This is a list of reviewers who served the Journal of Applied Remote Sensing in 2019.

Knowledge gained about mangrove species mapping is essential to understand mangrove species’ development and to better estimate their ecological service value. Spectral bands and spatial resolution of remote sensing data are two important factors for accurate discrimination of mangrove species. Mangrove species classification in Shenzhen Bay, China, was performed using Sentinel-2 (S2) multispectral instrument (MSI) data and Google Earth (GE) high-resolution imagery as data sources, and their suitability in mapping mangrove forest at a species level was examined. In the classification feature groups, the spectral bands were from the S2 MSI data and the textural features were based on GE imagery. The support vector machine classifier was used in mangrove species classification processing with eight groups of features, which were based on different S2 spectral bands and different GE spatial resolution textural features. The highest overall accuracy of our mapping results was 78.57% and the kappa coefficient was 0.74, which indicated great potential for using the combination of S2 MSI and GE imagery for distinguishing and mapping mangrove species.

The accuracy of differential interferometric synthetic aperture radar (DInSAR) in monitoring the ground subsidence is a major challenge to be addressed urgently. Using the repeat track DInSAR and GIS spatial analysis tools, eight C-band Sentinel-1A SAR images of the Guotun coal mine (China) were processed to determine the mining subsidence from November 27, 2015 to July 24, 2016. The mining data of 13 working faces and the DInSAR- and leveling-monitored results were compared. A method was proposed to solve the problem of time inconsistency between DInSAR- and leveling-monitored results. The location, spatial distribution, scope, and variations of mining subsidence monitored by Sentinel-1A repeat track DInSAR were consistent with the mining progress of the working faces. The accuracy of the DInSAR-monitored subsidence values was directly related to the coherence of the subsidence zones, and the absolute difference from the leveling-monitored values was small at the subsidence edge but large at the subsidence center.

There exist thousands of water bodies in watersheds, including large-scale water bodies, such as reservoirs, and small-scale water bodies, such as lakes, ponds, etc. In basin flood forecasting and other hydrology-related tasks, water bodies play an important role in the flooding process. The method of efficiently segmenting water bodies from remote sensing images (RSIs) is still a popular research topic in the fields of computer science and remote sensing. We propose a model based on mask R-CNN to automatically detect and segment water bodies in RSIs, thereby avoiding the complex operations of manual feature extraction when processing aerial images or satellite images because these images often have low resolution and complex background. RSIs were obtained from various remote-sensing research datasets and from snapshots from Google Earth. Data augmentation was introduced to enrich the training images dataset. Then, the proposed model was trained on the augmented dataset in two implementations: residual network (ResNet)-50 and ResNet-101. Experimental results show that the proposed method scores 90% on average for regular-shaped water bodies and 76% on average for irregular-shaped water bodies in terms of intersection over union, which indicates that the proposed models offer excellent feasibility and robustness for water-body segmentation.

Multiple vehicle tracking in aerial video has vital applications in intelligent transportation system. However, this is challenging owing to the demand for higher tracking efficiency. Moreover, it is particularly challenging to handle shadows and occlusions in complex traffic scenes. Currently, most approaches mainly address the multiple vehicle tracking problem based on consecutive frame pairs. However, they are inefficient and ineffective with respect to the tracking of vehicles that are shadowed or occluded by trees, buildings, or large vehicles. Therefore, in terms of consecutive multiframe images, we propose an efficient deshadowing approach for multiple vehicle tracking in aerial video via image segmentation and local region matching. By processing a series of frames at a time, this approach will address the occurrence of shadows and occlusions in multiple vehicle tracking. First, image segmentation and road segmentation are combined to construct the region adjacency graph within the road (RAGR) for each frame. Local region matching is then performed among the RAGRs based on the distance transform. In particular, to improve the region matching efficiency, a local matching criterion is proposed by introducing region of interest for the node of the RAGR. Finally, based on the motion dissimilarities and varying spatial distances of the background and moving vehicles, a batch-oriented foreground separation is achieved to separate vehicles (the foreground) from the background. Experiments and comparative analysis conducted on UAV123 and DARPA VIVID data sets demonstrate that the proposed approach for multiple vehicle tracking in aerial video can generate satisfactory and competitive results.

We report the lidar observations of a prolonged unusual event of atmospheric aerosol load extending up to about 15 km above the city of Sofia, Bulgaria. Based on the results obtained and air-transport modeling data, we show that the observed diffused aerosol layers consisted predominantly of Saharan mineral dust partly mixed with other aerosol types. We present retrieved time-averaged vertical profiles of the aerosol backscatter coefficient (at 532 and 1064 nm) in order to display the aerosol density height distribution, whereas the aerosol layering temporal dynamics is illustrated by color-coded maps of range-corrected lidar signals. We also calculated the profiles of the backscatter-related Ångström exponent (BAE), which showed the domination of moderately coarse aerosol particle modes. Further, frequency count distributions of the BAE occurrences were obtained and analyzed statistically. Particular attention was paid to the revealed BAE distribution skewness, in view of better identifying and characterizing the aerosol size and composition changes. The results obtained enabled us to argue that the atmosphere above Sofia in the period considered had been strongly affected by intrusion of Saharan dust. Desert aerosols had perturbed or covered over the top of the atmospheric boundary layer and filled up the free troposphere, considerably influencing the local meteorological conditions. We also ascertained that aged volcanic aerosols had been likely present and detected at tropopause altitudes.

The ability to accurately extract urban digital surface models (DSMs) at a reasonable cost is quite valuable. In this study, the ZiYuan-3 (ZY-3) satellite across-track and along-track modes were used to build stereo pairs and elevation models featuring different views. Image matching processing was used to generate optical point clouds for multiple stereo pairs, which were used to analyze the effects of point cloud fusion under different scenarios and optimize the ZY-3 multiview point cloud integration. Then, with the assistance of a high-precision airborne light detection and ranging point cloud, 12 indices, including the maximum, minimum, mean, and B10 to B90, were selected from the building roof optical point cloud in order to construct an urban building elevation model. The results were as follows: (1) The across-track optical point cloud formed by stereo pairs of ZY-3 nadir images is more suitable for urban DSM extraction due to the high spatial resolution and small base/height ratio of these images. (2) The B70 index performs best in the estimated urban building elevation model [R²  =  0.91, root mean square error (RMSE) = 5.59 m]; this model accurately identified 95.39% and 95.25% of the whole-building elevations in a narrow-scale test of 3170 buildings and a broad-scale test of 1,510,606 buildings, respectively. Moreover, the model R² and number of accurate building elevation recognitions gradually increased with increasing numbers of fused point clouds, while the RMSE decreased significantly. (3) A combined ZY-3 multiview point cloud fusion and multispectral data analysis technique delivered a more refined urban DSM and better characterization of building elevation under complex underlying surface scenarios than were obtained using traditional stereo pair technology. This research presents the methods and theory relating to the application of ∼2- to 6-m resolution satellite images for complex urban underlying surface scenarios.

The rice-cropping system in southern China experienced a major change in the past few decades. Rice-cropping systems not only affect the comprehensive utilization intensity of agricultural resources, food security, and the ecological environment but also reveal the conditions of the agricultural policy, regional economy, and rural labor force. Consequently, accurate detection of rice-cropping systems is an essential issue. We present a method for classifying rice-cropping systems using a pretrained convolutional neural network (CNN) that involves extracting deep features of spatial and spectral trajectories. The study area is a major rice-growing region located in Zhuzhou City, Hunan Province, China. Multitemporal Sentinel-2 satellite images were collected to obtain time series curves of texture (spatial trajectories) and the first derivative vegetation index (spectral trajectories) as inputs of the hierarchical classification model. The classification results of land cover and rice-cropping systems were acquired using the pretrained CNN, and the method achieved overall accuracy of 94.78% and 94.87%, respectively. This method was then compared with the support vector machine (SVM). The accuracy of rice-cropping systems using the pretrained CNN was 7.11% higher than that of the SVM. The adaptability of the model was also investigated using the time series curve in another year with relatively insufficient data. The model obtained satisfactory performance with the overall accuracy of land cover types and rice-cropping systems of 93.52% and 93.23%, respectively. We suggest that the pretrained CNN can improve the accuracy of rice-cropping system mapping by extracting deep features of spatial and spectral trajectories during the rice growth cycle and such method may be applied in other fine crop classifications.

Current very high-resolution (VHR) satellite images enable the detailed monitoring of the Earth and can capture the ongoing works of railway construction. We presented an integrated framework applied to monitoring railway construction in China using QuickBird, Gaofen-2, and Google Earth VHR satellite images. We also constructed a deep convolutional neural network-based semantic segmentation network to label the temporary works, such as borrow and spoil areas, camp, beam yard, and environmentally sensitive areas such as resident houses throughout the whole railway construction project using VHR satellite images. In addition, we employed the holistically nested edge detection subnetwork to refine the boundary details and a cross entropy loss function to fit the sample class disequilibrium problem. Our semantic segmentation network was trained on 471, validated on 58, and tested on the 58 VHR true color images. The experiment results showed that compared with the existing state-of-the-art DeepLabV3 plus approach, our approach has improvements with an overall accuracy of more than 80%.

Mangroves not only protect the coastline from ocean waves and storms but also provide economical, ecological, and social functions for the human being. Despite the important role mangrove plays, mangrove has suffered great loss over the past decades. Mapping the mangrove extent is essential for its conservation. We examined the potential of original bands, spectral indices, including combined mangrove recognition index (CMRI) and normalized difference vegetation index (NDVI), and principal component analysis (PCA) of Sentinel-2 to map the mangrove of Dongzhaigang, China in June 2019. Random forest (RF) approach was adopted for the image classification. Results showed that the overall accuracy and kappa coefficient of the image classification can achieve 90.47% and 0.86. The area of mangrove of Dongzhaigang, China in June 2019 is 17.83  km². The importance of each variable was calculated and among all the 15 variables, the CMRI ranked first, followed by NDVI, band 3 (green), band 2 (blue), and PCA1 (the first principal component) in turn. Based on the results, we can conclude that vegetation indices and PCA are essential for mangrove extent mapping since the variable importance of these features are relatively higher in the ranking of mean decrease Gini. Also band 3 (green band) provides important information for distinguishing mangrove from land and water class. Furthermore, combining vegetation indices, PCA, and RF model can accurately delineate the extent of mangrove.

Owing to the inaccessibility of most of the mineral deposits as well as cost and time requirements for conventional ground-based methods, remote sensing techniques become very useful for the rapid detection, mapping, and multitemporal tracking of different mineralization surface states. In this context, a methodological approach based on hyperspectral spectroscopy and Advanced Spaceborne Thermal Emission and Reflection (ASTER) data is proposed for the phosphate exploration and mapping in a specific geological context in Tunisia. ASTER’s thermal infrared (TIR) portion has been particularly expedited in phosphate detection and mapping. Visible near-infrared (VNIR), shortwave infrared (SWIR), and TIR atmospherically and geometrically corrected data are thus processed using the matched filtering (MF) method and VNIR-SWIR reflectance spectra of the collected samples. Results revealed that different absorption features of carbonate minerals (calcite and dolomite) in the SWIR region have to be considered for an efficient identification of rock phosphate. Proposed band combinations using MF fraction maps have successfully discriminated phosphate mineralization and outcrops. Moreover, the surface temperature map derived from TIR bands using the emissivity normalization method reveals the usefulness of the proposed method for rock phosphate delineation. Both geological map and published prospects map as well as x-ray diffraction (XRD) analysis results are used for the validation purposes. A high kappa coefficient value of 0.92, 0.97, and 0.88, is thus calculated, respectively, for VNIR, SWIR, and TIR data-based maps. The high kappa coefficient values implicate the contribution of proposed methodological approach in exploration and mapping phosphates in the Chaketma mine site in central west Tunisia.

To measure the image distortion during the actual operation of an airborne camera, improve the accuracy of geolocation using airborne cameras, and reduce the influence of image distortion on location results, a distortion measurement algorithm based on overlapping remote sensing images is proposed, which attempts to overcome the image distortion caused by external working environment. A distortion measurement algorithm based on the geographical locations of the image coincident points is presented, and the distortion rate for the camera is calculated. To correct the positioning error caused by image distortion, the coordinate values of the target projection in the charge-coupled device coordinate frame are processed based on the distortion rate. The simulation results show that positioning accuracy decreases with increase in the distortion rate. When the distortion rate is 4%, use of the correction algorithm can improve positioning accuracy by 21%. The flight test shows that the maximum distortion rate measured by the algorithm for the experimental camera is approximately 9%. After correction, positioning error is reduced by 50 to 100 m, and positioning accuracy is greatly improved to meet the needs of actual engineering.

Advancement in wireless communication as well as recording and transferring data over the internet provides a lot of possibilities for smart inspection and monitoring for machines and structures. The big data recorded and transferred through such a system must be analyzed efficiently on the go to provide accurate feedback to the system. Neural network (NN) data processing techniques are an effective methodology for fast and accurate analyses of the data and provide feedback to the system. An NN methodology is proposed for structural health monitoring of bridge structures. The proposed platform uses the direct and indirect sensors mounted on the bridge structure and on the passing vehicle, respectively. This proposed approach will decrease the cost and the potential damages to the sensors in direct methods, and will increase the accuracy and reliability of monitoring in indirect techniques. The methodology and data processing techniques have been validated using a lab-scaled test bed.

Invasive alien species are a major threat to global biodiversity and result in adverse environmental and socioeconomic implications, such as reduced ecosystem services, landscape productivity, and costly eradication initiatives. The ability to monitor the extent and spread of alien species invasions provides valuable insight for the mitigation of these adverse implications. The generation of Earth Observation (EO) Sentinel sensor provides unprecedented freely available imagery, which is suitable for both local and regional invasive species monitoring. Specifically, its radar (S1) and optical (S2) sensors offer unique tandem datasets valuable for landscape analysis. Hence, we sought to determine the synergistic potential of fused S1 dual-polarized synthetic aperture radar (SAR) imagery with S2 optical imagery for invasive alien species detection and mapping. S1 and S2 imageries were fused at the feature level, and the support vector machine algorithm used for the multiclass image classification. Results indicated that the fusion of the S1 dual-polarized imagery with S2 optical imagery produced the highest classification accuracy (85%), whereas stand-alone S2 optical bands produced the lowest classification accuracy (79%). Findings from this study underline the significant synergistic potential and complementarity of new-age S2 optical imagery and dual-polarized S1 SAR imagery for invasive alien species detection and mapping. Due to large swath, higher pixel resolution, free availability, and possible tandem complementarity between optical and SAR sensors, we recommend Sentinel EO imagery as an economically viable option for the invasive alien species detection and mapping.

Land surface temperature (LST) is a crucial parameter for global climate change studies. LST changes are also directly associated with the large-scale changes in land cover. Previous studies carried out a comparative analysis of satellite-derived LST response between periods before and after homogenous land cover changes. We present an alternative approach that quantifies long-term LST variability in response to various land use/land cover change (LULCC) patterns over Phuket Island, Thailand, from 2003 to 2017. First, four Moderate Resolution Imaging Spectroradiometer (MODIS) overpass times of LST time series were adjusted for seasonal effects using a cubic spline function to preserve the number of original data and enable estimates of LST dynamics and trends using the generalized least squared models. Second, LULCC patterns were classified according to land cover type conversion and spatial pattern transformations between the years 2000 and 2016. Spatial homogeneity and heterogeneity were quantified by the coverage percentage for each land use and land cover (LULC) type within a given location. Finally, the influence of LULCC patterns on the long-term spatiotemporal behavior of LST was assessed using the generalized estimating equation model. Results showed that different land cover transitions influence the dynamics of daytime LST but not the nighttime LST. The proportion of different land cover types within an LST pixel and transition amounts contributed to the quantity of increasing surface temperature, especially over impervious surface types. Diverse LULCC patterns with considerations of spatial heterogeneity improved our insight about a relatively strong effect of combined LULC types on LST responses. The climatic effect through the gradual conversion of heterogeneous land cover is necessary to be considered in climate research studies.

Lodging stress affects the yield, quality, and mechanical harvesting capacity of maize, thereby resulting in a reduction in maize production. The timely monitoring of maize lodging assists in determining the extent and degree of the effects. This study proposed a method for this purpose using Gaofen-1 (GF-1) optical satellite images, which can be significant for lodging field management, production assessments, and implementation of the related remedial measures. The evaluation index of the lodging grade was constructed by measuring the proportion and angle of maize lodging after the lodging incident. Variations in spectral reflectance and vegetation index were calculated before and after lodging. The competitive adaptive reweighted sampling algorithm was used to screen the optimal combination of variables sensitive to the evaluation index of the maize lodging grade. A remote sensing monitoring model was established by using random forest (RF) and partial least squares (PLS). Results show that the maize lodging monitoring model established by RF is better than that of PLS. The accuracy of monitoring the lodging damage using the GF-1 images reaches 79%. Meanwhile, the classification map of the lodging grade we constructed using an RF model is consistent with the actual lodging area. Hence, this method can effectively characterize the intensity of lodging stress and can also be used to assess the range of damages caused by regional-scale lodging.

We introduce a new framework for compressed sensing (CS) in synthetic aperture radar (SAR) imaging in the case of model error. Conventional CS-based autofocus methods solve a joint optimization problem to achieve both model error parameter estimation and SAR image formation simultaneously. Owing to the possibly nonconvex feature of the joint optimization problem, however, these algorithms may get stuck in local optima having large phase errors and thus fail to reconstruct the image. In contrast, we use phaseless measurements and pose imaging as a convex optimization problem. To solve the convex problem, we use the alternating direction method of multipliers-based approach, which is computationally efficient and easy to implement. The results from simulations with both point targets and extended targets validate the effectiveness of the proposed method.

Based on the optimal estimation (OE) theory, an inversion framework for aerosol retrieval from multiviewing polarimetric satellite measurements, is presented. The retrieved parameters include the wavelength-independent volume concentration and spectral aerosol optical depth (AOD). An OE-iteration approach is proposed to implement the framework. Polarization data from the polarization and directionality of the Earth’s reflectance sensor over the Chinese Beijing–Tianjin–Hebei region are selected to test the proposed algorithm. The validation against aerosol robotic network products produces high correlation coefficients (R) of 0.85, 0.9, and 0.86 for the fine-mode volume concentration, fine-mode AOD, and total AOD at 500 nm, respectively. These results indicate that the OE retrieval framework is practical and useful in the quantitative retrieval of AOD and fine-mode volume concentration from the multiangle polarization data.

Accurate simulation of potato leaf area index (LAI) under water stress is of great significance for rational selection of planting location and extraction of planting area. To obtain the simulation model of potato LAI under different water conditions [water saturation (WS), water deficiency (WD), and control check (CK)], several spectral parameters of vegetation indices (VIs), first-order differential parameters and spectral parameters of the continuum removal method were derived from canopy spectra, first-order differential spectra and continuum removal spectra for correlation with the corresponding LAI, respectively. Furthermore, three different types of parameters with high LAI correlation coefficients were used to build the simulation models and verify the accuracy. As a result, the selected modeling parameters were normalized difference VI, soil-adjusted VI, quotient of red edge area (SDr) and blue edge area (SDb),   (  SDr  −  SDb  )    /    (  SDr  +  SDb  )  , depth area ratio (W), and gross area of absorption peak (S). The most suitable models are as follows: the exponential model of W under the WS condition [coefficient of determination   (  R²  )    =  0.8262, root mean square error   (  RMSE  )    =  0.4338, and mean absolute error   (  MAE  )    =  0.3750], the power function model of S under the CK condition (R²  =  0.8133, RMSE  =  0.4695, and MAE  =  0.3798), and the power function model of SDr  /  SDb under the WD condition (R²  =  0.8407, RMSE  =  0.4459, and MAE  =  0.3882).

Rice is one of the world’s most dominant staple foods, and hence rice farming plays a vital role in a nation’s economy and food security. To examine the applicability of synthetic aperture radar (SAR) data for large areas, we propose an approach to determine rice age, date of planting (dop), and date of harvest (doh) using a time series of Sentinel-1 C-band in the entire Mekong Delta, Vietnam. The effect of the incidence angle of Sentinel-1 data on the backscatter pattern of paddy fields was reduced using the incidence angle normalization approach with an empirical model developed in this study. The time series was processed further to reduce noise with fast Fourier transform and smoothing filter. To evaluate and improve the accuracy of SAR data processing results, the classification outcomes were verified with field survey data through statistical metrics. The findings indicate that the Sentinel-1 images are particularly appropriate for rice age monitoring with R²  =  0.92 and root-mean-square error (RMSE) = 7.3 days (n  =  241) in comparison to in situ data. The proposed algorithm for estimating dop and doh also shows promising results with R²  =  0.92 and RMSE  =  6.2 days (n  =  153) and R²  =  0.70 and RMSE  =  5.7 days (n  =  88), respectively. The results have indicated the ability of using Sentinel-1 data to extract growth parameters involving rice age, planting and harvest dates. Information about rice age corresponding to the growth stages of rice fields is important for agricultural management and support the procurement and management of agricultural markets, limiting the negative effects on food security. The results showed that multitemporal Sentinel-1 data can be used to monitor the status of rice growth. Such monitoring system can assist many countries, especially in Asia, for managing agricultural land to ensure productivity.

We comprehensively considered the spatial resolution and seasonal differences in NO₂ column concentration data observed by different satellites to reconstruct an NO₂ column concentration dataset in China from 1997 to 2018. On this basis, the spatial and temporal variations in NO₂ column concentrations were analyzed over China from 1997 to 2018. The results show that the reconstructed NO₂ column concentration dataset with high spatial structure characteristics in China from 1997 to 2018 effectively compensates for the lack of long-term high-resolution monitoring of NO₂ pollution in China by a single satellite. (The accuracy of the reconstructed dataset was verified by season: correlation coefficient between GOME and SCIAMACHY, R  >  0.87, P  <  0.01; correlation coefficient between GOME2A and SCIAMACHY, R  >  0.91, P  <  0.01.) The analysis of the spatial and temporal variations showed that the NO₂ column concentrations in China from 1997 to 2018 generally showed high values in the east and low values in the west. The high concentrations of NO₂ are mainly concentrated in the northern and eastern coastal areas and the Pearl River Delta. From 1997 to 2010, the NO₂ column concentrations continued to increase with a growth rate of   (  0.180  ±  0.011  )    ×  10¹⁵  molec  /    (  cm² a  )   over China. From 2011 to 2018, the NO₂ column concentrations showed a downward trend, with a decreasing rate of   (  0.144  ±  0.037  )    ×  10¹⁵  molec  /    (  cm² a  )   over China. After 2011, NO₂ pollution in high pollution regions such as Beijing, Tianjin, Hebei, Shandong, the Yangtze River Delta, and the Pearl River Delta was alleviated. Energy saving and emission reduction measures can reduce nitrogen dioxide pollution while ensuring sustained gross domestic product (GDP) growth.

Spatiotemporal inconsistency in MODIS land surface temperature (LST), one of the most widely used geospatial data parameters, is a concern for its application in various studies. Moreover, there are limited methods that can address spatiotemporal reconstruction of LST in diverse physiography. The use of kernel-based spatiotemporal assimilation in a multitemporal approach to reconstruct LST in a complex physiographic region, northeast India, is addressed. Global land skin temperature (Ts) from Global Land Data Assimilation System (GLDAS) is chosen as a temporal covariate for MODIS LST due to its consistent temporal resolution. Considering the high temporal correlation between GLDAS Ts and MODIS LST, temporal assimilation is done in the first stage followed by spatial assimilation in the second stage. Due to data gaps, a kernel-based nonparametric estimator is adopted to map the spatiotemporal distribution and reconstruct LST with spatiotemporal consistency. This approach shows satisfactory performance in restoration of spatial variation in the study area with a coefficient of determination (R²  =  0.98), root mean square error (RMSE = 0.61 K), and absolute value of bias (B  =  0.22  K). Also, the credibility of reconstructed LST from this method is found to be as good as the baseline data (MODIS LST) given that there is no severe data deficiency with uneven spatial availability in the baseline data itself. Comparison of gap-filled LST showed positive correlation with ground-based measurements (average R  =  0.71), reflecting decent agreement with seasonality. Hence, the kernel-based nonparametric assimilation could be used to reconstruct LST using a multitemporal approach in complex physiographic regions, which could be used to fill the spatiotemporal data gaps and increase data consistency in MODIS LST.

In recent years, there has been an increasing interest in object detection in remote sensing imagery with onboard sensors and embedded platform based on deep convolutional neural networks. However, the limited cost, power consumption, compute complexity, and parameter size make the task challenging. The current object detection frameworks are mainly designed on the basis of graphics processing units (GPUs) and require further optimization in power consumption, and calculating quantity and parameter size. To address these issues, we propose an effective single-shot multibox detector (DF-SSD) framework, using the DepthFire module we designed to reform SqueezeNet as the backbone network to reduce the calculating quantity and improve the processing efficiency. To evaluate the effectiveness and superiority of DF-SSD, compared with the other state-of-the-art methods, extensive experiments are conducted on various hardware platforms, including GPU 1080ti, central processing unit i7-7700k, NVidia Jetson TX2, and Cambricon-1H8. Experiment results show that the designed algorithm can achieve a mean average precision of 75.2% on NWPU VHR-10 dataset, with 181, 5.2, 26.3, and 15 frames per second on the above four typical hardware platforms, respectively, which finally demonstrate the effectiveness and high accuracy of the proposed algorithm.

The objective of this study is to present a geographic information system based method for detecting and quantifying boulders (area, length, width, elevation, and transport distance) that couples image processing techniques with spatial analysis tools. This study focuses on a delta within Jezero Crater on Mars, the site of the upcoming Mars 2020 rover mission. Understanding the distribution of boulders and their properties is significant since they are critical for understanding surficial geology, sedimentology, and erosional/depositional processes on Mars. Additionally, boulders and large clasts pose a challenge to the traversability of terrain for both rovers and future human missions. A high-resolution imaging science experiment image of the Jezero delta was processed using a filtering/masking algorithm, which combines mean filtering, brightness interval masking, and range kernel filtering. User-determined conservative (no false-positive boulder pixels) and liberal (no false-negative pixels) brightness thresholds were used to generate boulder polygons and their minimum-bounding geometry. Using conservative parameters, 443,126 boulder-like features were identified, of which 153,582 (35%) were relevant boulders. For these boulders, the cumulative fractional area versus boulder width relationship flattens out as predicted in prior work, suggesting that approach is accurate, and further is simpler than existing approaches, and can identify boulders as small as 50 cm in diameter. The algorithm, which can be customized to various degrees of boulder rounding and works on images with different brightness and contrast levels and illumination angles, is applicable to boulder fields on Earth as well as Mars and other solar system bodies.

Change detection of hyperspectral (HS) images can be a motivating research area because of the dense spectral sampling of the HS dataset. The support vector domain description (SVDD) classifier has been widely used for change detection of multispectral images. Most of the research has focused only on the binary change detection problem. We took advantage of the high spectral resolution of HS images to extract different kinds of changes present in the image. As a result, we proposed a multiple-change detection algorithm for the HS dataset based on the sparse representation binary change mask and SVDD classifiers. Also, a geometric index, D2R2, was presented to integrate the binary SVDD classifiers. The experimental results showed that our proposed algorithm has comparable performance to some deep neural networks, including one-dimensional convolutional neural network (CNN), two-dimensional CNN, and Deep Pattern Net, with overall accuracies of 88.51% and 70.11% on the two Hyperion sensor datasets.

Rapid, precise, and quantitative assessment of soil total nitrogen (TN) of alpine grassland is crucial for monitoring the environmental conditions of grasslands. The right choice of preprocessing and calibration methods may improve the model performance of soil property estimation in visible/near infrared reflectance spectroscopy (VNIRS). VNIRS (350 to 2500 nm) measurements were made of 80 soil samples from the northeast part of the Tibetan Plateau. To reduce the signal noise, the Savitzky–Golay technique was implemented and six spectral transformations were evaluated with the aim of identifying the best transformation for predicting soil TN. The performance of three linear calibration models, namely, stepwise multiple linear regression, principal component regression, partial least squares regression (PLSR), and a combination of these models with a nonlinear back propagation neural network (BPNN) was tested for accuracy in the measurement of soil TN. Our results indicate the following: (1) models using the derivative transformations of spectra as inputs perform better than those using nonderivative transformations, and the first derivative of the reciprocal of reflectance [(1  /  R)^′] is the best spectral transformation for soil TN estimation; (2) models based on the PLSR and companion BPNN-latent variables (LVs) yield better predictions and smallest errors of soil TN. We demonstrate that (1  /  R)^′ is the best spectral transformation and consistently improves the estimation models of soil TN. Among multivariate techniques, the BPNN-LVs is recommended for estimation of soil TN with great accuracy.

Deep learning methods have been developed and widely used in land use classification with remote sensing images. In addition, due to the different datasets used in different studies, there is a lack of direct comparison between different deep learning model applications in land use classification. The open source dataset DeepSat was used to build and test a convolutional neural network (CNN) model. The convolution kernels in the model were extracted to further study the specific features learned by the deep learning model. In addition, different CNN-based models were compared to explore the impacts of model structures on model accuracy. The major conclusions from the research are: (1) CNN model is effective in land use classification, with an accuracy of 0.9998 and 0.9991 for the SAT-4 and SAT-6 data, respectively; (2) CNN does have a “learning” ability that can extract the most critical and effective information from training datasets; and (3) for remote sensing land use classification, increasing the number of convolution kernels is better than adding more convolutional layers. Max pooling is better than average pooling. In addition, a localized response normalized layer can also improve model accuracy.

Large intraclass variance and low interclass variance are among the most challenging problems in very high-resolution (VHR) image classification. Semantic segmentation constructed in a deep convolution neural network is used as a classification algorithm conducted via end-to-end training, which combines spectral–spatial features and context information. However, large-scale remote sensing images cannot be directly processed because they are limited by GPU memory and segmentation algorithm. At the same time, classification using single band combinations is also unsatisfactory due to the extraordinary complex features of VHR images. Therefore, a method is proposed based on multiple band combinations and patchwise scene analysis. A complex remote sensing image can be considered as into a combination of simple scenes from multiple patchwise images. And optimal band combinations of each patchwise image are selected according to their scene. The segmentation results of each patchwise image are merged to get the desired results according to geographical coordinates. Our method is validated on the ISPRS 2-D Semantic Labeling dataset of Potsdam, on which results competitive with the state-of-the-art are obtained. The proposed scheme has strong universality and can be used for large-scale high-resolution remote sensing image classification.

Shadow detection plays an important role in remote sensing applications. Shadow should be detected with damage assessment algorithms, and it should be removed from the ground surface with semantic labeling applications. The procedure of a typical shadow detection method includes defining a shadow index and thresholding it, either automatic or manually. An automatic shadow detection method is proposed to facilitate the process of automatic applications. Specifications of shadow and nonshadow areas are analyzed to construct a new spectral–spatial shadow index. Spectral elements of the index can handle the dark shadow extraction. The spatial element of the index provides a high separability between light shadow and dark nonshadow areas such as water bodies. Index definition follows a segmentation algorithm to provide a segment-based analysis. Thresholding is a nonseparable part of shadow detection methods as it divides the region into shadow and nonshadow areas. To avoid high false positive or negative results, the proposed thresholding method is dependent on the enhanced bimodality test. Bimodal distribution can easily be thresholded by typical techniques such as the Otsu thresholding, whereas a special clustering-based thresholding is proposed in the unimodal distribution. The evaluations show a great improvement in shadow detection of very high-resolution RGB images.

The Baltic Sea represents an optically complex case 2 water type, where high concentrations of water column constituents limit acquisition of benthic information. Different preprocessing steps were applied to the hyperspectral compact airborne spectrographic imager (CASI) image to extract as much useful benthic information as possible. Atmospheric correction, minimum noise fraction transform, and sun-glint correction were performed to acquire water surface reflectance data. Additionally, a water column removal procedure was applied to acquire bottom reflectance data. Although retrieved CASI water surface reflectance spectra generally matched the magnitudes and shapes of in situ measured spectra, then the applied water column correction algorithm did not yield accurate bottom reflectance spectra. Therefore, both benthic habitat and bathymetry maps were retrieved from the CASI sea surface image data set. An image-based supervised classification method produced a good quality benthic habitat map from the shallow Pakri study area (depth  <  3.0  m) with an overall accuracy of 80%. A site-specific algorithm was developed for the bathymetry retrieval utilizing a green–yellow CASI band ratio. Validation of the bathymetry map for depths shallower than 4.0 m revealed an R² value of 0.88 and a root-mean-square error of 0.32 m. The assessment of the benthic substrate detectability limits in the Baltic Sea revealed that, at the wavelength of deepest light penetration (near 570 nm), the depth restriction for CASI benthic substrate detection was 7.6 m for sand, 5.0 m for green macroalgae, 3.0 m for higher order vegetation, and 3.1 m for brown macroalgae. The depth limit to which bathymetric mapping is practical in our study site was estimated to be around 3.5 to 4.0 m.

Synthetic aperture radar (SAR) data that can collect information day and night is widely applied in both military and civilian life for security, environmental, and geographical systems. However, detection of rivers in such images is still a challenging problem because rivers are complex with various directions and branches. We aim to detect rivers from SAR images and propose an algorithm combining saliency features, multifeature fusion, and active contour model. The proposed method first filters the image and extracts the global saliency features, which are different from traditional river detection approaches that are mostly based on edge information. A feature fusion technique based on principal component analysis is then applied to merge the saliency features to achieve optimal feature map. Finally, an active contour model is applied to detect the river. Our major contributions are characterizing the rivers by their saliency features, introducing a feature fusion method, and designing an improvement strategy. Experimental results and assessments show that the algorithm is effective and can achieve competitive performance compared with other methods.

The quantity and distribution of photovoltaic power stations play an important role in policymaking related to clean energy. High-resolution satellite imagery with a wide field of view enables photovoltaic power station monitoring at low cost. The separability varies greatly due to distinct backgrounds in different regions, which makes photovoltaic power station identification a challenging task. We propose an end-to-end semantic segmentation network to identify photovoltaic power stations under multiple complex backgrounds in Gaofen 1 imageries. Our network adopts the encoder–decoder architecture in which three modules between the encoder and decoder are added, i.e., feature refinement residual module (FRRM), chained dilation attention module (CDAM), and global channel attention module (GCAM). FRRM uses the structure of residual connection to refine the features from each stage of the encoder. CDAM consisting of chained residual dilated convolutions with different dilation rates and channel attention can enlarge the receptive field without reducing image resolution and select useful features. GCAM utilizes high-level features containing rich semantic information as a guide to select more discriminative low-level features. Experiments demonstrate the effectiveness of each module and the ability of the proposed network to improve photovoltaic power station identification under complex backgrounds.

Land surface temperature (LST) is an essential parameter in investigating environmental, ecological processes and climate change, and thermal infrared remote sensing is a useful tool to acquire information regarding LST. Several accurate LST retrieval methodologies have been developed or refined in recent years and have demonstrated great potential. An assessment of various recent LST inversion single-channel (SC) algorithms is presented. These algorithms include improved mono-window, SC, and improved single channel (ISC). We compared the methods using two Brazilian sites, in which two kinds of validation were performed: field measurements with the satellite overpass and a comparative analysis using the web-based Atmospheric Correction Parameter Calculator tool and the radiative transfer equation (RTE) (assumed as reference). The three methods showed high coefficient of determination with the RTE (between 0.9 and 0.98). SC algorithm produced the furthest results from the reference and was statistically different. ISC algorithm provided the most reliable LST estimates, yielding root mean square errors between 1.53 and 1.91 K. LST can be retrieved through ISC algorithm only using meteorological station data, thus being an alternative for regions where radiosonde points have low density. Our findings contribute to more operational LST products from the Landsat series in humid places.

Foreign object debris (FOD) denotes any unwanted objects on an airport runway, which must be removed before aircraft take off or land. Millimeter-wave (mm-wave) radar is widely utilized to locate small FODs due to its high-range resolution. However, the main difficulty faced by mm-wave radar is to detect stationary little FODs in heavy ground clutter. We propose a layered FOD detection algorithm using clutter map constant false alarm rate (CFAR) and fractional Fourier transform (FrFT) for a mm-wave radar system working at 77 GHz. In the first stage, we utilized the traditional clutter map CFAR to suppress ground clutter while FOD returns and some false alarms were detected using an adaptive threshold. Then, we propose an FrFT-based pattern classification method to distinguish FODs and false alarms, where a two-dimensional feature vector is extracted and a one-class minimax probability machine classifier is trained to accomplish FOD and false alarm classification. Finally, the effectiveness of the proposed method is verified using some measured data obtained via the 77-GHz mm-wave radar.

Reliable crop maps are a vital source of information for many economic sectors. Hyperspectral images and efficient supervised classification algorithms can help in obtaining these maps. One of the objectives is to use a spectroradiometer as a source of a priori information for training supervised classification algorithms. Although there are many supervised classification algorithms, few can be trained using collected spectral signatures, such as spectral angle mapper (SAM). This algorithm relies on setting the angle threshold values manually to separate different classes, and it is usually a random process. So, the second objective is to use an automated process to optimize the angle threshold values to improve the efficiency of the SAM algorithm. An innovative cooperative classification algorithm for hyperspectral images based on enhancing the supervised SAM algorithm performance using a genetic algorithm (GA) is presented. The improvement is based on selecting global optimal threshold angle values for different classes using GA. The efficiency of the developed cooperative evolutionary algorithm is proved by classifying hyperspectral images to create maps of major crops, such as wheat, potato, and alfalfa. The source of the hyperspectral images is the Compact High-Resolution Imaging Spectrometer onboard of the Proba satellite. The evaluation of the results showed that the new cooperative evolutionary algorithm classified hyperspectral images with the highest accuracy compared to well-known reliable supervised classification algorithms.

In spite of widespread adoption, deep convolutional neural networks (DCNNs) have a reputation for remaining “black boxes.” Due to the feature extraction ability of DCNN itself, good predicted results can be achieved without any processing of the raw input images, but it is hard to explain what features of the raw images the results are based on. We propose a preprocessing feature fusion method, based on the Sobel and Canny edge operators, which we treat as the interpretation of the raw images. This method obtains the interpretable edge feature by fusing features in raw ship images, which helps to explain that the DCNN ship classification result is based on the ship edge feature. That is to say, through the interpretation of the raw ship images, we indirectly interpret the classification results of the DCNN to some degree. In subsequent experiments, the highest average accuracy of 85.31% was achieved from the fusion-1 dataset, whereas the highest average accuracy of 74.4% was achieved from the balanced fusion-1 dataset, and the best mean average accuracy of 74.3% is achieved from the cross-validated experiments, which helps to verify the validity of our preprocessing method.

Satellite image restoration from its degraded observation has a significant impact on the remote sensing industry. There are many potential applications that can directly benefit from this technique. A convolutional neural network (CNN) has been recently explored in image restoration and achieved remarkable performance. However, most deep CNN architectures in the literature do not properly handle the inherent trade-off between localization accuracy and the use of global context, which is vital for satellite images covering a ultrabroad area. We present a stacked lossless deconvolutional network (SLDN) for remote sensing image restoration. We fully exploit global context information while guaranteeing the recovery of fine details. Specifically, we design a lossless pooling by reformulating the pixel shuffle operator and incorporate it with a shallow deconvolutional network. The resulting lossless deconvolution blocks are stacked one by one to enlarge the receptive fields without any information loss. We further propose an attentive skip connection and progressive learning scheme to improve gradient flows throughout the SLDN. The SLDN can reconstruct high-quality satellite images without noticeable artifacts. An extensive ablation study is also provided to show that all the components proposed are useful for remote sensing image restoration. Experimental comparisons on various restoration tasks, including super-resolution, denoising, and compression artifact reduction, demonstrate the superiority of the proposed method over state-of-the-art methods both qualitatively and quantitatively.

Time series generated from NOAA operational satellite sensor temperature, noise, and calibration parameter statistics is a critical science analysis tool to trend instrument performance and detect and resolve on-orbit anomalies. Establishing this capability entails ingesting instrument engineering, housekeeping, and calibration data; performing statistics on them; and then storing and providing the resultant data and/or plots. For instruments with relatively small amounts of input and output data, this is a relatively easy task. For the NOAA Geostationary Operational Environmental Satellite R-Series Advanced Baseline Imager (ABI)—with three times more spectral information, four times the spatial resolution, and more than five times faster temporal coverage than previous GOES—instrument performance monitoring can be extremely complex because of the relatively large data volumes and number of parameters. Also of difficulty is that software and computing architecture needed to build such a system is usually proprietary and not openly documented. In order to fill this gap, we focus on the concept of operations and the results associated with the ABI instrument performance monitor. This monitoring system has proven to be extremely valuable in tracking instrument stability and detecting and performing initial diagnosis of ABI anomalies.

In order to effectively alleviate the pressure of high-resolution imaging and massive data storage and transmission, it is of great practical significance to introduce compressed sensing into remote sensing applications. From the perspective of imaging control strategy, the typical block-based compressed sensing (BCS) system is optimized. Based on the fact that there are generally significant differences between regions of remote sensing images, a self-adaptive BCS method is proposed. Compared with the traditional BCS system, the prior information of the imaging target is obtained first by adding a presampling process. On the one hand, it is used to generate a saliency information map, which guides the reasonable allocation of self-adaptive sampling ratios between blocks in the compressed sampling process, thereby improving the sampling efficiency. On the other hand, it is used to generate the weighted sparse coefficient matrix, which will be substituted into the theoretical model in the image restoration process, thus improving the image restoration efficiency. The experimental results show that the imaging quality of the proposed method has a significant improvement compared with the traditional system and is also superior to several existing self-adaptive methods. In addition, on the basis of the above method, a multiangle image restoration strategy is proposed, which further improves the image quality at the cost of four times the image restoration time.

To overcome the difficulties in tracking flying targets in the sky under different conditions, a combined scheme is proposed to get a long-term tracklet: (1) point object is efficiently detected via a fast hypothesis test for background extraction, and an improved correlation filtering algorithm is utilized for possibly near object with much texture; (2) tracker is initialized and managed by estimating continuous motion using an acceleration Kalman filter; and (3) prior knowledge of object is further incorporated to remove false object in tracklet association process. Outdoor experiments prove that the proposed techniques improve the accuracy and reliability for target objects undergoing significant appearance variation due to cloud shift, abrupt motion, and temporary occlusion, and it also extends the validity of our strategy for further valuable applications.

Hyperion images from Earth Observing-1 (EO-1) are being used in natural resources assessment and management. The evaluation and verification of Hyperion images for the above applications are validating the EO-1 mission. However, the presence of random and striping noises in Hyperion images affect the accuracy of the results. Therefore, reduction of random noise and stripes from Hyperion images becomes indispensable for the evaluation of the results in natural resources assessment and in optimum use of the data. Thus, a collective approach for correcting pixels with no-data values and removing random noise and stripes from Hyperion radiance images is developed. In the developed method, first, no-data valued pixels are identified and corrected using a local median filter. Minimum noise fraction transformation is then used to reduce random noise from noise-dominated bands. Further, spatial statistical techniques are used to reduce random noise from the rest of the bands. Finally, a local quadratic regression by a least squares method is used to correct bad columns and global stripes, and a local-spatial-statistics-based algorithm is used to detect and correct local stripes. The effectiveness and efficiency of the algorithm is demonstrated by application to two Hyperion images: one from the Udaipur area, western India, and another from the Luleå area, northern Sweden. The results show that the algorithm reduces random and striping noise without introducing unwanted effects and alterations in the original normal data values in the images.

Traditional waveform design methods mostly assume that the noise in the environment is Gaussian white noise. However, as the electronic warfare environment for the jammer becomes increasingly complicated, there is more likely to be a noise source that can emit fuzzy colored noise and a situation where the radar and jammer game, and the existing waveform design methods cannot meet jammer’s performance demand. Therefore, to reduce the radar’s estimation performance in complex environments, a fuzzy colored noise environment and a robust interference waveform design method under the hierarchical game model between the radar and jammer are proposed successively, which can minimize the mutual information of the radar echo signal and the target impulse response. Extensive simulation experiments show that the designed robust interference waveform can finally optimize the strategy of the jammer and provide a meaningful reference for the interference waveform energy allocation strategy in the future.

An unsupervised change-detection method specifically oriented to improve the change-detection accuracy of synthetic aperture radar (SAR) images is proposed. This method has four main steps. First, a preprocessing method is proposed based on the Kalman filter. We concentrate on reducing the false detection or missing detection rate caused by only using a single filter. Second, the difference image containing explicit information about the changed region is generated based on a standard log-ratio operator. Third, an improved nonsubsampled shearlet transform (NSST) algorithm based on the nonlocal means (NLM) filter is proposed. Simultaneously, we decompose the difference image into low-frequency and high-frequency subbands by the improved NSST algorithm. In particular, in order to effectively preserve the detailed information while denoising, the NLM filter is used to suppress the noise of the high-frequency subbands. Fourth, the final difference image is obtained by the fuzzy C-means algorithm, which is selected because of its high clustering performance and wide range of applications. Experiments conducted on three real datasets of images demonstrate the effectiveness of the proposed method.

A remote sensing image fusion method is proposed to enhance fusion image quality and fusion efficiency. This method is based on the combination of discrete wavelet and multiscale morphological transform in the IHS color space. First, the three-band multispectral image is converted into the IHS color space to obtain the intensity (I), hue (H), and saturation (S) components. Second, the discrete wavelet transform is used to decompose the I component and PAN images, each of which is decomposed into a set of high-frequency detail images at different scales and one low-frequency approximation image. Then, the dual-channel multiscale morphological transform is constructed based on the idea of à trous algorithm. Upon this, the high-frequency detail images are further decomposed by multiscale morphological gradient transform to extract the edge details information. Two low-frequency approximate images are decomposed by multiscale morphological top-hat and bottom-hat transform to extract the light and dark details information. The decomposition coefficients are fused by designed fusion rules. Finally, the corresponding inverse transform is conducted to reconstruct the fused image. The experimental results have shown that the proposed method performs well, in both subjective and objective evaluations. Compared with other seven hybrid methods popularly applied in image fusion, the proposed method is superior in the 12 indicators. In addition, it has significantly reduced the computational cost in the fusion process. The results have proved that the proposed method not only fully preserves the spectral information of the source MS image but also inherits the rich spatial details of the source PAN image while, achieves high efficiency in image fusion.

Proper estimation of the number of endmembers is imperative for proper and accurate spectral unmixing of hyperspectral image. The presence of noise and perturbation poses a challenge in estimation and leads to underestimation or overestimation of the number of endmembers. We propose a noise invariant tensor-based rank estimation approach, which avoids reshaping of the data. We perform a noise-robust PARAFAC tensor decomposition using a generalized low-rank factorization in the initial stage and record the core consistency and reconstruction error for a different number of components. In the next stage, we identify the lower and upper bounds of the number of components. Next, we propose a cost function containing both core consistency value and reconstruction error and identify the number of components from the minima of the cost function. The proposed estimation approach is effective for high, mid, and low signal to the noise level. We analyze the performance of the proposed estimation approach on several synthetic as well as real image experiments and verify the proficiency of our proposed estimation even at higher levels of noise.

Due to the large field angle and large geometric distortions of the HaiYang-1C ultraviolet imager (UVI), a single rational function model (SRFM) is unable to fully absorb the geometric distortions. The SRFM-based sensor orientation accuracy of the UVI images is thereby worse than expected. In order to improve the sensor orientation accuracy, a feasible sensor orientation method for the UVI images based on a piecewise rational function model (PRFM) is proposed. A complete UVI image is first logistically divided into five subimages, and the adjacent subimages have an overlap. Then, an SRFM is used to mathematically fit the physical sensor model of each subimage. Finally, the five SRFMs form a continuous PRFM, and the PRFM-based sensor orientation of the UVI images is performed. Three HaiYang-1C UVI images were tested. The experimental results showed that the unabsorbed geometric distortions fully propagated into the SRFM-based sensor orientation results. The orientation accuracy of the three images reached ∼7  pixels. In the PRFM-based sensor orientation, both the PRFM fitting errors and the inconsistent errors between the adjacent subimages could be negligible. The PRFM-based orientation accuracy was thereby noticeably improved and reached >1  pixel.

The adsorption of gas molecules (SO₂ and H₂S) on a hybrid of graphene oxide and tetrakis(4-carboxyphenyl)porphyrin (GO-TCPP) was studied via photoelectric conversion and protonation. π–π stacking interaction between TCPP and GO was employed to fabricate GO-TCPP film in a sodium dodecylbenzene sulfonate surfactant solution. A GO-TCPP  /  K⁺ exchanged glass optical waveguide sensor was prepared from the fabricated films, and the selectivity of the sensor was studied for volatile organic and acid gases. Results show that the sensor displayed good response to SO₂ and H₂S. Low-gas-concentration experiments detected SO₂ and H₂S concentrations of 1 and 100 ppb, respectively. Target gases were performed at different atmospheric humidities, indicating that humidity has little effect on the sensitivity of the components. Based on these results, the proposed SO₂ gas sensor has great potential for future applications in gas detection and related fields.

Satellite data are used in several environmental applications, particularly in air quality supervising, climate change monitoring, and natural disaster predictions. However, remote sensing (RS) data occur in huge volume, in near-real time, and are stored inside complex structures. We aim to prove that satellite data are big data (BD). Accordingly, we propose a software as an extract-transform-load tool for satellite data preprocessing. We focused on the ingestion layer that will enable an efficient RSBD integration. As a result, the developed software layer receives data continuously and removes ∼86  %   of the unused files. This layer also eliminates nearly 20% of erroneous datasets. Thanks to the proposed approach, we successfully reduced storage space consumption, enhanced the RS data accuracy, and integrated preprocessed datasets into a Hadoop distributed file system.

Achieving parameter measurements and three-dimensional visualization for leaves and wood that have been accurately separated from a whole plant is of great significance. We develop a leaf and wood separation method that combines locally convex connected patches (LCCPs) and K-means++ clustering with poplar seedlings as the research object. First, point cloud data of the research samples are collected using terrestrial laser scanning (TLS). Second, an individual poplar seedling point cloud is separated from the complex background using the pass-through filtering algorithm, and noise and outliers are filtered using the statistical filtering algorithm to improve the separation accuracy. Subsequently, the supervoxel segmentation method is introduced to oversegment the poplar seedlings into patches, and the LCCP clustering algorithm is utilized to segment the leaves and branches. The modified K-means++ clustering algorithm is used to further separate connected leaves and branches. To validate the separation method, numerous experiments on normal and water-deficient poplar seedlings are carried out, and the results illustrate that the segmentation accuracy of the leaves are 93.66% and 90.77%, respectively. The results indicate that the proposed method can effectively separate the leaves and wood of poplar seedlings from TLS data.

An accurate on-orbit characterization of the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the Suomi National Polar-orbiting Partnership satellite is essential to satisfy the requirements from NOAA, NASA, and the general research community for high-quality operational and research products. NASA’s Land Science Investigator-led Processing System (LSIPS) sensor data records (SDRs) are utilized to assess the stability of the calibrated top of atmosphere reflectance over the deep convective clouds (DCCs) and over the Libya-4 desert site. The results from DCCs and desert show detector-to-detector (D2D) reflectance differences in the VIIRS reflective solar bands, 10 moderate-resolution bands (M-bands, M1 to M5, M7 to M11), and three imagery resolution bands (I-bands, I1-I3). More importantly, the D2D differences exhibit an increase in magnitude over time for bands M1 to M5 and I1 to I2, which induces noticeable striping and higher uncertainties in the downstream data products. The largest D2D difference is observed in bands M1 and M2, with magnitudes greater than 1.8% in trends among the 16 detectors and standard deviations less than 0.12% over time. The D2D stability assessment results over DCCs are consistent with those observed over the Libya-4 desert site. To improve the quality of the calibrated SDR reflectance data, NASA’s VIIRS characterization support team has improved the calibration algorithm to incorporate corrections based on these DCC measurements to mitigate the striping (detector differences) observed in the current version of the SDR. These improvements are planned to be included in the next mission reprocessing of the LSIPS land products.

Satisfying the global food demand of the growing population from limited cropland resources is a huge challenge for scientists and policymakers. Approximately 14 mha land remains fallow (rice-fallow) in the dry season after monsoon season crop in South Asia. Accurate geospatial information of cropping intensity is key to identifying the potential areas for crop intensification. In this regard, we aim to map and evaluate cropping intensity dynamics over Odisha, using moderate-resolution multitemporal Landsat OLI (2013 to 14 to 2017 to 18) and TM (2006 to 07) images. The study performed k-mean clustering on principal components extracted from annual surface reflectance composites to identify the presence and absence of crop in the kharif and rabi seasons. The overall accuracy was estimated as 0.92  ±  0.01, 0.91  ±  0.01, 0.92  ±  0.01, 0.89  ±  0.01, and 0.91  ±  0.01 for cropping intensity map of 2006 to 07, 2013 to 14, 2014 to 15, 2015 to 16, 2016 to 17, and 2017 to 18 years, respectively. The analysis of this study revealed that ∼70  %   (3.5 mha) of the total kharif crop area remains fallow during the rabi season. The study illustrated that the dynamics in cropping intensity in the coastal region of Odisha was directly influenced by the occurrence of floods during 2013 to 14 and 2014 to 15. Whereas, extreme and moderate drought during 2015 to 16 and 2017 to 18, respectively, reduced the total crop growing area of rabi season as well as cropping intensity especially in middle and western Odisha. The results and findings of this study can be used for policy-level interventions to target improved agricultural practices for improved productivity and crop intensification. The methodology and approach used in this study can be scaled up in similar areas of South and East Asia and Africa to study the dynamics of cropping intensity.