ABSTRACT This paper proposes convolutional neural network jointed with long short-time memory (CNN_LSTM) and Seq2Seq based on convolutional operation (Convolutional Seq2Seq), which the fully connected operation of Seq2Seq is replaced by convolution, and the attention mechanism of Seq2Seq is improved to monitor changes in water bodies. Convolutional Seq2Seq and CNN_LSTM can extract the temporal and

ABSTRACT In this study, we introduce a non-local block of the attention mechanism into capsule neural network (CapsNet) to form a non-local capsule network (NLCapsNet) for hyperspectral remote sensing image (HSI) classification. The presented NLCapsNet uses global information from input images and has a powerful representation of the capacity and spatial relationships among HSI features. It can effectively

ABSTRACT The present study reports a univariate predictive model for Total Ozone (TO) concentration derived from Ozone Monitoring Instrument (OMI)/Aura observations. Using the Markovian approach through proper discretization method it has been observed that the second-order Markov Chain can represent the time series of TO Concentration. Considering daily data spanning from 2015 to 2019, consisting

ABSTRACT The nonstationary clutter is one of the biggest challenge to ocean remote sensing radar system such as high frequency (HF) mixed-mode surface wave radar. The performance of space-time adaptive processing (STAP) degrades badly with limited homogeneous secondary training data support. Single dataset algorithms overcome the problem by working on primary data solely. But the heavy computational

ABSTRACT Multichannel synthetic aperture radar (SAR) systems in azimuth can realize high-resolution and wide-swath (HRWS) imaging, which has attracted intensive research interests. The azimuth ambiguities caused by non-uniform sampling can be suppressed by digital beamforming technique. However, the presence of inevitable channel errors in the multichannel SAR system deteriorates the performance of

ABSTRACT Despite significant improvement in Moderate Resolution Imaging Spectroradiometer (MODIS) aerosol optical depth (AOD) retrieval, high-resolution–high-accuracy AOD retrieval remains a challenging task. This study utilises machine learning for AOD retrieval of MODIS data. The global long-time-series data of AERONET sites and corresponding MODIS data in time and space were used as sample training

ABSTRACT The lockdowns imposed worldwide to curb the coronavirus diseases (COVID-19) spread has positive effects on the environment. However, it is unclear what fraction is caused by weather and what is related to lockdown. Here we used Global Navigation Satellite System (GNSS) height anomaly time series to quantify the spatio-temporal characteristics of lockdown-induced noise anomalies at 231 selected

ABSTRACT ABSTRACT: The two-scale Model (TSM) is broadly employed for studying EM scattering from rough sea surface due to its advantage of simple and practical. However, it suffers from a drawback in the prediction of microwave scattering from sea surface. That is the TSM cannot provide accurate predictions for hh polarized scattering coefficients. To overcome this problem, a modified TSM (MTSM) is

ABSTRACT This study evaluates various combinations of data fusion techniques at Pixel, Feature, and Decision level for crop classification using Sentinel-1 Synthetic Aperture Radar (SAR) data and Resourcesat-2 LISS (Linear Imaging Self Scanning) III, optical data for Yadgir District of Karnataka, India. For Pixel level data fusion, techniques such as brovey transformation (BT), principal component

ABSTRACT If crops were polluted by heavy metals, the spectra will change. Therefore, the variation information of spectral changes has become an important basis for heavy metal pollution monitoring. Based on spectral frequency domain, we studied the spectra of maize leaves. Combined with time-frequency analysis method, we proposed DDCR-Db (Second-order Differential Continuum Removal- Daubechies) method

ABSTRACT In the present communication, using cloud-type distribution derived from CloudSat observations during the years 2006–2010 and vertical structure of LH derived from Tropical Rain Measuring Mission (TRMM) observations during the same period over the Indian summer monsoon (ISM) region, an attempt is made to relate the both. First, the regions where particular type of clouds forms predominantly

ABSTRACT Google Earth Engine (GEE) provides a convenient platform for applications based on optical satellite imagery of large areas. With such data sets, the detection of cloud is often a necessary prerequisite step. Recently, deep learning-based cloud detection methods have shown their potential for cloud detection but they can only be applied locally, leading to inefficient data downloading time

ABSTRACT Detection and monitoring of vigour decline are certainly useful for forest managers to support future interventions. In this work a new index, from Sentinel-2 retrieved NDVI (Normalized Difference Vegetation Index) time series, is proposed, hereafter called ‘normalized Vegetation Vigour Index (nVVI)’, which is specifically designed to quantify and map tree vigour decline. Results proved that

ABSTRACT Identification and characteristics of Internal Waves (IWs) are widely used in a variety of industries, including shipbuilding, maritime, as well as oil and marine engineering. In this study, IWs have been detected using field measurement and satellite remote sensing in the Persian Gulf during the period 2000–2015. In order to identify the potential areas for IWs after vertical changes in the

ABSTRACT This study evaluated the land surface model-simulated soil moisture (SM) product from the China Land Data Assimilation System (CLDAS). This was achieved using three densely instrumented in situ observation networks in China that have different environmental conditions, i.e., the Hebi, Naqu and Heihe sites. The European Space Agency Climate Change Initiative multi-satellite-retrieved SM product

ABSTRACT The ground breaking performance of fully convolutional neural nets (FCNs) for semantic segmentation tasks has yet to be achieved for landcover classification, partly because a lack of suitable training data. Here the FCN8 model is trained and evaluated in real-world conditions, so called in-the-wild, for the classification of opium poppy and cereal crops at very high resolution (1 m). Densely

ABSTRACT Satellite remote sensing has been one of the most main methods to acquire the sea surface temperature (SST). In this study, the new regional algorithms to estimate SST in the western north Pacific are developed using the data from the visible and infrared radiometer (VIRR) aboard the Chinese FengYun-3 C (FY-3 C) meteorological satellite. The new regional algorithms are based on the nonlinear

ABSTRACT A synthetic aperture radar (SAR) target classification method is proposed by properly selecting multiple views via the nonlinear correlation information entropy (NCIE). The optimal subset of multi-view SAR images are selected, which are assumed to share stable inner correlations. The joint sparse representation is adopted as the basic classification scheme for the selected views to exploit

ABSTRACT Cloud masking is a critical step in the estimation of Sea Surface Temperature (SST) from satellite observations. It can affect the validation statistics of SST on synoptic scales but also on long-term climatologies. One of the main challenges in cloud masking is the discrimination between clouds and ocean sharp fronts as both of these are associated with high spatial variability. In this study

ABSTRACT In this study, Sentinel-2 optical satellite imagery was acquired over the Peace Athabasca Delta and assessed for its open water classification capabilities using an object-oriented deep learning algorithm . The workflow involved segmenting the satellite data into meaningful image objects, building a Convolutional Neural Network (CNN), training the CNN, and lastly applying the CNN, resulting

ABSTRACT Target discrimination is an important part of the synthetic aperture radar automatic target recognition (SAR ATR). Nowadays, convolutional neural network (CNN) has been used in SAR ATR successfully. However, training CNN requires large amounts of labelled data and collection of the labelled SAR data is expensive and time demanding. It may yield overfitting when directly applying CNN to the

ABSTRACT Landsat satellite time series annual Best-Available-Pixel (BAP) composites for the period 1984–2017 of the Kenora Forest Management Unit in northwestern Ontario, Canada were sampled and stratified by forest stand conditions and aerial sketch map (ASM) compilations of mortality and defoliation. Pre- and post-disturbance multispectral image and textural data were classified using a logistic

ABSTRACT Personal laser scanning (PLS) has shown great potential in diameter of breast height (DBH) estimation. Compared with the current laser scanning technique, PLS is hardly restricted by occlusion and trafficability, and produces omnidirectional point cloud data. However, the DBH estimation accuracy using PLS data was not higher in the previous research. One of the leading sources of bias is overlapping

Accurate estimation of the fraction of absorbed photosynthetically active radiation (fPAR) is important for maize growth and yield estimations. Light detection and ranging (LiDAR)-derived canopy vertical structural and hyperspectral image-derived vegetation spectral information are complementary for vegetation fPAR estimation. This study explores the potential of artificial neural networks (ANNs) with

Semantic segmentation models based on deep learning have shown remarkable performance in road extraction from high-resolution aerial images. However, it is still a difficult task to segment multiscale roads with high completeness and accuracy from complex backgrounds. To deal with this problem, this letter proposes an end to end network named Multiple features integrated with convolutional long-short

Image matching is a quite time-consuming task for Structure-from-Motion (SfM). In this paper, a Bag-of-Words (BoW) model that reduces the feature dimensions and introduces image spatial locations is proposed to improve the efficiency and reliability of SfM. The whole workflow includes three steps. Firstly, principal component analysis (PCA) is used to reduce the high-dimensional features to low-dimensional

Chlorophyll plays an important role in crop photosynthesis, which is closely related to nitrogen (N). N deficiency first occurs in the lower leaves, but the spectral detection of the lower layer is insufficient due to leaf shading. The aim of this paper was to investigate the feasibility of estimating the chlorophyll content of leaves (LCC) and the vertical distribution of LCC in wheat using multi-angle

The mountain hazard triggered by earthquake usually starts near the ridgeline. Knowing the distribution of the ridgeline is conducive to making sound judgements on the possible hazards after an earthquake. However, in the complex scenes of PolSAR images containing buildings, the ridgeline recognition process will produce incorrect judgements due to the existence of buildings. Aiming at the ridgeline

Forest and urban tree inventory measurements are increasingly adopting Remote Sensing (RS) techniques due to the accurate and rapid estimates available compared to conventional methods. The focus of this study is to assess the accuracy and potential application of the Microsoft Azure Kinect – a lightweight depth sensor – for outdoor measurement of tree stem Diameter at Breast Height (DBH). Individual

In the present study, a novel water index has been proposed for mapping of open surface water in multispectral images and named as open surface water index (OSWI). The proposed index has been applied on Landsat-8 data to delineate the water area and the results are compared with those of existing water indices. The index is also applied on Sentinel-2A data and corresponding water pixels are extracted

Hyperspectral image (HSI) classification is one of the core techniques in HSI processing. In order to solve the problem of scarcity of labelled samples, a novel HSI classification framework based on mixture generative adversarial networks (MGAN) is proposed in this letter. Firstly, to overcome the drawback that MGAN cannot be directly applied for classification, a category multi-classifier is introduced

Millimetre wave radar is an emerging technology that can monitor vital signs without contact. This unique feature is very suitable for some particular situations, such as burn patient monitoring. Currently, electrocardiogram (ECG) is still the most common approach for monitoring heart disease. Deep learning algorithms have already been applied to classifying ECG recordings and have achieved good diagnostic

Soil moisture (SM) is a critical variable in energy and water partitioning at the interface between the land surface and atmosphere. In this study, we provided a robust method to retrieve soil moisture using optimal remotely sensed soil evaporative efficiency (SEE) information. Specifically, SEE was deduced from the triangle space constituted by remotely sensed land surface temperature (LST) and fractional

To improve the quality of green tea, low light stress has been used to increase the chlorophyll-a (chl-a) content of tea leaves, although shading treatments sometimes lead to early mortality of tea trees. Therefore, in situ measurement of chl-a and chlorophyll-b (chl-b), which are markers for evaluating light stress and response to changing environmental conditions, can be used to improve tea tree

The Hyperspectral InfraRed Atmospheric Sounder (HIRAS), the first Chinese sun-synchronous hyperspectral infrared sounder onboard FengYun 3D (FY-3D), was launched on 15 November 2017. It will play a significant role in improving forecast accuracy of numerical weather prediction (NWP) systems. However, sun glint effects on the FY-3D HIRAS 3.9 μm channels cannot be currently estimated by radiative transfer

In this letter, we proposed a matting model guided variational pansharpening method with fractional order gradient transferring constraint to fuse a low-resolution (LR) multispectral (MS) and a high-resolution (HR) panchromatic (Pan) image to an HR-MS image. Specifically for the proposed model, we not only used the commonly local spectral fidelity constraint for spectral information preserving, but

The present study evaluate horizontal and vertical accuracy of seven open-source digital elevation models (DEMs) having moderate-to-high resolutions viz. 30 m Shuttle Radar Topography Mission (SRTM1), Advanced Spaceborne Thermal Emission and Reflection Radiometer Global DEM (ASTER GDEM), Advanced Land Observing Satellite World 3D (AW3D30), and Cartosat DEM (CartoDEM), 90 m TerraSAR-X add-on for Digital

Many existing techniques for estimating the optical properties of a body of water directly from a hyperspectral remote sensing spectrum are based on the idea of ‘spectral matching’ – that is, input parameters in a forward model are systematically varied until the measured and modelled spectra are sufficiently similar. This is usually done by using numerical optimization methods with a simplified forward

In this study, aerosol optical depth (AOD) and fine-mode fraction (FMF) with a 3 km resolution are retrieved from Geostationary Ocean Colour Imager (GOCI) data using a multi-temporal method. The retrieved results are input into a physical model for estimating the fine particulate matter (PM2.5) based on instantaneous satellite-based measurements. The other two input parameters of the model are planetary

ABSTRACT This study validated SNAP-derived LAI from Sentinel-2 and its consistency with existing global LAI products. The validation and inter-comparison experiments were performed on two processing levels, i.e., Top-of-Atmosphere and Bottom-of-Atmosphere reflectances and two spatial resolutions, i.e., 10 m, and 20 m. These were chosen to determine their effect on retrieved LAI accuracy and consistency

This paper discusses the estimation of rice yield from Water Cloud Models (WCM) with different layers by a single X-band (8–12 GHz) Synthetic Aperture Radar (SAR) image, from a satellite named TerraSAR. A simulated annealing (SA) algorithm, with a ‘leave-one-out’ cross-validation method, was applied. The results were related to actual rice parameters, with a coefficient of determination (R 2) of 0

As an important part of the ecosystem, wetlands provide varies of ecological functions while they have also been increasingly threatened and degraded. Therefore, it is necessary to protect wetlands with effective monitoring measures such as classification. Considering the quick development of multispectral sensors, multispectral images are available to classify wetlands. The current method, however

Tree detection and counting have been performed using conventional methods or high costly remote sensing data. In the past few years, deep learning techniques have gained significant progress in the remote sensing area. Namely, convolutional neural networks (CNNs) have been recognized as one of the most successful and widely used deep learning approaches and they have been used for object detection

In this paper, a novel modification applies to the Band-Dependent Spatial Detail (BDSD) framework to improve the fusion performance interestingly. Joint multiplication terms have been added to the conventional BDSD framework to make it nonlinear BDSD. It has been shown that joint product terms would lead to a better estimation of detail maps for each spectral band. To support the idea presented here

The large number of bands, the huge amount of information and the high band correlation in hyperspectral images bring great difficulties to the band selection of hyperspectral images. In the basic Particle Swarm Optimization (PSO), the learning factor and inertia factor are fixed, which limits the exploratory ability of the algorithm and cannot balance the global and local relations well. Therefore

Monitoring the dynamics of snow and glaciers in the Indian Himalaya has always attracted the attention of the remote sensing community. The snow water equivalent (SWE) represents the amount of water contained in the snowpack, and it is a product of snow depth (SD) and snow density ( ρ s ). The estimation of SD at high spatial and temporal resolution is still a challenge, especially in rugged mountainous

Convolutional Neural Networks (CNNs) are widely used in various fields, and have shown good performance in hyperspectral image (HSI) classification. Recently, utilizing deep networks to learn spatial-spectral features has become of great interest. However, excessively increasing the depth of network may result in overfitting. Moreover, in HSI classification, the existing network models ignore the strong

Soil-available nutrients (SANs)are essential for crop growth and yield formation. Appropriate variable rate fertilization (VRF) can control SAN at a normal level to avoid unnecessary damage to sustainable production capacity. The precondition of optimizing the application of VRF is obtaining the real-time status of SAN. A new method for SAN estimation has been proposed by integrating modified World

Multichannel synthetic aperture radar (SAR) is an effective method to achieve high-resolution wide-swath (HRWS) imaging simultaneously. Many algorithms have been proposed to reconstruct stationary targets. However, moving targets cannot be well reconstructed through traditional reconstruct algorithm for static scenes and the ghost targets of the moving targets will have a great influence on SAR image

The convolutional neural network (CNN) is widely used for image classification because of its powerful feature extraction capability. The key challenge of CNN in remote sensing (RS) scene classification is that the size of data set is small and images in each category vary greatly in position and angle, while the spatial information will be lost in the pooling layers of CNN. Consequently, how to extract

Accurate soil moisture (SM) data are key for climate and surface-atmosphere simulations associated with prediction and analysis of weather and climate variability. Here, we assessed in situ vs. remotely sensed SM discrepancies in a humid watershed during an extreme drought and an arid watershed under peak dry-season conditions. Using in situ SM measurements from the Soil and Climate Analysis Network

This research compares the capabilities of various Sentinel-2-derived spectral vegetation indices (SVIs) in particular red-edge SVIs to detect and classify spruce budworm (Choristoneura fumiferana) (SBW) defoliation using Support Vector Machine (SVM) and Random Forest (RF) models. The results showed the superiority of RF in model building for defoliation detection and classification into three classes

Aircraft detection in remote sensing imagery has drawn much attention in recent years, which plays an important role in various military and civil applications. While many advanced works have been developed with powerful learning algorithms in natural images, there still lacks an effective one to detect aircraft precisely in remote sensing images, especially in some complicated conditions. In this

This paper presents a multispectral image fusion method for remote sensing applications by using an improved estimation of detail maps. This estimation is achieved by considering a linear combination of the other MultiSpectral (MS) bands as well as their joint multiplications. A comparison with seven existing methods is carried out based on three public domain datasets and by considering commonly used

This letter develops a synthetic aperture radar (SAR) target classification method based on bidimensional variational mode decomposition (BVMD) and multitask compressive sensing (MTCS). BVMD is employed to decompose SAR images to exploit the time-frequency properties of the described targets. The MTCS is used to jointly classify the original SAR image and its BVMD components. So, the merits of BVMD

(2020). Remote Sensing Letters contribution to the success of the Sustainable Development Goals - UN 2030 agenda. Remote Sensing Letters: Vol. 11, No. 8, pp. 715-719.

In this paper, a novel approach is proposed for Synthetic Aperture Radar (SAR) target classification based on multi-aspect multi-feature collaborative representation. Firstly, principal component analysis (PCA), wavelet and 2-dimensional slice Zernike moments (2DSZM) features are extracted from SAR images. Next, based on the strong correlation among the adjacent aspect SAR target images, we extend

An adaptive-trimming-depth-based constant false alarm rate (ATD-CFAR) ship detector of heterogeneous environment in SAR imagery is proposed in this letter. Traditional CFAR detectors generally use all samples in the background window for parameter estimation. However, in the heterogeneous regions, these detectors will overestimate the parameters used for statistical modelling due to the interference

In this letter, we proposed a novel deep feature manifold embedding method to improve feature extraction ability of traditional deep learning methods. This method first obtains deep features of hyperspectral image (HSI) from a trained autoencoder. Then, an intrinsic graph and a penalty graph are constructed to discover the discriminant manifold structure of deep features. Finally, the deep features

The instrument used for ocean colour remote-sensing works in the visible wavelengths, and the presence of clouds frequently lead to invalid observations. The formation of clouds is known to be influenced by mesoscale oceanic processes (e.g., eddies and temperature fronts), but these influences are often overlooked in missing data reconstructions. By analysing more than 10 years of chlorophyll-a (chl-a)