This a list of reviewers who served JARS in 2018.

As an important part of ocean surveillance systems, a typical application of inverse synthetic aperture radar (ISAR) images is target recognition and classification. Since the target projection on the ISAR image plane depends on relative radar–target geometry and the target’s motion, target recognition and classification becomes a more difficult task when the target is noncooperative. To overcome such a problem, shorebase-shipborne bistatic ISAR is proposed. Although the target of interest is noncooperative, the trajectory of the receiver ship platform can be controlled to change the relative radar–target geometry aiming to obtain a desired ISAR image projection. The Doppler frequency model of shorebase-shipborne bistatic ISAR imaging of the ship target in the presence of a low sea state is derived. An optimal trajectory of the receiver ship is presented that is able to obtain desired ISAR images, in which the cross-range coordinate of a scatterer is proportional to its coordinate x in space. The correctness of the proposed Doppler frequency model and effectiveness of trajectory selection are verified by means of several imaging simulations.

Region map is the sparse representation of a high-resolution synthetic aperture radar (SAR) image on the middle-level semantic layer in its semantic space. Based on the semantic information of the region map, the high-resolution SAR image is divided into hybrid, structural, and homogeneous pixel subspaces. The segmentation of SAR images can be divided into these three subspaces segmentation, of which the segmentation of hybrid subspace has more challenge because of complex structures. There are often many extremely inhomogeneous areas in the hybrid pixel subspace. Are these nonconnected areas in the same or different classes? To solve this problem, a Bayesian learning model with the constraint of sketch characteristic and an initialization method is proposed to construct a structural vector that can reflect the essential features of each extremely inhomogeneous area. Then, the unsupervised segmentation of the hybrid pixel subspace can be realized by using the structural vectors of these areas in this paper. Theoretical analysis and experimental results show that the performance of the hybrid pixel subspace segmentation realized by the structural vectors based on the Bayesian learning model proposed in the paper is better than that only used by hand designing features.

Commercial satellites for Earth observation can integrate conventional positioning and tracking systems for monitoring legal and illegal activities by sea, in order to effectively detect and prevent events threatening human life and environment. This study describes an object-oriented approach to detect vessels combining high- and medium-resolution optical and radar images. Once detected, the algorithm estimates their position, length, and heading and assigns a speed range. Tests are done using WorldView-2, QuickBird, GeoEye-1, Sentinel-2A, COSMO-SkyMed, and Sentinel-1 data imaged in several test sites including China, Australia, Italy, Hong Kong, and the western Mediterranean Sea. Validation of results with data from the automatic identification system shows that the estimates for length and heading have R²  =  0.85 and R²  =  0.92, respectively. Tests for evaluating speed from Sentinel-2 time-lag image displacement show encouraging results, with 70% of estimates’ residuals within ±2  m  /  s. Finally, our method is compared to the state-of-the-art search for unidentified maritime object (SUMO), provided by the European Commission’s Joint Research Centre. Finally, our method is compared to the state-of-the-art SUMO. Tests with Sentinel-1 data show similar results in terms of correct detections. Nevertheless, our method returns a smaller number of false alarms compared to SUMO.

This study is aimed at evaluating the effectiveness of different supervised and unsupervised methods with information derived from Landsat satellite images and fieldwork in order to maximize the land cover classification accuracy in an area with geomorphologic differences and heterogeneous edaphic characteristics located in the southwest of the Pampas (Argentina). We test two datasets: bands-based and indices-based and also we analyze the spectral behavior of each land cover identified by field trips and surveys with farmers to improve the spatial samples employed in the digital processing. Complementarily, we study the spatial and temporal information about the land cover changes during 2000 to 2016. The classification based on indices widely outperforms the analyses based on bands. The best methods to classify the land cover are the Mahalanobis distance and the maximum likelihood. The values of kappa coefficient and overall accuracy obtain from these two methods allow us to realize a multitemporal study. This study provides essential information for semiarid regions with rain-fed agriculture and livestock activities worldwide. The knowledge obtained quickly and accurately about the land covers and their changes provides essential information about the past and current situations and can be used to predict likely future trends.

Accurate estimation of grassland coverage is important in both monitoring grassland growth and promoting grassland management. This study, for the first time, aims to evaluate the alpine grassland coverage estimation performances of four widely used methods [i.e., random forest classification (RFC), regression analysis (RA), multiple endmember spectral mixture analysis (MESMA), and support vector machine regression (SVMR)] from three typical remote sensing images [i.e., hyperspectral (HJ-1A/HSI) image and multispectral (Landsat 8 and Sentinel-2A) images] in the Three-River Headwaters region, China. The total grassland coverage and three levels of grassland subcoverage (i.e., low, moderate, and high coverage) are estimated from the three images by each method. Meanwhile, the overall accuracy (OA) and root-mean-square error (RMSE) of each coverage result are assessed. The experimental results show that (1) for the total grassland coverage estimation, RFC method combined with Sentinel-2A data and RA method used both Landsat 8 and Sentinel-2A data generated the highest OA of 79.4%, whereas the highest OAs of SVMR and MESMA methods are 3% to 14.7% lower than those of RFC and RA methods, (2) for three grassland subcoverage estimation, Landsat 8 data combined with RFC method generated the greatest OAs (100%) and the lowest RMSEs (7.34%) for low grassland coverage and Sentinel-2A data combined with SVMR method obtained the highest OAs (≥83.3  %  ) for moderate and high grassland coverage, (3) Sentinel-2A and Landsat 8 data generated higher OAs than HJ-1A/HSI data in estimating the total coverage while Sentinel-2A produced better performance than Landsat 8 in estimating the grassland subcoverage, and (4) SVMR method performed more stably than other methods in estimating alpine grassland coverage.

Monitoring the marginal ice zone (MIZ) is becoming increasingly important due to recent evidence that the width of the MIZ is changing with climate. A method to automatically detect the MIZ in synthetic aperture radar (SAR) imagery is proposed. The method utilizes the curve-like features of MIZ in SAR images. A multiscale strategy, the curvelet transform, is chosen to extract features from the SAR images. The statistical and co-occurrence features of curvelet coefficients at an appropriate scale are used to identify the MIZ from open water and consolidated ice. Experimental results show a significant increase in classification accuracy (89.7%) compared with the most commonly used MIZ definition from passive microwave sea ice concentration (74%), especially in the diffuse MIZ.

Remotely sensed data have become a valuable source of data for mapping regional or global land cover. Referable to the spectral complexity and the quality of remotely sensed data selected, classification errors usually occurred in land cover mapping, where the errors occurred can result in an issue in which the spatial heterogeneity in remotely sensed data is inconsistent with that in land cover data. To ensure data quality, land cover data checking mainly depends on manual labor, which is time-consuming and labor-intensive. To tackle this issue, a spatial heterogeneity-based Voronoi matching method for automated detection of land cover data inconsistency is proposed. First, Harris corner-based Voronoi diagrams (HCVDs) are constructed utilizing the image noise filter and Harris corner detector. Second, the regional statistics of range on land cover data using HCVDs is calculated. Ultimately, by matching the heterogeneity in HCVDs and using HCVDs-based first-order neighbor filter, where inconsistent spatial heterogeneity occurs in land cover data can be found. A GlobeLand30 case study from Linqu County, Shandong, for the year 2010 in China is carried out. After detection and validation, this method is shown to be workable for detecting spatial regions where inconsistent spatial heterogeneity occurs in land cover data, and the data quality is improved by 6.3% in this study area after manual verification.

Due to the presence of phase noise and temporal decorrelation, high-quality pixel selection has been an indispensable part in ground-based synthetic aperture radar (GB-SAR) interferometry. The main pixel selection methods are permanent scatterer (PS) methods and coherent scatterer methods. However, both of them cannot properly solve the false-selection and the miss-selection caused by intrinsic scattering and temporal characteristics of the pixel on the whole dataset. Pseudo-PS and quasi-PS are first introduced into the high-quality pixel selection of GB-SAR interferometry, and a temporal selection method (TSM) is proposed to avoid the miss-selection of quasi-PS. Then, in order to achieve appropriate pixel density and high-quality pixels for deformation monitoring, a high-quality interferometric pixel selection method based on pseudo- and quasi-permanent scatterers analyses is proposed, which is a joint processing technique combining spatial variant apodization with TSM. Finally, the effectiveness of the proposed high-quality pixel selection method is validated by the field measurements of GB-SAR data.

Lunar calibration is a commonly used method to track a climate satellite sensor’s long-term radiometric stability. We present a modeling approach to examine the satellite sensor lunar observation uncertainties due to several important aspects related to the lunar image acquisition by the satellite sensor: lunar pixel shift, point spread function (PSF), lunar orientation, pitch, and oversampling rates. Our analyses can be summarized as follows. (1) The sensor observed lunar irradiance can vary due to small lunar pixel shift if the PSF is less than ideal. (2) During lunar calibration, an unstable oversampling rate due to spacecraft control will result in errors in observed lunar irradiance. A drift in oversampling rate would result in a bias in observed lunar irradiance and a random variation in oversampling rate would cause random error in lunar irradiance. Increasing the overall oversampling rates can reduce random error in observed lunar irradiance but would not change the biases in the observation. (3) Furthermore, the biases can vary when the Moon is observed at different orientations. Our results show impacts on observed lunar irradiance are on the order of 0.1%, which is a significant part of the overall uncertainty for a lunar irradiance measurement of a climate satellite sensor.

On June 24, 2017, a catastrophic landslide occurred in Maoxian County, Sichuan, China, burying an entire village and more than 100 people. We focused on mapping and predicting the deformation in Maoxian before the landslide based on 39 C-band synthetic aperture radar images acquired by the Sentinel-1 satellite. In order to identify sufficient measurement points in rural areas, the images were first filtered with joint scatterer interferometric synthetic aperture radar to increase the number of measurement points 20 times. Then, the cumulative time series deformation in the landslide area from November 7, 2014, to June 12, 2017, was obtained by employing a time series InSAR technique combined with an unscented Kalman filter (UKF) to predict the deformation by separating the prediction and update steps. Approximately, 30 suspected landslide points were found in the area. Qualitative and quantitative analyses of the experimental results suggested that combining time series InSAR with UKF can be used to predict deformation trends with millimetric accuracy at a large scale before landslides occur.

Remote sensing is providing an increasing number of crucial data about Earth. Systematic revisitation time allows the analysis of long time series as well as imagery utilization in the most interesting moments. Nevertheless, the current huge amount of data makes essential the usage of automatic methods to select the best captures, as many of them are not useful because of clouds, shadows, etc. Because of that, one of the characteristics of the more recent missions is the distribution, along with the spectral data, of a large amount of quality ancillary datasets. These datasets can act synergistically in the aim of selecting the best quality images, but the criteria they provide are not always enough. Indeed, these datasets are often used on a per pixel basis and the spatial pattern of the different spectral bands is forgotten, so ignoring the key information they can provide for our goals. With this aim, our work takes one of the most successful instruments in remote sensing, MODIS, and demonstrates, through geostatistical techniques, that the role of the spatial patterns of the spectral bands can effectively improve image selection in a complex (for climate, relief, and vegetation and crop phenology) region of 63,700  km². The results show that band 01 (red) is the preferred one, as it achieves a 13% higher success than when only using quality bands criteria: a 94% global accuracy (66 true classifications, and only four omissions and one commission error). A second, important finding, is that the geostatistical selection improves results when using any band, except for band 02 (NIR1), which makes our proposal potentially useful for most remote sensing missions. Finally, the method can be executed in a reasonable computing time due to previously developed high-performance computing techniques.

We are the first to apply the copper stress vegetation index (CSVI) on a remotely sensed image and the purpose is to verify the effectiveness of CSVI at the satellite-image scale. The study area was located at the Dexing Copper Mine, Jiangxi Province, China. First, the data preprocessing for the Hyperion image was conducted, including bands removal, radiometric calibration, and atmospheric correction. Second, the regions with high vegetation cover were extracted based on endmembers extraction and spectral unmixing. The CSVI was calculated on the high-vegetation-cover regions. Third, the samples of soil and leaves were collected from the study area and the copper contents in the samples were measured for the assessment. The results showed that the high values of the CSVI were near the functional regions for copper mining and the polluted rivers. What is more, there was a significant positive correlation between the CSVI calculated from the Hyperion image and the copper content in soil/leaves. We demonstrate that CSVI is applicable for monitoring the copper stress on vegetation using satellite hyperspectral images. In addition, we provide a complete example for the application of CSVI at the satellite-image scale for the first time, which is helpful for the community in remote sensing of copper-stressed vegetation.

The compressed sensing (CS)-based methods for tomography synthetic aperture radar (SAR) imaging have a good performance in the case of a large number of baselines. Unfortunately, for the current tomography SAR, the baselines are obtained in repeat-pass mode from a large number of parallel passes in the same scene. That is expensive and can be severely affected by temporal decorrelation. And the performance of CS-based methods degrades rapidly with the decrease of the number of baselines. Inspired by the block-sparsity theory, an extended block orthogonal matching pursuit algorithm by using the weighted operation and a new column-block selection strategy is proposed for tomography SAR imaging. By using neighboring pixels information in reconstruction, the proposed method can improve the performance of CS-based methods in the case of strong noise and a small number of baselines. Experimental results confirm the effectiveness of the proposed method.

Defoliation induced by the weevil Gonipterus scutellatus is causing significant damage to South Africa’s eucalyptus plantations. Therefore, the ability of remote sensing to detect and map G. scutellatus defoliation is essential for monitoring the spread of the weevil so that precautionary measures are set in place. In our study, an integrated approach using image texture in various processing combinations and an artificial neural network (ANN) were developed to detect and map G. scutellatus induced vegetation defoliation. A 0.5-m WorldView-2 pan-sharpened image was used to compute texture parameters from the gray-level occurrence matrix and gray-level co-occurrence matrix, using optimal moving windows for specific levels of G. scutellatus induced vegetation defoliation. In order to improve the accuracy of detecting and mapping G. scutellatus induced vegetation defoliation, a method involving a three-band texture processing combination was proposed and tested. Using a sequential forward selection algorithm allowed for the selection of optimal texture combinations, which were subsequently input into a backpropagation ANN. The results showed an improvement in detecting vegetation defoliation using single texture bands [R²  =  0.82, root mean square error (RMSE) = 0.95 (1.82% of the mean measured defoliation)] when compared to single spectral reflectance bands [R²  =  0.60, RMSE = 1.79 (3.43% of the mean measured defoliation)], two-band spectral reflectance combination model [R²  =  0.74, RMSE = 1.48 (2.83% of the mean measured defoliation)], and the three-band spectral reflectance combination model [R²  =  0.80, RMSE = 1.35 (2.59% of the mean measured defoliation)]. Further improvements were obtained using the two-band texture combination model [R²  =  0.85, RMSE = 1.05 (2.01% of the mean measured defoliation)] and the most promising result was obtained using the proposed three-band texture combination model [R²  =  0.90, RMSE = 0.85 (1.63% of the mean measured defoliation)]. Overall, our study highlights the potential of image texture combinations in improving the detection and mapping of vegetation defoliation.

The land cover classification is an important task in geoscience applications. Many methods and implementations are based on multispectral data processing. The presented work aims to benefit from the nonlinear source separation process to enhance land cover identification. The source separation technique aims to provide underlying images and to compensate the mixing process. Nonlinear separation is more realistic due to multiple distortions occurring on the radiance path from soil to sensors. The presented paper addresses pattern recognition for remote sensing and proposes a framework based on feature extraction and decisional fusion. The first stage performs a nonlinear separation model based on Bayesian inferences. Nonlinearity is approximated by a multilayer neuron network. The separation process updates knowledge about unknown sources and model parameters iteratively. The second stage performs feature extraction. Based on a decisional fusion, the third stage realizes a classification process. This fusion scheme assigns, first, a suitable feature to each source/pattern based on the learning data set. Second, a majority vote determines the final label. Experimentation results demonstrate that the proposed fusion method enhances the recognition accuracy and represents a powerful tool for land identification.

An anisotropic diffusion model based on a log-normal distribution of a local gray-level is used to propose a way to denoise the panchromatic images. The implication of the low-edge gradient of the feature space for denoising and smoothing the noisy image is adaptively adjusted by the adaptive threshold parameter in a diffusion coefficient function. Furthermore, to terminate the diffusion process, an entropy-based stopping criterion is implemented. The proposed model is compared with the existing models such as Perona–Malik (PM), adaptive PM, difference eigenvalue PM, modified PM, and Maiseli–Gao. In order to analyze the performance of the models, quantitative metrics such as standard deviation, entropy, and the signal-to-noise ratio of a two-dimensional line profile are used. For further analysis, the results of denoising models are segmented using entropy-based segmentation techniques such as Harvda, Renyi, Kapur, and Yen models. A misclassification error metric is used to evaluate the segmentation results. The metric results show that the proposed model effectively removes the noise and preserves the features of a panchromatic image.

Vegetation water content (VWC) is an important land surface parameter that is used in retrieving surface soil moisture from microwave satellite platforms. Operational approaches utilize relationships between VWC and satellite vegetation indices for broad categories of vegetation, i.e., “agricultural crops,” based on climatological databases. Determining crop type–specific equations for water content could lead to improvements in the soil moisture retrievals. Data to address this issue are lacking, and as a part of the calibration and validation program for NASA’s Soil Moisture Active Passive (SMAP) Mission, field experiments are conducted in northern central Iowa and southern Manitoba to investigate the performance of the SMAP soil moisture products for these intensive agricultural regions. Both sites are monitored for soil moisture, and the calibration and validation assessments had indicated performance issues in both domains. One possible source could be the characterization of the vegetation. In this investigation, Landsat 8 data are used to compute a normalized difference water index for the entire summer of 2016 that is then integrated with extensive VWC sampling to determine how to best characterize daily estimates of VWC for improved algorithm implementation. In Iowa, regression equations for corn and soybean are developed that provided VWC with root mean square error (RMSE) values of 1.37 and 1.10  kg  /  m², respectively. In Manitoba, corn and soybean equations are developed with RMSE values of 0.55 and 0.25  kg  /  m². Additional crop-specific equations are developed for winter wheat (RMSE of 0.07  kg  /  m²), canola (RMSE of 0.90  kg  /  m²), oats (RMSE of 0.74  kg  /  m²), and black beans (RMSE of 0.31  kg  /  m²). Overall, the conditions are judged to be typical with the exception of soybeans, which had an exceptionally high biomass as a result of significant rainfall as compared to previous studies in this region. Future implementation of these equations into algorithm development for satellite and airborne radiative transfer modeling will improve the overall performance in agricultural domains.

Oasis agroecosystems monitoring plays a significant role in the economic development, sustainable management, and policy-making of arid and Saharan regions. The aims of this study are to analyze the spatiotemporal changes of oasis vegetation and discuss possible driving forces of changes. This analysis employed field and ancillary data, geographic information system, Landsat imagery, and remote-sensing techniques. Minimum noise fraction is used for endmembers extraction, and spectral mixture analysis is applied to each image to extract vegetation fraction, which is used as an indicator of change. Change detection is performed in six oases in south-eastern Morocco over eight separate periods from 1984 to 2017 using Landsat data. The pattern of the spatiotemporal changes in vegetation cover is analyzed using time- and space-oriented change detection algorithms. Results indicate that the Mezguita, Tinzouline, and Ternata oases had an evenly distributed and constant expansion during the last three decades (of ∼41.35  %  ), whereas the oases of Fezouata and Ktaoua presented low expansion and randomly distributed change (∼29  %  ). However, the vegetation cover of M’Hamid oasis decreased by 23% over the studied period. The results show that spectral mixture analysis yields high accuracies for oasis vegetation extraction in arid areas and accounts for mixed pixel issues. The results are discussed considering also climate and socioeconomic factors, showing that the driving forces of these dynamics are primarily anthropogenic.

Generalization of the KA distribution is formulated by combining the class A and K distributions; the resulting distribution is termed as generalized KA distribution. It is obtained as a mixture of a generalized Rayleigh and a class A distribution with gamma-distributed mean intensity, and it may be used to describe clutter statistics. Its parameters are estimated by implementing the expectation maximization algorithm. The latter provides estimates in the framework of the maximum likelihood principle, and it is widely used when the data set is incomplete and/or of limited size. The numerical results show that the absolute relative error of the estimated parameters may be <8  %   even in the case of 100 data samples, whereas it is reduced significantly as the size of the data set increases.

A sparse and low-rank matrix decomposition-based method is proposed for anomaly detection in hyperspectral data. High-dimensional data are decomposed into low-rank and sparse matrices representing background and anomalies, respectively. The problem of the decomposition process is defined from the dictionary learning point of view. Therefore, our way of obtaining these matrices differs from previous studies. It aims to find a correct partition of the data and separate anomaly pixels from the background. After decomposition, Mahalanobis distance is applied to the sparse part of the data to get anomaly locations. Three hyperspectral data sets are used for evaluation. Experimental results suggest that anomaly detection performance of the proposed method surpasses those of the state-of-the-art methods.

Inverse synthetic aperture radar is driven to a larger bandwidth to achieve a finer resolution. However, severe amplitude-phase distortion of radar greatly destroys the amplitude and phase of the chirp signal and hence becomes a major challenge to the fulfillment of the desired high-resolution performance. An adaptive framework is proposed to compensate for the distortion of a wideband radar system without the measurement of its intrinsic characteristic. The distorted image of the target is obtained first and then used to get cross-range profiles (CRPs). Based on the CRPs, amplitude distortion and phase distortion are compensated by the noise level and the autofocus algorithm, respectively. The proposed framework facilitates the compensation and works well in practical applications. Performance of the proposed framework is validated on simulated and measured data.

Land-cover classification using remote sensing imagery is an important part of environmental research because it provides baseline information for ecological vulnerability and risk assessment, disaster management, landscape conservation, local and regional planning, and so on. Rural-land-cover classification is challenging for both object-based image analysis methods and classifiers. The objective of this study is to improve the object-oriented classification accuracy of rural land cover by combining two models based on high spatial resolution imagery. We apply the C5.0 algorithm in R to combine support vector machines (SVMs) and random forest (RF) to create the model RS_C5.0. The effectiveness of the model combination is assessed by comparing the classification results with the state-of-the-art machine learning algorithm, namely extreme gradient boosting (XGBoost). The comparisons are done based on the classification results of both the study area and the case area. Results show that in the classification of the study area, RF performs slightly better than SVM, and XGBoost performs worse than RF but better than SVM. However, in the classification of the case area, SVM performs slightly better than RF and both SVM and RF perform better than XGBoost. Furthermore, RS_C5.0 obtains the highest overall accuracies and kappa coefficients in the classifications of both the study area and the case area. In terms of training time, XGBoost runs the slowest in the classifications of both the study area and the case area. SVM and RF as well as the combined model (RS_C5.0) run much faster than XGBoost classifier. To summarize, the combination of SVM and RF classifiers using C5.0 algorithm is found to be a fast and effective way to improve rural-land-cover classification.

Satellite-based rainfall products are essential for retrieving rainfall, particularly in data-scarce and drought-prone countries like Ethiopia. However, their quality needs to be validated prior to their use. Therefore, we evaluated the performance of the Climate Hazards group Infrared Precipitation with Stations version 2.0 (CHIRPS), the Tropical Applications of Meteorology using Satellite data version 3.0 (TAMSAT3), and the African Rainfall Climatology version 2 (ARC2) satellite rainfall estimates in the Ethiopian Rift Valley Lakes Basin. Their skill of retrieval was evaluated against ground-measured rainfall at dekadal, monthly, and seasonal scales across agroclimatic zones over 2001 to 2017. Finally, these satellite products have demonstrated different levels of agreement with the reference data, being the highest for CHIRPS and the lowest for ARC2. At all timescales and agroclimatic zones, ARC2 has severely underestimated the actual rainfall while TAMSAT3 has persistently overestimated it. However, TAMSAT3 has demonstrated better performance than ARC2. Generally, except for its slightly larger dekadal false alarm ratio, CHIRPS has achieved the highest and most consistent agreement with the reference data at all the timescales and agroclimatic classes. Consequently, CHIRPS was further assessed for its suitability of drought monitoring, and it has exhibited promising skill in detecting specific historical drought events. Therefore, to overcome the scarcity of ground-measured rainfall data in the study area, we recommend the CHIRPS rainfall estimate to be used as an alternative data source for drought monitoring. Conversely, owing to its overestimation tendency, TAMSAT3 could be used for flood monitoring in this region.

Fluvial systems offer a challenging and varied environment for topographic survey, displaying a rapidly varying morphology, vegetation assemblage, and degree of submergence. Traditionally, theodolite- or global positioning satellite-based systems have been used to capture cross-section and breakline-based topographic data, which have subsequently been interpolated. Advances in survey technology have resulted in an improved ability to capture larger volumes of information with infrared terrestrial and aerial Lidar systems capturing high-density (<0.02  m) point data across terrestrial surfaces. The rise of structure from motion (SfM) photogrammetry, coupled with small, unmanned aerial vehicles (sUAV), has the potential to record elevation data at reach scale subdecimeter density. The approach has the additional advantage over Lidar of seeing through clear water to capture bed details and also generating orthorectified photographic mosaics of the survey reach. However, data accuracy has yet to be comprehensively assessed. We present a survey protocol for sUAV deployment and provide a reach scale comparison between a theodolite and SfM sUAV survey on the River Sprint at Kendal, the River Ehen at Egremont, England, and the Afon Elwy at Llanfair Talhaiarn, Wales. Comparative analysis between theodolite survey and SfM suggests similar accuracy and precision across terrestrial surfaces with error lowest over solid surfaces, increasing with vegetation complexity. Submerged SfM data captured bed levels generally to within ±0.25  m at depths of <2.4  m, with only a weak relationship recorded between error and flow depth. Significantly, associated error when linked to channel D₅₀ highlights the ability of unmanned aerial vehicles to capture accurate fluvial data across a range of river biotopes and depths to 2.4 m.

Hyperspectral technology is particularly good at identifying hydrothermally altered minerals and can offer support for three-dimensional (3-D) alteration modeling of hydrothermal deposit. We propose a 3-D alteration modeling method with hyperspectral core imaging to deepen hydrothermal alteration research for uranium exploration of the Heyuanbei uranium deposit in the western Xiangshan ore field, Jiangxi, South China. We use the following three steps: (1) identify and map the altered minerals of drill cores, (2) quantify all types of altered minerals using a pixel classification-based statistical algorithm, and (3) establish 3-D alteration model using a parallel section and a triangulated irregular network. Hyperspectral core imaging data from 14 drill cores from the Heyuanbei uranium deposit are used to establish a 3-D alteration model of the superimposed stratigraphy and structural information by this method. The modeling results indicate that large-scale illite and kaolinite–dickite alterations have developed in the Heyuanbei deposit. The illite alteration has the following obvious zoning characteristics: (1) the spatial correlation of short-wavelength illite and kaolinite–dickite, which developed in two volcanic formations, is high, and the alteration is stronger in the upper part of the drill cores than in the lower part of the drill cores; (2) the long-wavelength illite is mainly distributed in the lower volcanic formations. The zoning characteristics reveal that the hydrothermal fluid is acidic in the upper cores, whereas in the lower cores, alkaline fluid is present, and the fluid environment has the characteristics of a transition from early alkaline to late acidic conditions. Moreover, the short-wavelength illite is more closely related to fluorite-hydromica-type uranium mineralization than the long-wavelength illite, and the conversion action from long-wavelength illite to short-wavelength illite is favorable for uranium enrichment. Thus, short-wavelength illite, kaolinite, dickite, and fluorite can be used as prospective indicators for uranium exploration in the Heyuanbei mining area and nearby areas.

Surface roughness is an important characteristic in analyzing synthetic aperture radar images and three-dimensional (3-D) surfaces that influence the backscattering of electromagnetic waves. We propose an algorithm based on a fractal method for retrieving parameters of a surface. To estimate surface roughness parameters, 10,000 different 3-D surfaces with different rms-height and correlation length are simulated. 3-D simulation of surfaces is carried out based on fractal Brownian motion. The results show that the fractal method represents a good relationship between roughness parameters and fractal dimension. It can be seen that, in some cases, surfaces with different roughness have the same fractal dimension. Roughness index (RI) can be used as a complement to fractal dimension. We present empirical relationships among fractal dimension, roughness parameters, and RI. The method is implemented on L and C bands with HH and HV polarizations of AIRSAR data at different dates. The comparison between field measurement and estimated roughness showed that the accuracy of soil surface roughness estimation for band L with HH polarizations is better than bands L and C with polarizations HV and HH, respectively. The results of this method also are compared with the roughness estimates using the integral equation model (IEM). The analysis of outputs shows that the roughness estimation using the fractal and IEM is very similar in the low moisture at L band in HH polarization. The root mean square error of roughness for data at L band in HH polarization on July 1, 2002, is 0.54 and 0.53 cm for fractal and IEM, respectively.

The techniques of unsupervised band selection are important in the processing of hyperspectral images, such as dimension reduction and abstraction of meaningful information. These techniques aim to select the most informative bands from the original dataset using some similarity metrics and search strategies. For this purpose, the structural information of the images can be utilized in many applications, such as the compact representation and classification of hyperspectral images. Hierarchical clustering method was chosen as the searching strategy and interest points, such as clear edges, blobs, and boundaries of objects, were used as the structural information during the selection of representative bands. At the same time, a similarity criterion was developed to generate a representative band subset from the hyperspectral bands. Finally, the representative band subsets obtained from the proposed methods were classified by the k-nearest neighborhood algorithm and the classification accuracy results were taken as performance criteria. The performances of the proposed methods were compared with the Walumi and Waludi techniques that are common in the literature. Experimental results show that the proposed methods give better results than the others.

Aerosol types were characterized and classified using multispectral satellite data. The role of near-UV data in the detection of absorbing aerosols, such as biomass burning aerosols (BBA) or mineral dust particles (DUST), was examined on a global scale. An absorbing aerosol index (AAI) was proposed and defined as the ratio of the satellite-observed radiance (R) at a wavelength of 0.412  μm [R (0.412)] to that at 0.380  μm [R (0.380)] that can also detect nonabsorbing-type aerosols. Initially, the numerical AAI values were estimated for the BBAs and DUST from measurements collected by the Advanced Earth Observing Satellite-2/Global Imager (ADEOS-2/GLI). The Japanese short mission ADEOS-2 carried the GLI instrument with observation channels in the near-UV region. Not only the AAI index but also the short-wavelength infrared measurements were utilized to determine the dust detection index (DDI) defined as the ratio of R (2.210) to R (0.380) in order to discriminate BBAs from DUST. In addition, the AAI and DDI values were evaluated for the detection of clouds. The results allowed the classification criteria for DUST, BBA, other types of aerosols and clouds to be obtained. The Second-Generation Global Imager (SGLI) sensor is onboard the Japanese Global Change Observation Mission-Climate (GCOM-C) (SHIKISAI in Japanese) satellite launched on December 23, 2017. The SGLI has multiple channels (19) including near-UV and polarization sensors in the red and near-IR wavelengths. We also demonstrated the advantages of the SGLI for near-UV and polarization data for aerosol remote sensing. An understanding of aerosol types facilitated subsequent aerosol retrieval. Then, retrieval for classified aerosols was made based on the radiation simulations with multispectral radiance by GLI and polarization measurements by Polarization and Directionality of the Earth’s Reflectances (POLDER)-2, respectively, mounted on the ADEOS-2 satellite. The proposed algorithms are expected to be available not only for the analysis of the SGLI data but also for other future missions.

There are different algorithms to estimate the energy balance equation. This equation expresses net radiation converting to other forms of radiation fluxes. Thus, the equation consists of net radiation (R_n), sensible heat flux (H), latent heat flux (λE), and soil heat flux (G). The computation of the equation is helpful to consider energy exchange between the atmosphere and the surface. Therefore, testing different methods could help us in finding a more accurate way to estimate the energy balance equation. The simple algorithm for evapotranspiration retrieving (SAFER) algorithm is a method to estimate energy balance equation parameters. Mashhad area in northeast of Iran located in a semiarid region was chosen as the study area. R_n, H, λE, and G were calculated by the SAFER algorithm (method 1). In addition, R_n, H, λE, and G were computed based on a theoretical formula (method 2). Surface albedo (α₀) and surface temperature (T₀) were the two input parameters used in both methods. In method 1, α₀ and T₀ values are spatially distributed in Mashhad while they are constant values in method 2. The results of method 1 and method 2 were compared. The obtained results showed ±2  Mj m^−  2 d^−  1 differences for R_n, H, and λE variables in both methods. Albedo and surface temperature values that were spatially distributed in method 1 lead to a wide variation range in the calculated energy terms. However, these parameters, which were considered as constant values, made energy terms have limited variation range in method 2. Therefore, the differences in the obtained results between the two methods resulted from α₀ and T₀ values. The differences in the values were insignificant in the large-scale calculation, which can be ignored. The SAFER algorithm showed acceptable results in the semiarid of Mashhad region. This algorithm can be used for energy balance equation estimation in Mashhad.

Ability to acquire high-resolution, phase-preserved images is an important requirement in modern synthetic aperture radar (SAR) applications. Wavenumber domain algorithms are good choices to achieve this objective without approximation in the wavenumber reversal process. However, when squint angle gets larger, even wavenumber domain algorithms suffer from resolution and computational efficiency degradations. In order to accommodate wavenumber domain algorithms to highly squinted SAR imaging by exploiting squint minimization and two-dimensional interpolation schemes, the proposed extended wavenumber domain algorithm possesses the advantages of maximum spectrum utilization, processing efficiency, and good interpretation of the two-dimensional interpolation. It can also be easily incorporated with motion compensation without complicating the motion analysis. Simulations and real data experiments validate our proposed algorithm.

In the context of land-use and land-cover (LULC) classification, there is a lack of leverage of the recent increase of the ease of access to satellite imagery data, cloud computing platforms, and classification techniques. We present both the development of an operational method for LULC classification that considers these progresses and the implementation of this operational method for a mountainous area of Nepal. The operational method allows the comparison of three LULC maps, each derived with a different classification technique [classification and regression tree (CART), max entropy (MaxEnt), and random forest (RF)] applied to Sentinel-2 data on the Google Earth Engine platform. The results derived with the RF technique have the highest overall accuracy coefficient (92%). The probabilities that the RF technique produces a more accurate LULC map than the MaxEnt (95%) and CART (61%) techniques are based on Kappa statistics. Results of general linear models suggest that some LULC types have higher producer’s and user’s accuracies at a statistically significant level. The operational method can help the producers of LULC maps conduct future work on areas in developing countries, as such contributing to addressing various issues that involve land use.

A multitemporal approach to discriminate freshwater macrophyte vegetation types in the Lower Paraná River floodplain is addressed. During a low-intensity flood pulse, seven X-band HH COSMO-SkyMed HImage images are acquired, covering a nine-month period. Scenes are segmented with a mean-shift segmentation algorithm. Objects are classified with an expectation maximization algorithm into clusters with different temporal signatures and are assigned to six information classes: water, bulrush marshes, short broad-leaf marshes, tall broad-leaf marshes, short grasslands and grass marshes, and tall grasslands and grass marshes. Class interpretation is based on backscatter dynamics, with focus on their correlation with hydrometric water level measured in the Paraná River and/or with the floodplain area covered by water as estimated with a normalized difference vegetation index threshold criterion. The obtained product has a global accuracy of 75.4% and a kappa index of 67.2%. We point out the usefulness of X-band for flood monitoring and macrophyte vegetation type discrimination. However, we find limitations for the discrimination between high-biomass vegetation targets, such as tall broad-leaf marshes and tall grasslands. In a mosaic of herbaceous wetlands, the knowledge on the relation between vegetation and floods is essential for interpreting and predicting how backscattering coefficients and other synthetic aperture radar-derivated parameters vary with flooding.

In hyperspectral image (HSI) analysis, dimensionality reduction is a preprocessing step for HSI classification. Independent component analysis (ICA) reduces the spectral dimension and does not utilize the spatial information of the HSI. To solve it, tensor decompositions have been successfully applied to joint noise reduction in spatial and spectral dimensions of HSIs, such as parallel factor analysis (PARAFAC). However, the PARAFAC method does not reduce the dimension in the spectral dimension. We proposed a method to improve it, which combines ICA and PARAFAC to reduce both the dimension in the spectral dimension and the noise in the spatial and spectral dimensions. The experimental results indicate that this method improves the classification compared with the previous methods.

Aerial images acquired by multiple sensors provide comprehensive and diverse information of materials and objects within a surveyed area. The current use of pretrained deep convolutional neural networks (DCNNs) is usually constrained to three-band images (i.e., RGB) obtained from a single optical sensor. Additional spectral bands from a multiple sensor setup introduce challenges for the use of DCNN. We fuse the RGB feature information obtained from a deep learning framework with light detection and ranging (LiDAR) features to obtain semantic labeling. Specifically, we propose a decision-level multisensor fusion technique for semantic labeling of the very-high-resolution optical imagery and LiDAR data. Our approach first obtains initial probabilistic predictions from two different sources: one from a pretrained neural network fine-tuned on a three-band optical image, and another from a probabilistic classifier trained on LiDAR data. These two predictions are then combined as the unary potential using a higher-order conditional random field (CRF) framework, which resolves fusion ambiguities by exploiting the spatial–contextual information. We utilize graph cut to efficiently infer the final semantic labeling for our proposed higher-order CRF framework. Experiments performed on three benchmarking multisensor datasets demonstrate the performance advantages of our proposed method.

Impervious surface (IS) area is an important indicator of ecological environment condition in the basin. We propose an index for IS extraction [i.e., enhanced normalized difference impervious surfaces index (ENDISI)] by integrating the spectrum character of Landsat-operational land imager (OLI) images, and an automatic threshold selection method using the generalized Gaussian model. Dianchi and Erhai Basin are employed as study areas to test the ENDISI method at the plateau basin scale. The results show that: (1) the ENDISI can reduce the impacts of arid land, bare rock, and bare soil on IS extraction effectively; (2) ENDISI had a much higher separability degree between ISs and pervious surfaces compared with normalized difference built-up index, modified normalized difference IS index, and combinational biophysical composition index; and (3) the overall accuracy and kappa coefficient values of IS extraction via automatic threshold selection exceed 93.9% and 82.4%, respectively. Therefore, the ENDISI can serve as an effective index algorithm for rapid and high-precision IS extraction at the plateau basin scale.

Spectral noise causes distorted spectra, shifting the central wavelength and thus reducing the accuracy of surface parameter retrieval. A hybrid method combining mathematical-morphology and wavelet-transform (WB)-based filters was used to remove spectral noise. First, a generalized mathematical-morphology (GM) filter was used to remove large-amplitude noise, and then the processed spectra were smoothed using the WT-based filter to remove small-amplitude noise. The simulated noise spectrum and 76 measured canopy spectra for winter wheat were denoised with three filters: the combination filter (CF), GM, and WT. In the simulated experiments, five evaluation indices were calculated to evaluate the denoising effects. For measured spectra, qualitative analyses were performed based on spectral characteristics. Quantitative evaluations were conducted by deriving various vegetation indices from denoised spectra to retrieve wheat’s biophysical and biochemical parameters. The results indicated that the CF removed both large- and small-amplitude noise efficiently, improving signal-to-noise ratio and peak signal-to-noise ratio of simulated noise spectrum and retrieval accuracy of leaf water content (LWC) significantly. Meanwhile, it better maintained the waveform and smoothness of spectrum, improving the retrieval accuracies of leaf area index and chlorophyll data slightly. The coefficient of determination (R²) of developed model between the modified normalized difference water index and LWC was improved from 0.428 to 0.622 using the CF, 0.555 using the GM, and 0.549 using the WT. The R² and root mean square error between the measured and retrieval LWC were improved from 0.364 and 0.027 to 0.611 and 0.018 using the CF, whereas the corresponding values were 0.504 and 0.022 for the GM, and 0.478 and 0.023 for the WT.

Dimensionality is one of the greatest challenges when deciphering hyperspectral imaging data. Although the multiband nature of the data is beneficial, algorithms are faced with a high computational load and statistical incompatibility due to the insufficient number of training samples. This is a hurdle to downstream applications. The combination of dimensionality and the real-world scenario of mixed pixels makes the identification and classification of imaging data challenging. Here, we address the complications of dimensionality using specific spectral indices from band combinations and spatial indices from texture measures for classification to better identify the classes. We classified spectral and combined spatial–spectral data and calculated measures of accuracy and entropy. A reduction in entropy and an overall accuracy of 80.50% was achieved when using the spectral–spatial input, compared with 65% for the spectral indices alone and 59.50% for the optimally determined principal components.

The modulation frequencies of random stepped frequency radars are randomly distributed over the expected specific bandwidth. The characteristics of random modulation frequencies make it difficult to synthesize the high-resolution range profile fleetly. A zero-padding inverse discrete Fourier transform method is proposed to address the problem. Theoretical analysis shows that the proposed method has the same signal-to-noise ratio improvement performance as the classical methods, but relatively lower computational complexity. The experimental results validate the effectiveness and efficiency of the proposed method.

According to frequency-modulated continuous-wave synthetic-aperture radar, we establish the signal model of a moving target and propose an algorithm for target parameters estimation and imaging. The key procedures of moving target imaging include velocities estimation and range migration correction. Compared with conventional methods, the effect introduced from the along-track velocity is taken into account during the range migration correction to improve the correction accuracy. Moreover, fractional Fourier transform is adopted to estimate the Doppler parameters. To summarize, the proposed method has three advantages. First, by carrying out rough and precise estimation separately, both accurate velocity estimation and better robustness are obtained. Second, this method is available for moving targets with Doppler ambiguity. Third, the third-order phase error derived from moving target is taken into account to improve the azimuth compression result. Finally, a complete flowchart of the proposed method is presented and is verified by both simulation and real data.

Multiscale segmentation is the premise and key step of geographic object-based image analysis (GEOBIA), but scale selection remains a challenge in multiscale segmentation. Over the years, scale selection and evaluation in image segmentation has been extensively explored and many methods have been developed. In these methods, when a scale is chosen or evaluated, all the features are generally extracted from images. In addition, an optimal scale is generally selected based on the pre-estimation of the statistical variance in remote sensing images or determined based on the postsegmentation evaluation of segmented results. In this study, a method was proposed to identify the optimal scale of each segmented object during the segmentation through combining the a priori thematic map knowledge with image features. First, 25 image segmentations were obtained using multiresolution segmentation algorithm of Definiens Professional 9.0 with different scales. A global score (GS) value was assigned to each segmentation based on the calculation results of the weighted variance and global Moran’s I and the single-scale optimal segmentation result was determined according to each GS of 25 segmentation scales. Second, the image feature complexity information and the a priori thematic map complexity information of each segmentation object were extracted to calculate complexity values for each object. Third, the optimal scale of each segmentation object was determined through the iterative calculation with the multithreshold method. Finally, the segmentation results of the proposed method were evaluated. The proposed method was applied to process Gaofen-2 (GF-2), GF-1, Korea Multipurpose Satellite (KOMPSAT-2), IKONOS, QuickBird and WorldView-2 high resolution satellite images to obtain the segmentation results and classification results, compared with results obtained of the optimal single-scale segmentation and the unsupervised evaluation method. The experimental results of GEOBIA showed that the method was helpful for generating the segmentation object with the optimal scale.

An unsupervised hyperspectral hierarchical representation (HHR) framework is proposed by estimating density peak index (DPI). To accelerate our proposed algorithm, we first aggregate adjacent homogenous pixels into extended-pixels as basic initial units. Next, to rank the importance of initial units, both local density and interunit distance of each extended-pixel are taken into account to calculate the DPI. Then, a local–global merging procedure is conducted to construct the hierarchical structure. Selecting the extended-pixels with high DPI values as centers, the related adjacent extended-pixels are merged into local-regions. To generate final HHR, the local-regions are grouped globally based on their updated DPI rank. Considering the representative property of HHR, we also explore a hyperspectral classification strategy. Different from traditional segmentation-based classification methods, our classification scheme only focuses on the classification of the representative subregions, then the labels of other subregions directly refer to the results of representative subregions in the same HHR group. Thus, the computational load for classification can be significantly decreased than for pixel-wise classification. The experiments on three hyperspectral subsets show that the proposed HHR framework can yield remarkable performance in unsupervised clustering, and the classification scheme can obtain equivalent classification accuracy with distinctly limited calculation.

Due to sensor or ground station failure and poor atmospheric conditions, missing information is a common problem that reduces the usage of optical remote sensing data. A large number of algorithms have been developed to reconstruct missing information. Simple methods are compared for filling missing data in the case that one spectral band presents missing information but the other bands are complete. The methods are (1) spatial convolution filtering, (2) unsupervised classification using the complete bands and assignment of the cluster’s average value to missing stripe pixels, (3) global regression models, (4) geographically weighted regression (GWR) models, and (5) a combination of classification and linear models. To evaluate the performance of the different methods, missing line stripes of different sizes are simulated and reconstructed using the five methods. Then, root mean square error (RMSE), correlation, maximum deviance, and bias are computed. To identify the conditions related to the performance of each method, some characteristics of the missing lines, as correlation between spectral bands and spatial autocorrelation, are also calculated. The unsupervised classification based on the larger number of clusters, in particular when combined with linear regression, presents a low RMSE due to its ability to identify the spectral signature of the objects present in the scene. The GWR model performs better than global regression model because it is able to fit the relationship between the missing band and the complete bands locally, which is an important advantage in heterogeneous landscape. Spatial filtering is the most inaccurate method except for one pixel-width missing line.

One practical way to characterize the spatial performance of an in-flight satellite optical imaging system is to determine the modulation transfer function (MTF) from its remotely sensed images on the ground. Currently, there are many MTF measurement techniques designed to provide accurate measurements for high-resolution imaging systems. However, these existing techniques usually rely on the presence and manual identification of a well-separated characteristics target. We introduce an insight to effectively estimate a degradation function based on the MTF by analyzing the nonlocal self-similarity characteristics, namely the structural component, in the observed image. To this end, this paper presents several strategies to realize the aim of this work. First, we propose a segmentation method to select the ideal candidates for MTF estimation. Second, we develop an adaptive structure selection method that removes detrimental structures and selects useful information for point spread function (PSF) estimation. Finally, we put forward a robust estimation method by introducing a spatial prior that is able to simultaneously suppress noises while preserving the sparsity and continuity of the PSF kernels. For the experiments, datasets used in the analyses are synthetically blurred images simulated from level 2A product of IKONOS and real remotely sensed images from level 0 product of RazakSAT. The experimental results demonstrate that the proposed method is practical and effective, with <2.3  %   of relative error at the Nyquist frequency as compared to the well-established edge method. This indicates that the proposed method can warrant a reliable result for on-orbit spatial characterization.

We discuss the problem of extended binary phase-shift keying (EBPSK)–modulated radar waveform design in the presence of colored Gaussian disturbance subject to peak-to-average-power ratio and power constraints. The research work aims at improving the target detection and parameter estimation performance with respect to Doppler shifts. We propose an EBPSK coding design algorithm by optimizing the output signal-to-noise ratio (SNR) in correspondence of the unknown target Doppler frequency. Under the same constraints as in the previous problem, we present EBPSK coding cognitive waveform design algorithm by estimating the target scattering coefficients (TSC) based on maximum output SNR criterion. The simulation experimental results showed that extended target detection performances of optimized EBSPK signals are better than those of the unoptimized EBSPK signal. The designed EBPSK signal attains better TSC estimation performance with relatively lower complexity than the EBPSK signal with arbitrary parameters in the cognitive radar system.

A local-region approach based on bag-of-visual-words model for high-resolution satellite image (HRSI) retrieval and classification is proposed. The proposed method effectively describes HRSI and, hence, considerably reduces the semantic gap. The local representation is achieved by an almost complete description of key points, through the proposed color-texture-structure-spectral-speeded-up robust features (CTSS-SURF) descriptor. The CTSS-SURF can effectively overcome the challenges of HRSI, such as scale, illumination, shift, and rotation variation. The regional representation is achieved by dividing images into several parts and then designing regional feature vectors. For both representations, an improved procedure for dictionary creation is proposed to increase the dictionary discriminative ability and reduce the computational cost. An extensive experimental evaluation on the UC Merced Land Use Dataset has been performed and compared with other feature extraction methods. For retrieval, the proposed method achieves 81.24 in mean average precision value and an accuracy of 98.23% for scene classification. It outperforms many state-of-the-art methods, including convolutional neural networks. The impressive results demonstrate the superiority of the proposed approach for HRSI.

Geological maps are commonly used to investigate the distribution of geological natural resources, such as minerals. However, the existing 1:200,000 geological map created in the 1990s for Guangxi, China, cannot support efficient investigation and interpretation of the geological surface changes. Therefore, we propose the application of remotely sensed multispectral imagery to update the existing 1:200,000 geological map at a scale of 1:100,000. To this end, the analysis of the spectral characteristics of six types of lithologies from the USGS spectral library and from actual measurements by Field Spec^®4, ASD Inc. is conducted first. With the analyzed results of the spectral characteristics, the carbonate rock is separated from the other rocks using band ratios, and then the study area is separated carbonate and noncarbonate areas. In the noncarbonate area, five types of rocks, named, shale, marble, sandstone, granite, and basalt, are classified using supervised classification, in which the training data sets are from the 1:200,000 geographic map. The field verification of the classified results shows that a classification accuracy of 66% is reached, which meets the accuracy requirement for the creation of 1:100,000 geological maps on that basis of the standard formulated by China Geological Map Remote Sensing Interpretation technology. The 1:100,000 geological map created will be delivered to the Guangxi Geological Bureau, China, for applications by the geological and remote sensing communities.

Detailed mapping of urban surfaces is one of the most challenging tasks in remote sensing due to the three-dimensional structure of cities, spatial diversity, and material spectral variability. Satellite urban applications demand better spatial, spectral, and temporal resolution, although there are strict technical constraints among them. Therefore, the development of sophisticated methods that exploit both high spectral and spatial data sources becomes necessary. A hierarchical multiple endmember spectral mixture analysis (MESMA) approach is developed and applied on Sentinel-2 imagery for the detailed quantification of the urban land cover, taking advantage of Worldview-2 high spatial resolution. The case study is the urban and peri-urban area of Heraklion, Greece. The area to point regression kriging (ATPRK) method is applied to downscale Sentinel-2 bands from 10 and 20 m to 2 m (WorldView-2 spatial resolution) and create a spectral library (SL) of urban materials, which contain 180 separate spectra. The urban SL is then used in the developed hierarchical MESMA approach to estimate the abundances of 11 urban land cover classes based on the original Sentinel-2 image. The estimated land cover fractions validate against a very high-resolution (1 m) land cover map of the area. It is proved that the complexity of the urban land cover can be efficiently investigated by the proposed methodology. Error analysis shows good accuracy of the results in all estimated class fractions. Moreover, the good validation results lead to the conclusion that ATPRK fusion algorithm between Sentinel-2 and WorldView-2 bands produced reliable urban material spectra, capable for advanced spectral analysis on Sentinel-2 imagery. The developed methodology is easily transferable to other cities since it is based exclusively on earth observation data and is suitable for multiple urban applications related to urban climate, urban sprawl, and urban regeneration.

We propose an information fusion algorithm—common matrix approach-based edge detection—for edge detection in multispectral images based on a tensor projection approach. In order to merge the edge information situated on different spectral bands, the concept of the common matrix approach is enhanced for each band to project a tensor of gradients onto a derived basis. Once a single-gradient map is acquired from the tensor, a smart thinning and thresholding methodology is utilized at the edge extraction stage in terms of getting rid of spots and discontinuous edge pixels. Upon inspecting the experiments’ results on datasets including the Computer Vision Laboratory at Columbia University, Scene, scanning electron microscopy using the energy-dispersive x-ray microanalysis, and particular hyperspectral remote sensing datasets, it is observed that the proposed method achieves pleasing results in terms of subjective and objective measures in comparison to some recently proposed edge detectors on multispectral and hyperspectral images.

We propose a tensor representation for polarimetric synthetic aperture radar data and extend the usage of tensor learning technique for feature dimension reduction (DR) in image classification. Under the tensor algebra framework, each pixel is modeled as a third-order tensor object by combining multiple polarimetric features and incorporating neighborhood spatial information together. A set of training tensors are determined according to the prior knowledge of the ground truth. Then a tensor learning technique, i.e., multilinear principal component analysis, is applied on the training tensors set to find a tensor subspace that captures most of the variation in the original tensor objects. This process serves as a feature DR step, which is critical for improving the subsequent classification accuracy. Further, the projected tensor samples after DR are fed to the k-nearest neighbor classifier for supervised classification. The performance is verified in both simulated and real datasets. The extracted features are more discriminative in the feature space, and the classification accuracy is significantly improved by at least 10% compared with other existing matrix-based methods.

Urban and peri-urban agriculture (UPA) is of high importance for city dwellers especially in medium-sized cities in sub-Saharan Africa as it fulfills a wide range of multifunctional urban needs. However, the extent of recent agricultural land use dynamics in sub-Saharan African cities under rapid urbanization so far remains unclear. This study assessed the operationality of a rule-based, hierarchical classification of RapidEye satellite imagery to monitor land use and land cover (LULC) dynamics from 2010 to 2017 with a focus on UPA and its patch sizes on the example of the rapidly growing medium-sized city Nakuru, Kenya. The hierarchical system was able to provide a high-resolution, multitemporal, area-wide data set, and proved beneficial through a high flexibility in considering each class separately. Though the approach offered limitations in distinguishing between agricultural crops, the produced data set gave insights about intense LULC dynamics and high land use pressure indicated, in particular, by an increase of sealed areas by 83%, an ongoing expansion of agricultural land and high variations in the detected classes between the years. In the context of data shortage, the findings contribute to the understanding of urbanization patterns in sub-Saharan Africa.

In the process of remote sensing image registration, a robust point-matching algorithm is very important to get better results. However, it may be difficult to find enough correct correspondences when remote sensing images exhibit various noises and local deformations. Although existing methods can remove the outliers, many inliers will be removed at the same time. We propose a point-matching method for remote sensing image registration based on phase congruency normalized cross-correlation and global constraint. First, the classical synthetic aperture radar scale-invariant feature transform algorithm is used to obtain the initial correspondences. Then, the feature distance and phase consistency normalized cross-correlation are combined to remove the outliers in the local neighborhood blocks of the initial correspondences. Finally, the proposed method focuses on increasing the number of inliers and raising the correct ratio by a global constraint. The experimental results compared with several state-of-the-art algorithms show that our method can get more correspondences with higher accuracy.

Hyperspectral remote sensing images (HSIs) are rich in spatial and spectral information, thus they help to enhance the ability to distinguish geographic objects. In recent years, great progress have been made in image classification using deep learning (such as 2D-CNN and 3D-CNN). Compared with traditional machine learning methods, deep learning methods can automatically extract the abstract features from low to high levels and convert the images into more easily recognizable features. Most HSI classification tasks focus on spectral information but often ignore the rich spatial structures in HSIs, leading to a low classification accuracy. Moreover, most supervised learning methods use shallow structures in HSI classifications and hence exhibit weak performance in finding sparse geographic objects. We proposed to use the three-dimensional (3-D) structure to extract spectral–spatial information to build a deep neural network for HSI classifications. Based on DenseNet, the 3D densely connected convolutional network was improved to learn spectral-spatial features of HSIs. The densely connected structure can enhance feature transmission, support feature reuse, improve information flow in the network, and make deeper networks easier to train. The 3D-DenseNet has a deeper structure than 3D-CNN, thus it can learn more robust spectral–spatial features from HSIs. In fact, the deeper network structure has a regularized effect, which can effectively reduce overfitting on small sample datasets. The network uses HSIs instead of feature engineering as input data and is trained in an end-to-end manner. The experimental results of this model on the Indian Pines datasets and the Pavia University datasets show that deeper neural networks further improve the classification of complex objects, especially in the areas where geographic objects are sparse. It effectively improves the classification accuracy of HSIs.

High discriminative feature representation is key in remote sensing scene classification. Existing mid-level feature methods for solving the classification show poor performance. The reason includes two aspects. First, the discrimination power of the feature generated by the feature coding method is limited. Second, semantic information hidden in the scene images are not utilized. These essentially prevent them from achieving better performance. To solve these issues, we propose a hierarchical feature coding model with two stacked feature encoding layers. Specifically, in the first coding layer, semantic information from convolutional layers of deep models and complementary structure and spectral features are extracted and encoded into bag of visual word (BOVW) histogram features. Then in the second layer, Dirichlet-based Gaussians mixture model Fisher kernel is adopted to transform the BOVW histogram features to the more discriminative and effective feature vectors. Thus, through feeding the output of the first layer into the second layer, the complex feature representation is refined. Finally, the concatenated feature vectors are put into support vector machine classifier for classification. Experiments on two public high-resolution remote sensing scene datasets demonstrate that the performance of our hierarchical coding method is comparable to the previous state-of-the-art methods, including most multifeature fusion methods and convolutional neural network-based methods.

In irregular pulse repetition interval (PRI) radar, successive pulses each with different PRIs are used as the transmission waveform. After analyzing the signal model of irregular PRI radar, we propose a coherent integration method based on Radon-iterative adaptive approach (Radon-IAA) to deal with the problems of irregular range cell migration (RCM) and the irregular phase fluctuations among different pulses introduced by the irregular PRI. In our method, the irregular RCM is compensated by searching through the motion parameters, and the irregular phase fluctuations can be resolved by an IAA-based spectral analysis method. The effectiveness of the proposed method is verified by simulation experiments.

A convolutional neural network (CNN) architecture has been proposed for estimating the digital surface model (DSM) from a single airborne or spaceborne image, which is inherently an ambiguous and illposed problem. Deriving the three-dimensional information and reconstructing the geometry of a surface from a monocular image require a deep network that has the ability to extract the local and global characteristics of the surface. In order to address this challenging issue, a deep CNN with residual blocks is employed as a downsampling part of the network, and an upsampling procedure is presented for improving the output accuracy. Moreover, an approach is proposed for connecting the estimated DSM patches and generating a seamless continuous surface. In order to assess the proposed methodologies, scenarios are designed and implemented in various datasets. The final results show that the root mean square error (RMSE) of the proposed approach, when the training and testing images are selected from remote sensing images, is about 3.5 m. In addition, evaluating the capability of the proposed approach for depth estimation using the terrestrial images of outdoor scenes reports about 4 m for RMSE, which is better than the other methods in the literature and demonstrates the promising performance of the proposed approach.

In multitemporal very-high-resolution urban remote sensing images, buildings, especially high-rise buildings, show difference in terms of morphology due to the different view angles. In the coregistered images, the pixels of the same building are not corresponding to each other, which causes false alarm in change detection. Our objective is to find out the matching points located on the roofs of high-rise buildings. When the difference of view angle between the coregistered images is fixed, we discover that there are spatial translation relationships, i.e., local translation transformation and fixed angle offset, between the point matches of high-rise building roofs. Therefore, using these relationships, a method that can sift out point matches of roofs to their correct positions is proposed. The experimental results show that most point matches located on the roof of the same building can be fast and correctly sifted out.

Water body classification is a topic of great interest, especially for the effective management of floods. Synthetic aperture radar (SAR) imaging has demonstrated a great potential for water monitoring, given its capacity to register images independent of weather conditions. Several algorithms for water detection using SAR images are based on optimal thresholding techniques. However, these simple methodologies produce false classification results when small water bodies embedded in mountain ranges are presented in the image. We present an unsupervised and easy-to-implement methodology, based on local Moran index of spatial association in combination with morphological closing operations, for inland water body extraction. According to several experiments, we demonstrate that our method is capable of effectively extracting lakes and rivers located at different land surface reliefs without the requirement of a training step. In addition, comparisons with the state-of-the-art techniques demonstrate the effectiveness of our procedure, performing an overall accuracy of 96.37% and Kappa  =  0.927.

Hyperspectral sparse unmixing aims to identify the best subset of endmembers provided in previously available libraries (which may be very large) and their corresponding proportions (fractional abundance). A sparse unmixing method, local collaborative sparse unmixing (LCSU), including spatial information in standard collaborative formulations, has produced better unmixing results. However, the local window size taken in this method can always have an effect on unmixing result. In detail, LCSU tends to be affected by the distribution of ground objects as well as collaborative sparse unmixing. To address this limitation, we introduce a robust strategy to enforce the local collaborative property in homogeneous areas. The proposed method first divides the denoised image into superpixels according to the quarternion theory, which involves spatial information, and then implements collaborative sparse unmixing in each superpixel. Compared with the most advanced sparse unmixing methods, experiments on two different synthetic datasets and a real hyperspectral dataset effectively verify the advantages of our proposed method.

Intercomparison between the two MODIS instruments is very useful for both the instrument calibration and its uncertainty assessment. Terra and Aqua MODIS have almost identical relative spectral response, spatial resolution, and dynamic range for each band, so the site-dependent effect from spectral mismatch for their comparison is negligible. Major challenges in cross-sensor comparison of instruments on different satellites include differences in observation time and view angle over selected pseudoinvariant sites. The simultaneous nadir overpasses (SNO) between the two satellites are mostly applied for comparison and the scene under SNO varies. However, there is a dearth of SNO between the Terra and Aqua. This work focuses on an intercomparison method for MODIS thermal emissive bands using Himawari-8 Advanced Himawari Imager (AHI) as a reference. Eleven thermal emissive bands on MODIS are at least to some degree spectrally matched to the AHI bands. The sites selected for the comparison are an ocean area around the Himawari-8 suborbital point and the Strzelecki Desert located south of the Himawari-8 suborbital point. The time difference between the measurements from AHI and MODIS is <5  min. The comparison is performed using 2017 collection 6.1 L1B data for MODIS. The MODIS–AHI difference is corrected to remove the view angle dependence. The Terra–Aqua MODIS difference for the selected TEB is up to 0.6 K with the exception of band 30. Band 30 has the largest difference, which is site dependent, most likely due to a crosstalk effect. Over the ocean, the band 30 difference between the two MODIS instruments is around 1.75 K, while over the desert; the difference is around 0.68 K. The MODIS precision is also compared from the Gaussian regression of the double difference. Terra bands 27 to 30 have significant extra noise due to crosstalk effects on these bands. These Terra–Aqua comparison results are used for MODIS calibration assessments and are beneficial for future calibration algorithm improvement. The impact of daytime measurements and the scene dependence are also discussed.

Low-cost receivers have already entered the mass market over the past few years. The high-precision real-time kinematic and precise-point positioning evaluation based on low-cost receivers has been analyzed and verified by many other studies in the past. With the integration of the multiglobal navigation satellite system (GNSS) and advances in technology, low-cost GNSS receivers are expected to be widely used in vehicle navigation. To investigate the kinematic single-point positioning (SPP) performance of low-cost receivers, we first assembled a low-cost receiver that can receive both American global positioning system (GPS) L1 and Chinese BeiDou Navigation Satellite System (BDS) B1 signals simultaneously. Then, a vehicle-positioning experiment is conducted in Fuzhou, China. For comparison purposes, the assembled low-cost receiver and a higher-grade receiver are connected to a survey-grade antenna via a power divider. The positioning accuracy and position dilution of the precision of GPS, BDS, and GPD/BDS SPP with both receivers are compared and analyzed. The results indicate that the low-cost receiver can achieve horizontal precision better than 1 m in the combined GPS/BDS case, when the car is positioned in the urban main road and expressway. This result is competitive and similar to that of the higher-grade receiver. Moreover, although the positioning precision of the low-cost receiver is worse than that of the higher-grade receiver in the GPS-only and BDS-only SPP case, it outperforms the higher-grade receiver in terms of satellite tracking, especially for tracking GPS satellite.

The visible infrared imaging radiometer suite (VIIRS) is a key sensor onboard the newly launched (November 18, 2017) NOAA-20 satellite and the early launched Suomi National Polar-orbiting Partnership (SNPP) satellite. The two VIIRS sensors are nearly identical in design. Its onboard calibration components include a solar diffuser (SD) and a SD stability monitor for the reflective solar bands, a V-grooved blackbody for the thermal emissive bands (TEBs), and a space view port for background subtraction. These onboard calibrators are located at fixed scan angles. The response versus scan angle (RVS) of the VIIRS scan mirror was characterized prelaunch in lab-ambient conditions and is currently used to calibrate the on-orbit response for all scan angles relative to the calibrator’s scan angle. A spacecraft-level pitch maneuver was scheduled during the initial intensive calibration and validation testing for both the NOAA-20 and SNPP. The pitch maneuver provided a rare opportunity for VIIRS to make observations of deep space over the entire range of the scan angle, which can be used to characterize the TEB RVS. We analyze the NOAA-20 pitch maneuver data and assess the derived TEB RVS. A comparison between the RVS determined by the pitch maneuver observations and prelaunch lab measurements is conducted for each band, detector, and mirror side of the half-angle mirror.

Discriminating maize varieties is crucial for crop monitoring, high-throughput phenotyping, and yield forecasting. Crop experts discriminate maize varieties using morphological and biophysical characteristics. However, visual classification suffers from inconsistency, low-throughput, and is only applicable in small scale. A cost-effective and accurate in situ varietal discrimination using multitemporal hyperspectral data and multivariate techniques is explored with threefold objectives, namely, to (1) discriminate maize varieties, (2) define suitable spectral bands, and (3) determine the optimum phenological stage(s) for varietal discrimination. Spectral data in 0.5-nm discrete narrowbands between 400- and 900-nm range are taken from 25 varieties measured using Apogee spectrometer at five phenological stages. Prior to discrimination analysis using partial least squares-discriminant analysis (PLS-DA), three preprocessing transformations are performed: autoscaling, Savitzky–Golay smoothing, and generalized least squares weighting. Ten optimal bands are identified for maize varietal discrimination across the visible and near-infrared section of the wavelength. The significant bands are located in the blue (400 and 455 nm), green (545 nm), red and red-edge (625, 680, 705, and 720 nm), and near-infrared (765, 840, and 895 nm) ranges of the spectrum. Flowering and onset of senescence are identified to be the most ideal phenological stages for accurate maize varietal discrimination. The overall discrimination accuracy improves by 52% and 63% with autoscaling at flowering and senescence, and by 55% and 62% with generalized least squares weighting at flowering and onset of senescence, respectively, compared to no preprocessing transformation. Of the three preprocessing transformations used, autoscaling and generalized least squares weighting are the most effective. Therefore, with appropriate preprocessing transformation, hyperspectral data and PLS-DA are effective in discriminating maize varieties.

With the rapid development of remote-sensing earth observation technology, hyperspectral imagery has shown exponential growth. The quick and accurate retrieval of hyperspectral images has become a practical challenge in applications. Moreover, open network sharing has rendered network information security increasingly important. It is necessary to prevent breach of confidentiality events during retrieval, particularly for hyperspectral images containing crucial information. Therefore, a method for hyperspectral image secure retrieval based on encrypted deep spectral–spatial features is proposed. In principle, our method includes the following steps: (1) Considering the powerful feature learning capability of deep networks, deep spectral–spatial features of hyperspectral image are extracted with a deep convolutional generative adversarial network. (2) For high-dimensional deep features, t-distributed Stochastic neighbor embedding based nonlinear manifold hashing is utilized to reduce the dimensionality of deep spectral–spatial features. (3) To ensure data security during retrieval, deep spectral–spatial features are encrypted with feature randomization encryption. (4) Multi-index hashing is utilized to measure similarities among the deep spatial–spectral features of hyperspectral images. (5) Relevance feedback based on feature reweighting is introduced to further improve retrieval accuracy. Four experiments are conducted to prove the effectiveness of the proposed method based on retrieval and security performance. Our experimental results on two hyperspectral datasets show that our method can effectively protect the security of image content with sufficient image retrieval accuracy.

Combining data of different types and from different sources for the classification of tree species has gained popularity recently, but training models on such datasets often requires more computational demands and does not always result in higher accuracy due to feature redundancy and irrelevance. Thus preprocessing data using dimensionality reduction (DR) methods can be employed to improve the classification accuracy and reduce computations. The objective of this research is to investigate and compare tree species classification performance for different classification algorithms [naive Bayes (NB), logistic regression (LR), random forest (RF), and support vector machine (SVM)], combined with various DR methods (correlation-based feature selection filter, information gain, wrapper methods, and principal component analysis). Two primary datasets are used—QuickBird and LiDAR, as well as derived topography data. When DR is used prior to classification, the NB classifier had a significant improvement in accuracy. SVM and RF had the best classification accuracy without DR. The overall accuracies (OA) of SVM and RF are 88.2% and 87.2% (kappa 0.84 and 0.83), respectively, followed closely by LR (OA: 84.8%, kappa: 0.79) and more distantly by NB (OA: 79%, kappa: 0.72). It is recommended to use SVM and RF without DR or NB with DR for tree species classification.

The Moderate Resolution Imaging Spectroradiometer (MODIS) has provided a large number of high-quality remote sensing data for Earth observation since its launch. Although the reflectance data is calibrated by on-board calibrators, it also needs to be verified by other independent methods in order to ensure the accuracy and reliability of the data. Thus, the Rayleigh scattering method is used to evaluate the MODIS/Aqua reflectance data in our study. In order to obtain high-precision Rayleigh scattering calculation results, the atmospheric and oceanic parameters (AOPs) corresponding to the local time and place, such as wind speed, aerosol optical depth, ozone amount, chlorophyll concentration, and seawater salinity, are put into a radiative transfer model to calculate after a series of screening. The current research selects the pixels with strong Rayleigh scattering characteristics in four days from a global ocean scene. The simulated reflectance is compared with the MYD021KM reflectance product in five visible bands, which presents the total uncertainties as, respectively, 1.39% (412 nm), 0.14% (469 nm), −0.18  %   (488 nm), −0.47  %   (531 nm), −0.41  %   (551 nm). The verification results prove that the MODIS reflectance product remains at a high level of precision without significant deviation after having operated in orbit for 16 years, and the MODIS product has high interband self-consistency. The sensitivity analysis shows that the wind speed and chlorophyll concentration perturbed more to the simulated reflectance than other AOPs of the selected samples. It is believed that the methodology can be applicable to other visible light sensors for validating their reflectance product accurately.

The phenomenon of mixed pixels is serious in hyperspectral images because of its low spatial resolution; so, unmixing mixed pixels is very important for further application of hyperspectral images. The geometrical description in the characteristic space and the sparseness of the abundance matrix are the most important characteristics of the hyperspectral data, as well as the spatial construction information. Traditional algorithms always consider one characteristic and neglect the other. We propose the endmember initials determined by spatial homogeneous area analysis-minimum volume and L_1/2 regularization sparseness-constrained non-negative matrix factorization algorithm, considering all the above characteristics. In estimating the initial endmember matrix, spatial homogeneous area analysis is performed on endmember pixels calculated by vertex component analysis to achieve the goal of accelerating the algorithm’s convergence. At the same time, we join the minimum volume endmember constraint and L_1/2 regularization abundance constraint of stronger sparseness as the final model of our algorithm. The experimental results of both the synthetic and the real data show that, compared with some classical algorithms, the proposed algorithm has a higher accuracy in hyperspectral unmixing.

This erratum corrects an error in one of the author affiliations in the original paper.

This is an erratum to correct an error in a figure in the original paper.