Optical Remote Sensing Research Articles

LMSD-Net: A Lightweight and High-Performance Ship Detection Network for Optical Remote Sensing Images

Ship detection technology has achieved significant progress recently. However, for practical applications, lightweight ship detection still remains a very challenging problem since small ships have small relative scales in wide images and are easily missed in the background. To promote the research and application of small-ship detection, we propose a new remote sensing image dataset (VRS-SD v2) and provide a fog simulation method that reflects the actual background in remote sensing ship detection. The experiment results show that the proposed fog simulation is beneficial in improving the robustness of the model for extreme weather. Further, we propose a lightweight detector (LMSD-Net) for ship detection. Ablation experiments indicate the improved ELA-C3 module can efficiently extract features and improve the detection accuracy, and the proposed WGC-PANet can reduce the model parameters and computation complexity to ensure a lightweight nature. In addition, we add a Contextual Transformer (CoT) block to improve the localization accuracy and propose an improved localization loss specialized for tiny-ship prediction. Finally, the overall performance experiments demonstrate that LMSD-Net is competitive in lightweight ship detection among the SOTA models. The overall performance achieves 81.3% in AP@50 and could meet the lightweight and real-time detection requirements.

Optical Remote Sensing Ship Recognition and Classification Based on Improved YOLOv5

Due to the special characteristics of the shooting distance and angle of remote sensing satellites, the pixel area of ship targets is small, and the feature expression is insufficient, which leads to unsatisfactory ship detection performance and even situations such as missed and false detection. To solve these problems, this paper proposes an improved-YOLOv5 algorithm mainly including: (1) Add the Convolutional Block Attention Module (CBAM) into the Backbone to enhance the extraction of target-adaptive optimal features; (2) Introduce a cross-layer connection channel and lightweight GSConv structures into the Neck to achieve higher-level multi-scale feature fusion and reduce the number of model parameters; (3) Use the Wise-IoU loss function to calculate the localization loss in the Output, and assign reasonable gradient gains to cope with differences in image quality. In addition, during the preprocessing stage of experimental data, a median+bilateral filter method was used to reduce interference from ripples and waves and highlight the information of ship features. The experimental results show that Improved-YOLOv5 has a significant improvement in recognition accuracy compared to various mainstream target detection algorithms; compared to the original YOLOv5s, the mean Average Precision (mAP) improved by 3.2% and the Frames Per Second (FPN) accelerated by 8.7%.

Integration of microwave satellite soil moisture products in the contextual surface temperature-vegetation index models for spatially continuous evapotranspiration estimation

The contextual surface temperature-vegetation index (TVX) models have been widely used for the retrieval of soil moisture (SM) and evapotranspiration (ET). One of the key premises for their application is to determine quantitatively the theoretical boundaries of this contextual TVX space. Usually these theoretical boundaries are determined based on land surface energy balance principle. Although sound in physical mechanism, the complex parameterization involved has hampered its remote sensing (RS) applications. Our recent studies show that ground-based SM observations can be used for a quick and continuous retrieval of these theoretical boundaries. The main disadvantage lies in its reliance on in-situ SM observations. Fortunately the development of microwave RS technology has made it possible to monitor SM at both regional and global scale. Under this background, a practical framework based on the contextual TVX models was proposed in this paper for continuous ET estimation through the combination of microwave-based SM products and MODIS (Moderate Resolution Imaging Spectroradiometer) products. Daily actual ET over the Southern Great Plains of the United States of America was calculated as the product of net radiation and evaporative fraction (EF). The focus of this framework was to investigate how optical and microwave RS can be combined to achieve a quick and continuous retrieval of EF independently of ground observations. Specifically, EF was retrieved from the contextual TVX models, the theoretical boundaries of which were calibrated at the annual scale based on an optimization scheme. Assuming the seasonal variation of the theoretical dry edge follows a cosine function of solar zenith, the essence of this optimization scheme is to search for the optimal amplitude parameter that maximizes the correlation between microwave-based SM observations and soil moisture index retrieved from the contextual TVX models. Two microwave-based SM products including SMAP (Soil Moisture Active Passive) and SMOS (Soil Moisture and Ocean Salinity) were adopted to demonstrate this optimization scheme. The correlation coefficient, root mean square error and bias of EF estimates calibrated by SMAP were 0.700, 0.146 and 0.047, and those achieved by SMOS were 0.663, 0.174 and 0.168, respectively. The optimization scheme has not only made it possible to conduct a continuous monitoring of ET based entirely on RS observations but also achieved the downscaling of microwave satellite SM products. Thus the combination of optical and microwave RS holds great potential for coupled and continuous estimation of SM and ET over large scale.

Extracting water body data based on SDWI and threshold segmentation: A case study in permafrost area surrounding Salt Lake in Hoh Xil, Qinghai-Tibet Plateau, China

There are a large number of lakes, rivers, and other natural water bodies distributed in the permafrost area of the Qinghai-Tibet Plateau (QTP). The changes in water bodies will affect the distribution of water resources in surrounding areas and downstream areas, resulting in environmental impact and bringing potential flood disasters, which will induce more serious issues and problems in alpine and high-altitude areas with a fragile habitat (such as the QTP in China). Generally, effective, reasonable, and scientific monitoring of large-scale water bodies can not only document the changes in water bodies intuitively, but also provide important theoretical reference for subsequent environmental impact prediction, and disaster prevention and mitigation in due course of time. The large-scale water extraction technology derived from the optical remote sensing (RS) image is seriously affected by clouds, bringing about large differences among the extracted water result products. Synthetic aperture radar (SAR) RS technology has the unique advantage characteristics of all-weather, all-day, strong penetration, and not being affected by clouds, which is hopeful in extracting water body data, especially for days with cloudy weather. The data extraction of large-scale water bodies based on SAR images can effectively avoid the errors caused by clouds that become prevalent at present. In this paper, the Hoh Xil Salt Lake on the QTP and its surrounding five lakes are taken as the research objects. The 2-scene Sentinel-1 SAR image data covering the whole area on 22 August 2022 was used to verify the feasibility of extracting water body data in permafrost zones. Furthermore, on 22 August 2022, the wealth here was cloudy, which made the optical RS images, e.g., Sentinel-2 images full of clouds. The results show that: using the Sentinel-1 image and threshold segmentation method to extract water body data is efficient and effective with excellent results in permafrost areas. Concretely, the Sentinel-1 dual-polarized water index (SDWI), calculated by combining dual vertical–vertical (VV) polarized and vertical–horizontal (VH) polarized data is a useful index for water extraction and the result is better than each of the VV or VH polarized images.

High Accuracy Solar Diffuser BRDF Measurement for On-Board Calibration in the Solar Reflective Band

In the solar reflective band, an on-board calibration method based on a solar diffuser (SD) can realize full aperture, full field of view, and end-to-end absolute radiometric calibration of optical remote sensors. The SD’s bidirectional reflectance distribution function (BRDF) is a key parameter that affects the accuracy of the on-board calibration. High-accuracy measurement of the SD BRDF is required in the laboratory before launch. Due to the uncertainty of the goniometer system, polarization effects, and other factors, the measurement uncertainty of the SD BRDF at large incident angles is much higher than that at a 0° incident zenith angle and 45° reflection zenith angle. In this paper, an absolute BRDF measurement facility is reported. The goniometric system consists of a high-brightness integrating sphere as a radiation source, a six-axis robot arm, and a large rotation stage. The measurement wavelength range was from 350 nm to 2400 nm. An improved data processing method based on the reciprocity theorem was proposed to reduce the measurement uncertainty of the SD BRDF at large incident angles. At an incident zenith angle of 75°, the improved data processing method reduced the measurement uncertainty of the SD BRDF by 52% at 410 nm to 480 nm, by 70% at 480 nm to 1000 nm, and by 20% at other bands compared to the absolute measurement method. The influence of the radiation source, goniometer system, detection system, and other factors on the measurement uncertainty are analyzed in this paper. The results show that the measurement uncertainty (coverage factor k = 2) of the SD BRDF was better than 1.04% at 350 nm to 410 nm, 0.60% at 410 nm to 480 nm, 0.43% at 480 nm to 1000 nm, and 0.86% at 1000 nm to 2400 nm.

Alignment Integration Network for Salient Object Detection and Its Application for Optical Remote Sensing Images.

Salient object detection has made substantial progress due to the exploitation of multi-level convolutional features. The key point is how to combine these convolutional features effectively and efficiently. Due to the step by step down-sampling operations in almost all CNNs, multi-level features usually have different scales. Methods based on fully convolutional networks directly apply bilinear up-sampling to low-resolution deep features and then combine them with high-resolution shallow features by addition or concatenation, which neglects the compatibility of features, resulting in misalignment problems. In this paper, to solve the problem, we propose an alignment integration network (ALNet), which aligns adjacent level features progressively to generate powerful combinations. To capture long-range dependencies for high-level integrated features as well as maintain high computational efficiency, a strip attention module (SAM) is introduced into the alignment integration procedures. Benefiting from SAM, multi-level semantics can be selectively propagated to predict precise salient objects. Furthermore, although integrating multi-level convolutional features can alleviate the blur boundary problem to a certain extent, it is still unsatisfactory for the restoration of a real object boundary. Therefore, we design a simple but effective boundary enhancement module (BEM) to guide the network focus on boundaries and other error-prone parts. Based on BEM, an attention weighted loss is proposed to boost the network to generate sharper object boundaries. Experimental results on five benchmark datasets demonstrate that the proposed method can achieve state-of-the-art performance on salient object detection. Moreover, we extend the experiments on the remote sensing datasets, and the results further prove the universality and scalability of ALNet.

Combined PS-InSAR Technology and High-Resolution Optical Remote Sensing for Identifying Illegal Underground Mining in the Suburb of Yangquan City, Shanxi Province, China

Illegal mining is one of the biggest problems in many coal mines. With the rapid development of the economy, driven by huge economic benefits, some outlaws illegally exploit mineral resources without a mining license, which is destructive and a potential safety hazard. In order to avoid inspection by law enforcement officials, some outlaws, regardless of the cost or risk, privately and surreptitiously excavate coal mines in self-built houses. The coal resources they excavate are shallow coal resources. Because surface buildings can maintain strong and stable radar scattering characteristics over a long time series, in this study, we combined PS-InSAR technology and high-resolution optical remote sensing to extract the subsidence information of surface buildings corresponding to PS point sets and analyzed their spatiotemporal characteristics. Finally, we developed a fast and accurate method for detecting suspected illegal mining sites from building subsidence information over a larger area. We also carried out a case study using Shandi Village, a suburb of Yangquan City, Shanxi Province, China, as our research object. QuickBird-2, WorldView-2 data, and 20 PALSAR scenes were selected for the experimental research, and two illegal mining sites were detected from 29 December 2006 to 9 January 2011. By comparing our results with previous investigation data, it was found that the accuracy rate reached 40% in local areas, and the detection rate reached 66.67%. In addition, the mining periods were basically consistent. This research shows that our method is feasible and has certain engineering applicability and practical value.

Wetland Mapping in Great Lakes Using Sentinel-1/2 Time-Series Imagery and DEM Data in Google Earth Engine

The Great Lakes (GL) wetlands support a variety of rare and endangered animal and plant species. Thus, wetlands in this region should be mapped and monitored using advanced and reliable techniques. In this study, a wetland map of the GL was produced using Sentinel-1/2 datasets within the Google Earth Engine (GEE) cloud computing platform. To this end, an object-based supervised machine learning (ML) classification workflow is proposed. The proposed method contains two main classification steps. In the first step, several non-wetland classes (e.g., Barren, Cropland, and Open Water), which are more distinguishable using radar and optical Remote Sensing (RS) observations, were identified and masked using a trained Random Forest (RF) model. In the second step, wetland classes, including Fen, Bog, Swamp, and Marsh, along with two non-wetland classes of Forest and Grassland/Shrubland were identified. Using the proposed method, the GL were classified with an overall accuracy of 93.6% and a Kappa coefficient of 0.90. Additionally, the results showed that the proposed method was able to classify the wetland classes with an overall accuracy of 87% and a Kappa coefficient of 0.91. Non-wetland classes were also identified more accurately than wetlands (overall accuracy = 96.62% and Kappa coefficient = 0.95).

Lidar measurements of wake around a bridge deck

Remote wind sensing technologies allow for measurements in a spatial domain beyond the one accessible by anemometers fixed to a mast. Remote optical wind sensors installed at an existing bridge site can both provide new information on the wind flow upstream of the bridge and capture the disturbed flow around the structure. These possibilities are explored in the presented measurement campaign with three synchronised continuous-wave Doppler lidars at the Lysefjord bridge in Norway. In particular, the interaction between the oncoming wind flow and the bridge deck is studied in terms of the mean velocity deficit wake profile as well as the spectral turbulence characteristics. The lidar measurements are validated by comparing the lidar observations at locations undisturbed by the structure with corresponding sonic anemometer data recorded 6 m and 10 m above the deck. The consistency between the two sets of data is tested in terms of the turbulence time series, their spectral content and the statistical moments. Within the wake of the bridge, significant values of vertical velocity and inclination angles, as well as turbulence intensities are seen. Moreover, the wake region can be distinguished by its singular spectral content, cross-correlation coefficient, and coherence. The results show that the tailored configuration of the lidars succeeded in providing information on 3D turbulence around the bridge, thereby capturing the bridge aerodynamic characteristics in full-scale.

Pixel-Wise Attention Residual Network for Super-Resolution of Optical Remote Sensing Images

The deep-learning-based image super-resolution opens a new direction for the remote sensing field to reconstruct further information and details from captured images. However, most current SR works try to improve the performance by increasing the complexity of the model, which results in significant computational costs and memory consumption. In this paper, we propose a lightweight model named pixel-wise attention residual network for optical remote sensor images, which can effectively solve the super-resolution task of multi-satellite images. The proposed method consists of three modules: the feature extraction module, feature fusion module, and feature mapping module. First, the feature extraction module is responsible for extracting the deep features from the input spatial bands with different spatial resolutions. Second, the feature fusion module with the pixel-wise attention mechanism generates weight coefficients for each pixel on the feature map and fully fuses the deep feature information. Third, the feature mapping module is aimed to maintain the fidelity of the spectrum by adding the fused residual feature map directly to the up-sampled low-resolution images. Compared with existing deep-learning-based methods, the major advantage of our method is that for the first time, the pixel-wise attention mechanism is incorporated in the task of super-resolution fusion of remote sensing images, which effectively improved the performance of the fusion network. The accuracy assessment results show that our method achieved superior performance of the root mean square error, signal-to–reconstruction ratio error, universal image quality index, and peak signal noise ratio compared to competing approaches. The improvements in the signal-to-reconstruction ratio error and peak signal noise ratio are significant, with a respective increase of 0.15 and 0.629 dB for Sentinel-2 data, and 0.196 and 1 dB for Landsat data.

Optical Remote Sensing in Provisioning of Ecosystem-Functions Analysis—Review

Keeping natural ecosystems and their functions in the proper condition is necessary. One of the best contactless monitoring methods is remote sensing, especially optical remote sensing, which is used for vegetation applications. In addition to satellite data, data from ground sensors are necessary for validation or training in ecosystem-function quantification. This article focuses on the ecosystem functions associated with aboveground-biomass production and storage. The study contains an overview of the remote-sensing methods used for ecosystem-function monitoring, especially methods for detecting primary variables linked to ecosystem functions. The related studies are summarized in multiple tables. Most studies use freely available Sentinel-2 or Landsat imagery, with Sentinel-2 mostly producing better results at larger scales and in areas with vegetation. The spatial resolution is a key factor that plays a significant role in the accuracy with which ecosystem functions are quantified. However, factors such as spectral bands, algorithm selection, and validation data are also important. In general, optical data are usable even without supplementary data.

Multi-Temporal SamplePair Generation for Building Change Detection Promotion in Optical Remote Sensing Domain Based on Generative Adversarial Network

Change detection is a critical task in remote sensing Earth observation for identifying changes in the Earth’s surface in multi-temporal image pairs. However, due to the time-consuming nature of image collection, labor-intensive pixel-level labeling with the rare occurrence of building changes, and the limitation of the observation location, it is difficult to build a large, class-balanced, and diverse building change detection dataset, which can result in insufficient changed sample pairs for training change detection models, thus degrading their performance. Thus, in this article, given that data scarcity and the class-imbalance issue lead to the insufficient training of building change detection models, a novel multi-temporal sample pair generation method, namely, Image-level Sample Pair Generation (ISPG), is proposed to improve the change detection performance through dataset expansion, which can generate more valid multi-temporal sample pairs to overcome the limitation of the small amount of change information and class-imbalance issue in existing datasets. To achieve this, a Label Translation GAN (LT-GAN) was designed to generate complete remote sensing images with diverse building changes and background pseudo-changes without any of the complex blending steps used in previous works. To obtain more detailed features in image pair generation for building change detection, especially the surrounding context of the buildings, we designed multi-scale adversarial loss (MAL) and feature matching loss (FML) to supervise and improve the quality of the generated bitemporal remote sensing image pairs. On the other hand, we also consider that the distribution of generated buildings should follow the pattern of human-built structures. The proposed approach was evaluated on two building change detection datasets (LEVIR-CD and WHU-CD), and the results proved that the proposed method can achieve state-of-the-art (SOTA) performance, even if using plain models for change detection. In addition, the proposed approach to change detection image pair generation is a plug-and-play solution that can be used to improve the performance of any change detection model.

Reduction and reconstruction strategy of active thermal control system based on unsupervised learning and thermo-optics for spaceborne high-resolution remote sensor

The development of small Earth observation satellites with high performance optical remote sensors has received much attention in recent years due to its responsiveness and immediacy. As one of the most resource-intensive subsystems of spacecrafts, the active thermal control system (ATCS) should be optimized; however, manual thermal design does not guarantee an optimal solution and requires months. For this purpose, we presented a ATCS’s reduction and reconstruction strategy based on an unsupervised learning framework, in which factor loadings are applied to combine kernel-based principal component analysis and Gaussian mixture model clustering. This strategy was implemented on a spaceborne high-resolution remote sensor (HRS) with 0.5 m ground sample distance. As a result, with the cumulative variance contribution rate > 90 % and 10 Gaussian distributions, the reconstructed ATCS of HRS had a 78.6 % reduction rate and took only 3 h, which suggests the high efficiency and low computational cost of the proposed strategy. Finally, the performed orbital thermal and thermo-optical integrated analyses indicated that temperature fluctuations were within ± 0.02–0.6 °C and optical surface errors were within the tolerance. The average modulation transfer function in the tangential and sagittal directions decreased by only 2 % and 2.5 %, respectively, and the radius of the spot root mean square was considerably smaller than the size of one pixel (8.75 μm). These results prove the feasibility and validity of the proposed strategy. This study solved the problem of identifying the optimal ATCS for optical remote sensors and significantly reducing the ATCS’s resource occupancy, which provided a reliable research basis for the development of high-performance small Earth observation satellites. Additionally, the proposed strategy is versatile and can be used in thermal engineering applications of various complex spacecrafts and payloads.

Automated object recognition in high-resolution optical remote sensing imagery.

This paper reports the background and results of the Automated Object Recognition in Optical Remote Sensing Imagery, which is one of the tracks in 2022 International Algorithm Case Competition, as well as summarize the challenges, champion solutions, and future directions.

RAIH-Det: An End-to-End Rotated Aircraft and Aircraft Head Detector Based on ConvNeXt and Cyclical Focal Loss in Optical Remote Sensing Images

In airport ground-traffic surveillance systems, the detection of an aircraft and its head (AIH) is an important task in aircraft trajectory judgment. However, accurately detecting an AIH in high-resolution optical remote sensing images is a challenging task due to the difficulty in effectively modeling the features of aircraft objects, such as changes in appearance, large-scale differences, complex compositions, and cluttered background. In this paper, we propose an end-to-end rotated aircraft and aircraft head detector (RAIH-Det) based on ConvNeXt-T (Tiny) and cyclical local loss. Firstly, a new U-shaped network based on ConvNeXt-T with the same performance as the Local Vision Transformer (e.g., Swin Transformer) is presented to assess the relationships among aircraft in the spatial domain. Then, in order to enhance the sharing of more mutual information, the extended BBAVectors with six vectors captures the oriented bounding box (OBB) of the aircraft in any direction, which can assist in head keypoint detection by exploiting the relationship between the local and overall structural information of aircraft. Simultaneously, variant cyclical focal loss is adopted to regress the heatmap location of keypoints on the aircraft head to focus on more reliable samples. Furthermore, to perform a study on AIH detection and simplify aircraft head detection, the OBBs of the “plane” category in the DOTA-v1.5 dataset and the corresponding head keypoints annotated by our volunteers were integrated into a new dataset called DOTA-Plane. Compared with other state-of-the-art rotated object and keypoint detectors, RAIH-Det, as evaluated on DOTA-Plane, offered superior performance.

Mapping Pluvial Flood-Induced Damages with Multi-Sensor Optical Remote Sensing: A Transferable Approach

Pluvial floods caused by extreme overland flow inland account for half of all flood damage claims each year along with fluvial floods. In order to increase confidence in pluvial flood susceptibility mapping, overland flow models need to be intensively evaluated using observations from past events. However, most remote-sensing-based flood detection techniques only focus on the identification of degradations and/or water pixels in the close vicinity of overflowing streams after heavy rainfall. Many occurrences of pluvial-flood-induced damages such as soil erosion, gullies, landslides and mudflows located further away from the stream are thus often unrevealed. To fill this gap, a transferable remote sensing fusion method called FuSVIPR, for Fusion of Sentinel-2 & Very high resolution Imagery for Pluvial Runoff, is developed to produce damage-detection maps. Based on very high spatial resolution optical imagery (from Pléiades satellites or airborne sensors) combined with 10 m change images from Sentinel-2 satellites, the Random Forest and U-net machine/deep learning techniques are separately trained and compared to locate pluvial flood footprints on the ground at 0.5 m spatial resolution following heavy weather events. In this work, three flash flood events in the Aude and Alpes-Maritimes departments in the South of France are investigated, covering over more than 160 km2 of rural and periurban areas between 2018 and 2020. Pluvial-flood-detection accuracies hover around 75% (with a minimum area detection ratio for annotated ground truths of 25%), and false-positive rates mostly below 2% are achieved on all three distinct events using a cross-site validation framework. FuSVIPR is then further evaluated on the latest devastating flash floods of April 2022 in the Durban area (South Africa), without additional training. Very good agreement with the impact maps produced in the context of the International Charter “Space and Major Disasters” are reached with similar performance figures. These results emphasize the high generalization capability of this method to locate pluvial floods at any time of the year and over diverse regions worldwide using a very high spatial resolution visible product and two Sentinel-2 images. The resulting impact maps have high potential for helping thorough evaluation and improvement of surface water inundation models and boosting extreme precipitation downscaling at a very high spatial resolution.

An Optical Remote Sensing Image Matching Method Based on the Simple and Stable Feature Database

Satellite remote sensing has entered the era of big data due to the increase in the number of remote sensing satellites and imaging modes. This presents significant challenges for the processing of remote sensing systems and will result in extremely high real-time data processing requirements. The effective and reliable geometric positioning of remote sensing images is the foundation of remote sensing applications. In this paper, we propose an optical remote sensing image matching method based on a simple stable feature database. This method entails building the stable feature database, extracting local invariant features that are comparatively stable from remote sensing images using an iterative matching strategy, and storing useful information about the features. Without reference images, the feature database-based matching approach potentially saves storage space for reference data while increasing image processing speed. To evaluate the performance of the feature database matching method, we train the feature database with various local invariant feature algorithms on different time phases of Gaofen-2 (GF-2) images. Furthermore, we carried out matching comparison experiments with various satellite images to confirm the viability and stability of the feature database-based matching method. In comparison with direct matching using the classical feature algorithm, the feature database-based matching method in this paper can essentially improve the correct rate of feature point matching by more than 30% and reduce the matching time by more than 40%. This method improves the accuracy and timeliness of image matching, potentially solves the problem of large storage space occupied by the reference data, and has great potential for fast matching of optical remote sensing images.

Overview of the Application of Remote Sensing in Effective Monitoring of Water Quality Parameters

This study provides an overview of the techniques, shortcomings, and strengths of remote sensing (RS) applications in the effective retrieval and monitoring of water quality parameters (WQPs) such as chlorophyll-a concentration, turbidity, total suspended solids, colored dissolved organic matter, total dissolved solids among others. To be effectively retrieved by RS, these WQPs are categorized as optically active or inactive based on their influence on the optical characteristics measured by RS sensors. RS applications offer the opportunity for decisionmakers to quantify and monitor WQPs on a spatiotemporal scale effectively. The use of RS for water quality monitoring has been explored in many studies using empirical, analytical, semi-empirical, and machine-learning algorithms. RS spectral signatures have been applied for the estimation of WQPs using two categories of RS, namely, microwave and optical sensors. Optical RS, which has been heavily applied in the estimation of WQPs, is further grouped as spaceborne and airborne sensors based on the platform they are on board. The choice of a particular sensor to be used in any RS application depends on various factors including cost, and spatial, spectral, and temporal resolutions of the images. Some of the known satellite sensors used in the literature and reviewed in this paper include the Multispectral Instrument aboard Sentinel-2A/B, Moderate Resolution Imaging Spectroradiometer, Landsat Thematic Mapper, Enhanced Thematic Mapper, and Operational Land Imager.

Polarimetric Imaging via Deep Learning: A Review

Polarization can provide information largely uncorrelated with the spectrum and intensity. Therefore, polarimetric imaging (PI) techniques have significant advantages in many fields, e.g., ocean observation, remote sensing (RS), biomedical diagnosis, and autonomous vehicles. Recently, with the increasing amount of data and the rapid development of physical models, deep learning (DL) and its related technique have become an irreplaceable solution for solving various tasks and breaking the limitations of traditional methods. PI and DL have been combined successfully to provide brand-new solutions to many practical applications. This review briefly introduces PI and DL’s most relevant concepts and models. It then shows how DL has been applied for PI tasks, including image restoration, object detection, image fusion, scene classification, and resolution improvement. The review covers the state-of-the-art works combining PI with DL algorithms and recommends some potential future research directions. We hope that the present work will be helpful for researchers in the fields of both optical imaging and RS, and that it will stimulate more ideas in this exciting research field.

A Multi-Branch Training and Parameter-Reconstructed Neural Network for Assessment of Signal-to-Noise Ratio of Optical Remote Sensor on Orbit

Signal-to-Noise Ratio (SNR) is the benchmark to evaluate the quality of optical remote sensors. For SNR estimation, most of the traditional methods have complicated processes, low efficiency, and general accuracy. In particular, they are not suitable for the distributed computation on intelligent satellites. Therefore, an intelligent SNR estimation algorithm with strong computing power and more accuracy is urgently needed. Considering the simplicity of distributed deployment and the lightweight goal, our first proposition is to design a convolutional neural network (CNN) similar to VGG (proposed by Visual Geometry Group) to estimate SNR for optical remote sensors. In addition, considering the advantages of multi-branch structures, the second proposition is to train the CNN in a novel method of multi-branch training and parameter-reconstructed inference. In this study, simulated and real remote sensing images with different ground features are utilized to validate the effectiveness of our model and the novel training method. The experimental results show that the novel training method enhances the fitting ability of the network, and the proposed CNN trained in this method has high accuracy and reliable SNR estimation, which achieves a 3.9% RMSE for noise-level-known simulated images. When compared to the accuracy of the reference methods, such as the traditional and typical SNR methods and the denoising convolutional neural network (DnCNN), the performance of the proposed CNN trained in a novel method is the best, which achieves a relative error of 5.5% for hyperspectral images. The study is fit for optical remote sensing images with complicated ground surfaces and different noise levels captured by different optical remote sensors.