Showing papers in "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing in 2019"

EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification

Abstract: In this paper, we present a patch-based land use and land cover classification approach using Sentinel-2 satellite images. The Sentinel-2 satellite images are openly and freely accessible, and are provided in the earth observation program Copernicus. We present a novel dataset, based on these images that covers 13 spectral bands and is comprised of ten classes with a total of 27 000 labeled and geo-referenced images. Benchmarks are provided for this novel dataset with its spectral bands using state-of-the-art deep convolutional neural networks. An overall classification accuracy of 98.57% was achieved with the proposed novel dataset. The resulting classification system opens a gate toward a number of earth observation applications. We demonstrate how this classification system can be used for detecting land use and land cover changes, and how it can assist in improving geographical maps. The geo-referenced dataset EuroSAT is made publicly available at https://github.com/phelber/eurosat .

Advanced Multi-Sensor Optical Remote Sensing for Urban Land Use and Land Cover Classification: Outcome of the 2018 IEEE GRSS Data Fusion Contest

Abstract: This paper presents the scientific outcomes of the 2018 Data Fusion Contest organized by the Image Analysis and Data Fusion Technical Committee of the IEEE Geoscience and Remote Sensing Society. The 2018 Contest addressed the problem of urban observation and monitoring with advanced multi-source optical remote sensing (multispectral LiDAR, hyperspectral imaging, and very high-resolution imagery). The competition was based on urban land use and land cover classification, aiming to distinguish between very diverse and detailed classes of urban objects, materials, and vegetation. Besides data fusion, it also quantified the respective assets of the novel sensors used to collect the data. Participants proposed elaborate approaches rooted in remote-sensing, and also in machine learning and computer vision, to make the most of the available data. Winning approaches combine convolutional neural networks with subtle earth-observation data scientist expertise.

Hyperspectral Anomaly Detection by Fractional Fourier Entropy

Abstract: Anomaly detection is an important task in hyperspectral remote sensing. Most widely used detectors, such as Reed–Xiaoli (RX), have been developed only using original spectral signatures, which may lack the capability of signal enhancement and noise suppression. In this article, an effective alternative approach, fractional Fourier entropy (FrFE)-based hyperspectral anomaly detection method, is proposed. First, fractional Fourier transform (FrFT) is employed as preprocessing, which obtains features in an intermediate domain between the original reflectance spectrum and its Fourier transform with complementary strengths by space-frequency representations. It is desirable for noise removal so as to enhance the discrimination between anomalies and background. Furthermore, an FrFE-based step is developed to automatically determine an optimal fractional transform order. With a more flexible constraint, i.e., Shannon entropy uncertainty principle on FrFT, the proposed method can significantly distinguish signal from background and noise. Finally, the proposed FrFE-based anomaly detection method is implemented in the optimal fractional domain. Experimental results obtained on real hyperspectral datasets demonstrate that the proposed method is quite competitive.

Pansharpening via Detail Injection Based Convolutional Neural Networks

Abstract: Pansharpening aims to fuse a multispectral (MS) image with an associated panchromatic (PAN) image, producing a composite image with the spectral resolution of the former and the spatial resolution of the latter. Traditional pansharpening methods can be ascribed to a unified detail injection context, which views the injected MS details as the integration of PAN details and bandwise injection gains. In this paper, we design a new detail injection based convolutional neural network (DiCNN) framework for pansharpening with the MS details being directly formulated in end-to-end manners, where the first detail injection based CNN (DiCNN1) mines MS details through the PAN image and the MS image, and the second one (DiCNN2) utilizes only the PAN image. The main advantage of the proposed DiCNNs is that they provide explicit physical interpretations and can achieve fast convergence while achieving high pansharpening quality. Furthermore, the effectiveness of the proposed approaches is also analyzed from a relatively theoretical point of view. Our methods are evaluated via experiments on real MS image datasets, achieving excellent performance when compared to other state-of-the-art methods.

Analysis of CYGNSS Data for Soil Moisture Retrieval

Abstract: Data from the CYGNSS mission, originally conceived to monitor tropical cyclones, are being investigated here for land applications as well. In this paper, a methodology for soil moisture (SM) retrieval from CYGNSS data is presented. The approach derives Level 3 gridded daily SM estimations, over the latitudinal band covered by CYGNSS, at a resolution of 36 km × 36 km, using the CYGNSS reflectivity over land, coupled with ancillary vegetation and roughness information from the SMAP mission. The results are compared globally with SM measurements from SMAP, which are assumed to be ground truth, showing a good agreement, and a global root-mean-square difference of 0.07 cm3/cm3. A more extensive comparison is performed over two test regions—Texas in the United States and New South Wales in Australia—where reference data from SMAP are complemented with measurements from the SMOS mission. The results over both regions are generally consistent with the global results, and a good agreement is observed between CYGNSS and reference SM measurements from SMAP and SMOS. The study demonstrates that SM can be successfully retrieved from the CYGNSS mission on a global scale and using ancillary information about the overlying vegetation and the characteristics of the soil. The results open up further future perspectives for global, high-resolution SM products from spaceborne Global Navigation Satellite System-Reflectometry data.

Incremental Learning for Semantic Segmentation of Large-Scale Remote Sensing Data

Abstract: In spite of remarkable success of the convolutional neural networks on semantic segmentation, they suffer from catastrophic forgetting: a significant performance drop for the already learned classes when new classes are added on the data having no annotations for the old classes. We propose an incremental learning methodology, enabling to learn segmenting new classes without hindering dense labeling abilities for the previous classes, although the entire previous data are not accessible. The key points of the proposed approach are adapting the network to learn new as well as old classes on the new training data, and allowing it to remember the previously learned information for the old classes. For adaptation, we keep a frozen copy of the previously trained network, which is used as a memory for the updated network in the absence of annotations for the former classes. The updated network minimizes a loss function, which balances the discrepancy between outputs for the previous classes from the memory and updated networks, and the misclassification rate between outputs for the new classes from the updated network and the new ground-truth. For remembering, we either regularly feed samples from the stored, little fraction of the previous data or use the memory network, depending on whether the new data are collected from completely different geographic areas or from the same city. Our experimental results prove that it is possible to add new classes to the network, while maintaining its performance for the previous classes, despite the whole previous training data are not available.

Hyperspectral Band Selection via Adaptive Subspace Partition Strategy

Abstract: Band selection is considered as a direct and effective method to reduce redundancy, which is to select some informative and distinctive bands from the original hyperspectral image cube. Recently, many clustering-based band selection methods have been proposed, but most of them only take into account redundancy between bands, neglecting the amount of information in the subset of selected bands. Furthermore, these algorithms never consider the hyperspectral bands as ordered. Based on these two facts, we propose a novel approach for hyperspectral band selection via an adaptive subspace partition strategy (ASPS). The main contributions are as follows: 1) the ASPS is adopted to partition the hyperspectral image cube into multiple subcubes by maximizing the ratio of interclass distance to intraclass distance; 2) unlike previous methods, we estimate the band noise and select the band containing minimum noise (high-quality band) in each subcube to represent the whole subcube; and 3) adaptive subspace partition is viewed as a general framework and thus forms the variant version. Experimental results on three public datasets show that the proposed method achieves satisfactory results in both accuracy and efficiency than some state-of-the-art algorithms.

Assessment of CYGNSS Wind Speed Retrieval Uncertainty

Abstract: Measurements of near surface wind speed made by the Cyclone Global Navigation Satellite System (CYGNSS) constellation of GNSS-R satellites are evaluated and their uncertainty is assessed in two ways. A bottom-up assessment begins with a model for the error in engineering measurements and propagates that error through the wind speed retrieval algorithm analytically. A top-down assessment performs a statistical comparison between CYGNSS measurements and coincident “ground truth” measurements of wind speed. Results of the two approaches are compared. Overall performance, as determined by the top-down method, is decomposed using the bottom-up approach into its contributing sources of error. Overall root mean square (RMS) uncertainty in the CYGNSS retrievals is 1.4 m/s at wind speeds below 20 m/s. At higher wind speeds, an increase in the retrieval error is primarily caused by a decrease in the sensitivity of the ocean scattering cross section to changes in wind speed. In tropical cyclones, retrieval errors are compounded by unaccounted departures from a fully developed sea state. Overall RMS uncertainty in the CYGNSS retrievals is 17% at wind speeds above 20 m/s.

Deep Learning Ensemble for Hyperspectral Image Classification

Abstract: Deep learning models, especially deep convolutional neural networks (CNNs), have been intensively investigated for hyperspectral image (HSI) classification due to their powerful feature extraction ability. In the same manner, ensemble-based learning systems have demonstrated high potential to effectively perform supervised classification. In order to boost the performance of deep learning-based HSI classification, the idea of deep learning ensemble framework is proposed here, which is loosely based on the integration of deep learning model and random subspace-based ensemble learning. Specifically, two deep learning ensemble-based classification methods (i.e., CNN ensemble and deep residual network ensemble) are proposed. CNNs or deep residual networks are used as individual classifiers and random subspaces contribute to diversify the ensemble system in a simple yet effective manner. Moreover, to further improve the classification accuracy, transfer learning is investigated in this study to transfer the learnt weights from one individual classifier to another (i.e., CNNs). This mechanism speeds up the learning stage. Experimental results with widely used hyperspectral datasets indicate that the proposed deep learning ensemble system provides competitive results compared with state-of-the-art methods in terms of classification accuracy. The combination of deep learning and ensemble learning provides a significant potential for reliable HSI classification.

Development of the CYGNSS Geophysical Model Function for Wind Speed

Abstract: Geophysical model functions (GMFs) are developed which map the Level 1 observables made by the Cyclone Global Navigation Satellite System (CYGNSS) radar receivers to ocean surface wind speed. The observables are: 1) the normalized bistatic radar cross section ( σo ) of the ocean surface; and 2) the slope of the leading edge of the radar return pulse scattered by the ocean surface. GMFs are empirically derived from measurements by CYGNSS which are nearly coincident with independent estimates of the 10-m-referenced ocean surface wind speed ( u 10). Two different sources of “ground truth” wind speed are considered: numerical weather prediction model outputs and measurements by the NOAA P-3 hurricane hunter during eyewall penetrations of major hurricanes. The GMFs derived in each case have significant differences that are believed to result from differences in the state of development of the long wave portion of the ocean surface height spectrum that result from characteristic differences in wave age and fetch length near versus far from a hurricane.

The CYGNSS Level 1 Calibration Algorithm and Error Analysis Based on On-Orbit Measurements

Abstract: The calibration algorithm used by the Cyclone Global Navigation Satellite System (CYGNSS) mission to produce version 2.1 of its Level 1 (L1) science data products is described. Changes and improvements have been made to the algorithm, relative to earlier versions, based on the first year of on-orbit result. The L1 calibration consists of two parts: first, the Level 1a (L1a) calibration converts the raw Level 0 delay Doppler maps (DDMs) of processed counts into received power in units of watts. Second, the L1a DDMs are then converted to Level 1b DDMs of bistatic radar cross section values by unwrapping the forward scattering model, which are then normalized by the surface scattering area to arrive an observation of $ \sigma _0$ . An update to the bottom up term-by-term error analysis is also presented, using on-orbit results to better quantify the accuracy of the rolled-up L1 calibration. The error analysis considers uncertainties in all known input calibration parameters. Finally, a method for calibrating the time delay of CYGNSS measurements is presented.

Remote Sensing Image Super-Resolution Using Sparse Representation and Coupled Sparse Autoencoder

Abstract: Remote sensing image super-resolution (SR) refers to a technique improving the spatial resolution, which in turn benefits to the subsequent image interpretation, e.g., target recognition, classification, and change detection. In popular sparse representation-based methods, due to the complex imaging conditions and unknown degradation process, the sparse coefficients of low-resolution (LR) observed images are hardly consistent with the real high-resolution (HR) counterparts, which leads to unsatisfactory SR results. To address this problem, a novel coupled sparse autoencoder (CSAE) is proposed in this paper to effectively learn the mapping relation between the LR and HR images. Specifically, the LR and HR images are first represented by a set of sparse coefficients, and then, a CSAE is established to learn the mapping relation between them. Since the proposed method leverages the feature representation ability of both sparse decomposition and CSAE, the mapping relation between the LR and HR images can be accurately obtained. Experimentally, the proposed method is compared with several state-of-the-art image SR methods on three real-world remote sensing image datasets with different spatial resolutions. The extensive experimental results demonstrate that the proposed method has gained solid improvements in terms of average peak signal-to-noise ratio and structural similarity measurement on all of the three datasets. Moreover, results also show that with larger upscaling factors, the proposed method achieves more prominent performance than the other competitive methods.

A Lightweight and Discriminative Model for Remote Sensing Scene Classification With Multidilation Pooling Module

Abstract: With the growing spatial resolution of satellite images, high spatial resolution (HSR) remote sensing imagery scene classification has become a challenging task due to the highly complex geometrical structures and spatial patterns in HSR imagery. The key issue in scene classification is how to understand the semantic content of the images effectively, and researchers have been looking for ways to improve the process. Convolutional neural networks (CNNs), which have achieved amazing results in natural image classification, were introduced for remote sensing image scene classification. Most of the researches to date have improved the final classification accuracy by merging the features of CNNs. However, the entire models become relatively complex and cannot extract more effective features. To solve this problem, in this paper, we propose a lightweight and effective CNN which is capable of maintaining high accuracy. We use MobileNet V2 as a base network and introduce the dilated convolution and channel attention to extract discriminative features. To improve the performance of the CNN further, we also propose a multidilation pooling module to extract multiscale features. Experiments are performed on six datasets, and the results verify that our method can achieve higher accuracy compared to the current state-of-the-art methods.

Noise-Robust Hyperspectral Image Classification via Multi-Scale Total Variation

Abstract: In this paper, a novel multi-scale total variation method is proposed to extract structural features from hyperspectral images (HSIs), which consists of the following steps. First, the spectral dimension of the HSI is reduced with an averaging-based method. Then, the multi-scale structural features (MSFs), which are insensitive to image noise, are constructed with a relative total variation-based structure extraction technique. Finally, the MSFs are fused together using kernel principal component analysis (KPCA), so as to obtain the KPCA-fused MSFs for classification. Experimental results on three publicly available hyperspectral datasets, including both well-known, long-used data, and a recent dataset obtained from an international contest, demonstrate the competitive performance over several state-of-the-art classification approaches in this field. Moreover, the robustness of the proposed method to the small-sample-size problem and serious image noise is also demonstrated.

Landslide Detection of Hyperspectral Remote Sensing Data Based on Deep Learning With Constrains

Abstract: Detecting and monitoring landslides are hot topics in remote sensing community, particularly with the development of remote sensing technologies and the significant progress of computer vision. To the best of our knowledge, no study focused on deep learning-based methods for landslide detection on hyperspectral images. We proposes a deep learning framework with constraints to detect landslides on hyperspectral image. The framework consists of two steps. First, a deep belief network is employed to extract the spectral–spatial features of a landslide. Second, we insert the high-level features and constraints into a logistic regression classifier for verifying the landslide. Experimental results demonstrated that the framework can achieve higher overall accuracy when compared to traditional hyperspectral image classification methods. The precision of the landslide detection on the whole image, obtained by the proposed method, can reach 97.91%, whereas the precision of the linear support vector machine, spectral information divergence, and spectral angle match are 94.36%, 84.50%, and 86.44%, respectively. Also, this article reveals that the high-level feature extraction system has a significant potential for landslide detection, especially in multi-source remote sensing.

CNN-Based Multilayer Spatial–Spectral Feature Fusion and Sample Augmentation With Local and Nonlocal Constraints for Hyperspectral Image Classification

Abstract: The extraction of joint spatial-spectral features has been proved to improve the classification performance of hyperspectral images (HSIs). Recently, utilizing convolutional neural networks (CNNs) to learn joint spatial-spectral features has become of great interest. However, the existing CNN models ignore complementary spatial-spectral information among the shallow and deep layers. Moreover, insufficient training samples in HSIs afflict these CNN models with overfitting problem. In order to address these problems, a novel CNN method for HSI classification is proposed. It considers multilayer spatial-spectral feature fusion and sample augmentation with local and nonlocal constraints, which is abbreviated as MSLN-CNN. In MSLN-CNN, a triple-architecture CNN is constructed to extract spatial-spectral features by cascading spectral features to dual-scale spatial features from shallow to deep layers. Then, multilayer spatial-spectral features are fused to learn complementary information among the shallow layers with detailed information and the deep layers with semantic information. Finally, the multilayer spatial-spectral feature fusion and classification are integrated into a unified network, and MSLN-CNN can be optimized in the end-to-end way. To alleviate the small sample size problem, the unlabeled samples having high confidences on local spatial constraint and nonlocal spectral constraint are selected and prelabeled. The nonlocal spectral constraint considers the structure information with spectrally similar samples in the nonlocal searching, while the local spatial constraint utilizes the contextual information with spatially adjacent samples. Experimental results on several hyperspectral datasets demonstrate that the proposed method achieves more encouraging classification performance than the current state-of-the-art classification methods, especially with the limited training samples.

Registration of Multimodal Remote Sensing Image Based on Deep Fully Convolutional Neural Network

Abstract: Multimodal image registration is the fundamental technique for scene analysis with series remote sensing images of different spectrum region. Due to the highly nonlinear radiometric relationship, it is quite challenging to find common features between images of different modal types. This paper resorts to the deep neural network, and tries to learn descriptors for multimodal image patch matching, which is the key issue of image registration. A Siamese fully convolutional network is set up and trained with a novel loss function, which adopts the strategy of maximizing the feature distance between positive and hard negative samples. The two branches of the Siamese network are connected by the convolutional operation, resulting in the similarity score between the two input image patches. The similarity score value is used, not only for correspondence point location, but also for outlier identification. A generalized workflow for deep feature based multimodal RS image registration is constructed, including the training data curation, candidate feature point generation, and outlier removal. The proposed network is tested on a variety of optical, near infrared, thermal infrared, SAR, and map images. Experiment results verify the superiority over other state-of-the-art approaches.

A Machine Learning Framework for Drill-Core Mineral Mapping Using Hyperspectral and High-Resolution Mineralogical Data Fusion

Abstract: Mining companies heavily rely on drill-core samples during exploration campaigns as they provide valuable geological information to target important ore accumulations. Traditional core logging techniques are time-consuming and subjective. Hyperspectral (HS) imaging, an emerging technique in the mining industry, is used to complement the analysis by rapidly characterizing large amounts of drill-cores in a nondestructive and noninvasive manner. As the accurate analysis of drill-core HS data is becoming more and more important, we explore the use of machine learning techniques to improve speed and accuracy, and help to discover underlying relations within large datasets. The use of supervised techniques for drill-core HS data represents a challenge since quantitative reference data is frequently not available. Hence, we propose an innovative procedure to fuse high-resolution mineralogical analysis and HS data. We use an automatic high-resolution mineralogical imaging system (i.e., scanning electron microscopy-mineral liberation analysis) for generating training labels. We then resample the MLA image to the resolution of the HS data and adopt a soft labeling strategy for mineral mapping. We define the labels for the classes as mixtures of geological interest and use the classifiers (random forest and support vector machines) to map the entire drill-core. We validate our framework qualitatively and quantitatively. Thus, we demonstrate the ability of the proposed technique to fuse and up-scale high-resolution mineralogical analysis with drill-core HS data.

Hyperspectral Image Classification Method Based on CNN Architecture Embedding With Hashing Semantic Feature

Abstract: Deep convolutional neural networks (CNN) have led to a successful breakthrough for hyperspectral image (HSI) classification. In this paper, a CNN system embedded with an extracted hashing feature is proposed for HSI classification that utilizes the semantic information of the HSI. First, a series of hash functions are constructed to enhance the presentation of the locality and discriminability of classes. Then, the sparse binary hash codes calculated by the discriminative learning algorithm are combined into the original HSI. Next, we design a CNN framework with seven hidden layers to obtain the hierarchical feature maps with both spectral and spatial information for classification. A deconvolution layer aims to improve the robustness of the proposed CNN network and is used to enhance the expression of deep features. The proposed CNN classification architecture achieves powerful distinguishing ability from different classes. The extensive experiments on real hyperspectral images results demonstrate that the proposed CNN network can effectively improve the classification accuracy after the embedding of the extracted semantic features.

A Generative Discriminatory Classified Network for Change Detection in Multispectral Imagery

Abstract: Multispectral image change detection based on deep learning generally needs a large amount of training data. However, it is difficult and expensive to mark a large amount of labeled data. To deal with this problem, we propose a generative discriminatory classified network (GDCN) for multispectral image change detection, in which labeled data, unlabeled data, and new fake data generated by generative adversarial networks are used. The GDCN consists of a discriminatory classified network (DCN) and a generator. The DCN divides the input data into changed class, unchanged class, and extra class, i.e., fake class. The generator recovers the real data from input noises to provide additional training samples so as to boost the performance of the DCN. Finally, the bitemporal multispectral images are input to the DCN to get the final change map. Experimental results on the real multispectral imagery datasets demonstrate that the proposed GDCN trained by unlabeled data and a small amount of labeled data can achieve competitive performance compared with existing methods.

Image Registration With Fourier-Based Image Correlation: A Comprehensive Review of Developments and Applications

Abstract: Fourier-based image correlation is a powerful area-based image registration technique, which involves aligning images based on a translation model or similarity model by means of the image information and operation in the frequency domain. In recent years, Fourier-based image correlation has made significant progress and attracted extensive research interest in a variety of applications, especially in the field of photogrammetry and remote sensing, leading to the development of a number of subpixel methods that have improved the accuracy and robustness. However, to date, a detailed review of the literature related to Fourier-based image correlation is still lacking. In this review, we aim at providing a comprehensive overview of the fundamentals, developments, and applications of image registration with Fourier-based image correlation methods. Specifically, this review introduces the principal laws underlying these methods, presents a survey of the existing subpixel methods calculated both in the spatial domain and in the frequency domain, summarizes the major applications from three aspects, and discusses the challenges and possible directions of future research. This review is expected to be beneficial for researchers working in the relevant fields to obtain an insight into the current state of the art, to develop new variants, to explore potential applications, and to suggest promising future trends of image registration with Fourier-based image correlation.

Densely Based Multi-Scale and Multi-Modal Fully Convolutional Networks for High-Resolution Remote-Sensing Image Semantic Segmentation

Abstract: Automatic and accurate semantic segmentation from high-resolution remote-sensing images plays an important role in the field of aerial images analysis. The task of dense semantic segmentation requires that semantic labels be assigned to each pixel in the image. Recently, convolutional neural networks (CNNs) have proven to be powerful tools for image classification, and they have been adopted in the remote-sensing community. But many limitations still exist when modern CNN architectures are directly applied to remote-sensing images, such as gradient explosion when the depth of the network increases, over-fitting with limited labeled remote-sensing data, and special differences between remote-sensing images and natural images. In this paper, we present a novel architecture that combines the thought of dense connection and fully convolutional networks, referred as DFCN, to automatically provide fine-grained semantic segmentation maps. In addition, we improve DFCN with multi-scale filters to widen the network and to increase the richness and diversity of extracted information, making the network more powerful and expressive than the naive convolution layer. Furthermore, we investigate a multi-modal network that incorporates digital surface models (DSMs) into a DFCN structure, and then we propose dual-path densely convolutional networks where the encoder consists of two paths that, respectively, extract features from spectral data and DSMs data and then fuse them. Finally, through conducting comprehensive experimental evaluations on two remote sensing benchmark datasets, we test our proposed models and compare them with other deep networks. The results demonstrate the effectiveness of proposed approaches; they can achieve competitive performance compared with the current state-of-the-art methods.

HyperPNN: Hyperspectral Pansharpening via Spectrally Predictive Convolutional Neural Networks

Abstract: Hyperspectral (HS) pansharpening intends to synthesize a HS image with a registered panchromatic image, to generate an enhanced image with simultaneous high spectral resolution and high spatial resolution. However, the spectral range gap between the two kinds of images and the need to resolve details for many continuous narrow bands make the technique prone to spectral distortion and spatial blurring. To mitigate the problems, we propose a new HS pansharpening framework via spectrally predictive convolutional neural networks (HyperPNN). In our proposed HyperPNN, spectrally predictive structure is introduced to strengthen the spectral prediction capability of a pansharpening network. Following the concept of the proposed HyperPNN, two specific pansharpening convolutional neural network (CNN) models, i.e., HyperPNN1 and HyperPNN2, are designed. Experimental results from three datasets suggest the excellent performance of our CNN-based HS pansharpening methods.

Pyramid Fully Convolutional Network for Hyperspectral and Multispectral Image Fusion

Abstract: Low spatial resolution hyperspectral (LRHS) and high spatial resolution multispectral (HRMS) image fusion has been recognized as an important technology for enhancing the spatial resolution of LRHS image. Recent advances in convolutional neural network have improved the performance of state-of-the-art fusion methods. However, it is still a challenging problem to effectively explore the spatial information of HRMS image. In this paper, we propose a pyramid fully convolutional network made up of an encoder sub-network and a pyramid fusion sub-network to address this issue. Specifically, the encoder sub-network aims to encode the LRHS image into a latent image. Then, this latent image, together with a HRMS image pyramid input, is used to progressively reconstruct the high spatial resolution hyperspectral image in a global-to-local manner. Furthermore, to sharpen the blurry predictions easily obtained by the standard $l_{2}$ loss, we introduce the gradient difference loss as a regularization term. We evaluate the proposed method on three datasets acquired by three different satellite sensors. Experimental results demonstrate that the proposed method achieves better performance than several state-of-the-art methods.

Design and Performance of a GPS Constellation Power Monitor System for Improved CYGNSS L1B Calibration

Abstract: The Cyclone Global Navigation Satellite System (CYGNSS) uses a bistatic radar configuration with the Global Positioning System (GPS) constellation as the active sources and the CYGNSS satellites as the passive receivers. The GPS effective isotropic radiated power (EIRP), defined as the product of transmit power and antenna gain pattern, is of great importance to the accurate Level 1B calibration of the CYGNSS mission. To address the uncertainties in EIRP, a ground-based GPS constellation power monitor (GCPM) system has been built to accurately and precisely measure the direct GPS signals. A PID thermal controller successfully stabilizes the system temperature over the long term. Radiometric calibration is performed to determine the system dynamic range and to calibrate GCPM gain. Single PRN calibration using a GPS signal simulator is used to compute the scale factor to convert the received counts into power in watts. The GCPM received power is highly repeatable and has been verified with DLR/GSOC's independent measurements. The transmit power (L1 C/A) of the full GPS constellation is estimated using an optimal search algorithm. Updated values for transmit power have been successfully applied to CYGNSS L1B calibration and found to significantly reduce the PRN dependence of CYGNSS L1 and L2 data products.

Crop Type Identification and Mapping Using Machine Learning Algorithms and Sentinel-2 Time Series Data

Abstract: In this paper, the random forests method and the support vector machine in machine learning are explored and compared to the traditional statistical-based maximum likelihood method with 126 features from Sentinel-2A images. The spectral reflectance of 12 bands, 96 texture parameters, 7 vegetation indices, and 11 phenological parameters are successfully extracted from Sentinel-2A images in 2017. The classification result shows that the optimal combination of 13 features yields overall accuracies of traditional classification and machine learning classification of 88.96% and 98%, respectively. Short-wave infrared information shows a significant effect on distinguishing rice, corn, and soybean. The water vapor band plays a significant role in distinguishing between corn and rice. In the multiclassification problem, the machine learning methods have robustness with the identification accuracy of greater than 95% for each crop type, whereas the traditional classification result shows imbalanced accuracies for different crops.

A Novel Deep Structure U-Net for Sea-Land Segmentation in Remote Sensing Images

Abstract: Sea-land segmentation is an important process for many key applications in remote sensing. Proper operative sea-land segmentation for remote sensing images remains a challenging issue due to complex and diverse transition between sea and land. Although several convolutional neural networks (CNNs) have been developed for sea-land segmentation, the performance of these CNNs is far from the expected target. This paper presents a novel deep neural network structure for pixel-wise sea-land segmentation, a residual Dense U-Net (RDU-Net), in complex and high-density remote sensing images. RDU-Net is a combination of both downsampling and upsampling paths to achieve satisfactory results. In each downsampling and upsampling path, in addition to the convolution layers, several densely connected residual network blocks are proposed to systematically aggregate multiscale contextual information. Each dense network block contains multilevel convolution layers, short-range connections, and an identity mapping connection, which facilitates features reuse in the network and makes full use of the hierarchical features from the original images. These proposed blocks have a certain number of connections that are designed with shorter distance backpropagation between the layers and can significantly improve segmentation results while minimizing computational costs. We have performed extensive experiments on two real datasets, Google-Earth and ISPRS, and compared the proposed RDU-Net against several variations of dense networks. The experimental results show that RDU-Net outperforms the other state-of-the-art approaches on the sea-land segmentation tasks.

Object Tracking in Satellite Videos Based on a Multiframe Optical Flow Tracker

Abstract: Object tracking is a hot topic in computer vision. Thanks to the booming of the very high resolution (VHR) remote sensing techniques, it is now possible to track targets of interests in satellite videos. However, since the targets in the satellite videos are usually too small in comparison with the entire image, and too similar with the background, most state-of-the-art algorithms failed to track the target in satellite videos with a satisfactory accuracy. Due to the fact that optical flow shows great potential to detect even the slight movement of the targets, we proposed a multiframe optical flow tracker for object tracking in satellite videos. The Lucas–Kanade optical flow method was fused with the HSV color system and integral image to track the targets in the satellite videos, while multiframe difference method was utilized in the optical flow tracker for a better interpretation. The experiments with five VHR remote sensing satellite video datasets indicate that compared with state-of-the-art object tracking algorithms, the proposed method can track the target more accurately.

Evaluation of Machine Learning Algorithms in Spatial Downscaling of MODIS Land Surface Temperature

Abstract: Land surface temperature (LST) is described as one of the most important environmental parameters of the land surface biophysical process. Commonly, the remote-sensed LST products yield a tradeoff between high temporal and high spatial resolution. Thus, many downscaling algorithms have been proposed to address this issue. Recently, downscaling with machine learning algorithms, including artificial neural networks (ANNs), support vector machine (SVM), and random forest (RF), etc., have gained more recognition with fast operation and high computing precision. This paper intends to make a comparison between machine learning algorithms to downscale the LST product of the moderate-resolution imaging spectroradiometer from 990 to 90 m, and downscaling results would be validated by the resampled LST product of the advanced spaceborne thermal emission and reflection radiometer. The results are further compared with the classical algorithm—thermal sharpening algorithm (TsHARP), using images derived from two representatives kind of areas of Beijing city. The result shows that: 1) all machine learning algorithms produce higher accuracy than TsHARP; 2) the performance of TsHARP on urban area is unsatisfactory than rural because of the weak indication of impervious surface by normalized difference vegetation index, however, machine learning algorithms get the desired results on both two areas, in which ANN and RF models perform well with high accuracy and fast arithmetic, SVM also gets a good result but there is a smoothing effect on land surface; and 3) additionally, machine learning algorithms are promising to achieve a universal framework which can downscale LST for any area within the training data from long spatiotemporal sequences.

Hyperspectral Image Denoising via Subspace-Based Nonlocal Low-Rank and Sparse Factorization

Abstract: Hyperspectral images (HSIs) are unavoidably contaminated by different types of noise during data acquisition and transmission, e.g., Gaussian noise, impulse noise, stripes, and deadlines. A variety of mixed noise reduction approaches are developed for HSI, in which the subspace-based methods have achieved comparable performance. In this paper, a novel subspace-based nonlocal low-rank and sparse factorization (SNLRSF) method is proposed to remove the mixture of several types of noise. The SNLRSF method explores spectral low rank based on the fact that spectral signatures of pixels lie in a low-dimensional subspace and employs the nonlocal low-rank factorization to take the spatial nonlocal self-similarity into consideration. At the same time, the successive singular value decomposition (SVD) low-rank factorization algorithm is used to estimate three-dimensional (3-D) tensor generated by nonlocal similar 3-D patches. Moreover, the well-known augmented Lagrangian method is adopted to solve final denoising model efficiently. The experimental results over simulated and real datasets demonstrate that the proposed approach outperforms the related state-of-the-art methods in terms of visual quality and quantitative evaluation.