the moving weighted harmonic analysis (MWHA) method, which incorporates a moving support domain to assign the weights for all the points, making the determination of the frequency number much easier. Additionally, a four-step process flow is designed to make the data approach the upper NDVI envelope, so that the actual change in the vegetation can be detected. A total of 487 test pixels selected from SPOT VEGETATION 10-day MVC NDVI time-series data from January 1999 to December 2001 were used to illustrate the effectiveness of the new method by comparing the MWHA results with the results of another four existing methods. Finally, the long-term SPOT VEGETATION 10-day maximum-value compositing (MVC) NDVI time series for China from April 1998 to May 2014 was reconstructed by the use of the proposed method, and a test region in China was utilized to validate the effectiveness of the proposed MWHA method. All the results indicate that the reconstructed high-quality NDVI time series fits the actual growth profile of the vegetation and is suitable for use in further remote sensing applications.

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.

The Earth Observing System of the National Aeronautics and Space Administration pays a great deal of attention to the long-term global observations of the land surface, biosphere, atmosphere, and oceans. Specifically, the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on board the twin satellites Terra and Aqua plays a vital role in the mission. Unfortunately, around 70% of the detectors in Aqua MODIS band 6 have malfunctioned or failed. Consequently, many of the derivatives related to band 6, such as the normalized difference snow index, suffer from the adverse impact of dead or noisy pixels. In this letter, the missing or noisy information in Aqua MODIS band 6 is successfully completed using a robust multilinear regression (M-estimator) based on the spectral relations between working detectors in band 6 and all the other spectra. The experimental results indicate that the proposed robust M-estimator multiregression (RMEMR) algorithm can effectively complete the large areas of missing information while retaining the edges and textures, compared to the state-of-the-art methods.

With regard to quantitative remote sensing products in the visible and infrared ranges, thick clouds and accompanying shadows are an inevitable source of noise. Due to the absence of adequate supporting information from the data themselves, it is a formidable challenge to accurately restore the surficial information underlying large-scale clouds. In this paper, dictionary learning is expanded into the multitemporal recovery of quantitative data contaminated by thick clouds and shadows. This paper proposes two multitemporal dictionary learning algorithms, expanding on their KSVD and Bayesian counterparts. In order to make better use of the temporal correlations, the expanded KSVD algorithm seeks an optimized temporal path, and the expanded Bayesian method adaptively weights the temporal correlations. In the experiments, the proposed algorithms are applied to a reflectance product and a land surface temperature product, and the respective advantages of the two algorithms are investigated. The results show that, from both the qualitative visual effect and the quantitative objective evaluation, the proposed methods are effective.

Because of malfunction or noise in 15 out of the 20 detectors, band 6 (1.628-1.652 μm) of the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor aboard the Aqua satellite contains large areas of dead pixel stripes. Therefore, the corresponding high-level products of MODIS are corrupted by this periodic phenomenon. This paper proposes an improved Bayesian dictionary learning algorithm based on the burgeoning compressed sensing theory to solve this problem. Compared with other state-of-the-art methods, the proposed method can adaptively exploit the spectral relations of band 6 and other spectra. The performance of the proposed method is demonstrated by experiments on both simulated Terra and real Aqua images.