Search

Seagrass beds play a vital ecological role by regulating climate, maintaining biodiversity, and sequestering carbon. Achieving accurate classification of seagrass beds across both time and space is of significant importance for efficient monitoring. This paper addresses the challenges of seagrass bed classification in remote sensing imagery and the difficulty of achieving temporal and spatial transferability of classification model. Leveraging a long-term Landsat satellite image time series spanning from 1976 to 2022 in the Yellow River Delta, we enhance the DeepLab V3+ network and propose a cross-temporal-spatial domain seagrass bed transfer learning classification approach. We utilize selectively chosen images from a few years in partial regions for training, introduce freeze-training and-unfreeze-training during the model training phase, and then transfer the model to achieve cross-temporal and cross-temporal-spatial domain classification on images from other years and regions. Additionally, we discuss the factors influencing the classification results. The research results indicate that: (1)The method we proposed has shown good results, achieving classification accuracy of 83.17% and 80.99% for seagrass beds in both cross-temporal and cross-temporal-spatial domains. (2) The classification results indicate that this method achieves an average classification accuracy higher by 16.18% compared to networks such as UNet, PSPNet, HRNet, UNet++, SegFormer, etc. under cross-temporal conditions and higher by 17.56% under cross-spatiotemporal conditions for seagrass bed classification. (3) Based on the classification results, it is observed that from 1973 to 1986, the overall area of seagrass beds in the Yellow River Delta had an initial growth trend, with the largest seagrass bed area recorded in 1986, reaching 4281.289 hectares. Subsequently, the seagrass beds showed an overall decreasing trend, and by 2019 they had almost completely disappeared. The cross-temporal-spatial domain seagrass bed transfer learning classification method proposed in this paper can provide valuable support for the monitoring, restoration, management, and utilization of seagrass ecosystems. [ABSTRACT FROM AUTHOR]

To address the landslide recognition problem in remote sensing images, this paper designs a visual transformer network model based on DEM (digital elevation model) feature enhancement, which is experimentally validated on the Bijie landslide dataset and Landslide4Sense2022 dataset. The lion optimizer is used during training. The results show that 98.49% accuracy and 97.24% F1 score are achieved on Bijie dataset, and 88.22% accuracy and 90.16% F1 score on Landslide4Sense2022 dataset, which is a significant improvement in landslide recognition compared with other mainstream network models. Therefore, it can be found that this paper's method is effective in the recognition of landslide from remote sensing images. Firstly, the swin transformer network model was successfully applied to remote sensing landslide image classification by means of transfer learning. Secondly, an information enhancement approach based on DEM features was designed. [ABSTRACT FROM AUTHOR]

This paper addresses the challenge of training deep learning models for fire smoke scene detection from multi-sensor, multi-spectral satellite imagery, where spectral bands vary and training data is scarce, especially for new sensors. We introduce a novel cross-sensor transfer learning approach enhanced by spectral pattern extraction, notably aided by a deep learning (DL) module called Input Amplification (IA). Our contribution includes the creation of two datasets, Landsat_smk2770 and Sentinel2_smk, with varying spectral bands. We employ a model named IA_VIB_SD which incorporates IA and is designed specifically for fire smoke scene detection. We investigated the transferability of the Landsat_smk2770-trained IA_VIB_SD model to the Sentinel2_smk dataset using various transfer learning techniques, and compared the performance of the transferred model with the benchmark IA_VIB_SD model exclusively trained on pure Sentinel2_smk data. Our proposed transfer learning approach resulted in a transferred model with an average accuracy 5% higher than the benchmark model. Notably, Our proposed transfer learning approach outperformed existing transfer learning methods by more than 1% in terms of the accuracy, even when trained on a mere 10% of the Sentinel2_smk dataset. This research advances fire smoke scene detection from multi-sensor, multi-spectral satellite imagery. [ABSTRACT FROM AUTHOR]

Automatic building detection from remote sensing images holds significant applications in various domains, including cartography, land-use change detection, urban planning, and 3D intelligent city construction. Driven by the innovation of deep learning technology, convolutional neural networks (CNNs) have emerged as a powerful framework for automatic interpretation, exhibiting hierarchical feature generation and robust representation for building extraction. However, the inherent computational limitations of convolutional operators hinder the acquisition of an effective global receptive field. Recent advancements in deep learning have witnessed the adoption of the multi-head self-attention (MHSA) mechanism by Transformer, presenting the superiority of global context construction. Nonetheless, this network structure encounters difficulties in effectively modelling two-dimensional spatial structure features with scale variation in semantic segmentation and often has huge parameters with high algorithm complexity. To address the above problems, this paper proposed a hybrid deep neural network synergistically harnessing the superiority of CNN and Transformer to extract buildings from remote sensing images. Meanwhile, an improved MHSA mechanism is developed to facilitate the construction of multiscale global representations, concurrently reducing computational complexity. Furthermore, the cross-level MHSA modules the semantic correlation between the deep and shallow-level features to refine spatial fine-grained information. The proposed method was comprehensively evaluated on two building datasets via quantitative analysis and visualization. The experimental results confirmed that the developed network and modules outperform other state-of-the-art methods and can effectively improve the accuracy (93.4% F1-score on the WHU dataset, 89.38% F1-score on the CAMB dataset) with high efficiency for building extraction. [ABSTRACT FROM AUTHOR]

Semantic segmentation of high-resolution remote sensing images is important in land cover classification, road extraction, building extraction, water extraction, etc. However, high-resolution remote-sensing images have a lot of details. Due to the fixed receptive field of convolution blocks, it is impossible to model the correlation of global features. In addition, complex fusion methods cannot integrate spatial and global context information. In order to solve these problems, this paper proposes a cross-linear attention network (CLANet) to capture spatial and context information in images. The structure consists of a spatial branch and a context branch. The spatial branch is constructed by stacked convolution to better capture spatial information. The context branch models the global information based on the transformer deformation module. In addition, to effectively fuse spatial and context information, this paper also designs a feature fusion module (FFM), which uses a cross-linear attention mechanism for feature aggregation. Finally, this paper conducts many experiments on the ISPRS Vaihingen and the ISPRS Potsdam datasets. Among them, 82.28% of mIoU achieves on the ISPRS Vaihingen dataset. The experimental results show that CLANet has better performance and effect than the methods in recent years. [ABSTRACT FROM AUTHOR]

Eddies in the Marginal Ice Zones (MIZ) contribute to an enhanced melting of sea ice by forcing contact between the sea ice and the warmer water off the ice edge. The horizontal current field can be retrieved by tracking the movement of ice floes between sequential satellite images. However, the traditional method of current retrieval imposes stringent requirements on the time intervals between sequential images, and the performance of the traditional method diminishes rapidly with increasing time intervals. In this paper, an improved method of retrieving currents from multi-sensor remote sensing images is proposed, which can significantly improve the accuracy of current retrieval. The method first extracts feature from remote sensing images based on texture features and a support vector machine model and then extracts prior information about the ice-edge eddy from the first two remote sensing images. Under the guidance of prior information, horizontal currents of the eddy are retrieved by tracking the movement of sea ice from sequential remote sensing images. The proposed method was used to retrieve currents from Sentinel-1 and Sentinel-2 sequential images over the MIZ of the Fram Strait. Comparisons with the traditional method show encouraging results that the proposed method can significantly reduce the number of mismatches and improve the reliability of current vectors. [ABSTRACT FROM AUTHOR]

Preserving spectral and spatial information in the satellite image fusion problem is one of the essential challenges. This paper presents a GAN-based method for fusing panchromatic and multispectral satellite images. The proposed method utilises the idea of Cycle-GAN with two generators, for spectral and spatial information preservation, based on residual in residual dense block super-resolution architecture. First, generator-1 translates panchromatic and multispectral satellite images to a high-resolution fused image and then preserves details by utilising generator-2. The goal is to reach the spatial and spectral details of panchromatic and multispectral satellite images, respectively. Two discriminators are employed, one for spectral and the other for spatial transformations, and weighted L1 loss is used for both as cycle losses. By leveraging the unique capabilities of the two generators, the proposed method achieves high-quality image fusion results with improved spectral and spatial resolutions that are evaluated over two different datasets. The experimental results demonstrate the effectiveness of the proposed method, with enhancements ranging from approximately 2% to 30% and 0.5% to 35% in WorldView2 and GeoEye-1 images, respectively, in five metrics including PSNR. Moreover, we show significant enhancements of approximately 7% to 50% and 22% to 29% in the same datasets in metrics like SAM and ERGAS. [ABSTRACT FROM AUTHOR]

Recently, there is a growing interest in analysing the degrading effect of satellite micro-vibrations due to the rapid growth in satellite technologies and the urgent need to precisely extract a huge amount of information from satellite images. Different kinds of micro-vibration have a notable effect on the quality of satellite images. The main objective of this paper is to demonstrate and analyse the effect of all types of micro-vibration on the quality of images acquired by high-resolution satellites. An algorithm to simulate micro-vibrations is proposed. A very high-resolution satellite image from the Pleiades-neo satellite is selected as an example to be used in addressing the degrading effects of micro-vibrations. In this paper, the modulation transfer function (MTF) is used as a major function to model the degradation that has been conducted. Also, several quality metrics are used to quantitatively assess the degradation. The key result of this paper is the significant effect of micro-vibrations on the quality of remote sensing satellite images which is attributed to the main influential parameters. These parameters like blur diameter, vibration displacement, number of Time Delay and Integration (TDI) stages of the camera, and the ratio of the integration time to the vibration period. [ABSTRACT FROM AUTHOR]

Presence of cloud-covered pixels is inevitable in optical remote-sensing images. Therefore, the reconstruction of the cloud-covered details is important to improve the usage of these images for subsequent image analysis tasks. Aiming to tackle the issue of high computational resource requirements that hinder the application at scale, this paper proposes a Feature Enhancement Network(FENet) for removing clouds in satellite images by fusing Synthetic Aperture Radar (SAR) and optical images. The proposed network consists of designed Feature Aggregation Residual Block (FAResblock) and Feature Enhancement Block (FEBlock). FENet is evaluated on the publicly available SEN12MS-CR dataset and it achieves promising results compared to the benchmark and the state-of-the-art methods in terms of both visual quality and quantitative evaluation metrics. It proved that the proposed feature enhancement network is an effective solution for satellite image cloud removal using less computational and time consumption. The proposed network has the potential for practical applications in the field of remote sensing due to its effectiveness and efficiency. The developed code and trained model will be available at . [ABSTRACT FROM AUTHOR]

Atmospheric correction eliminates corruption in reflectance captured by satellite images due to atmospheric elements like gases, aerosols, and water vapours. Existing physics-based approaches employ radiative transfer models constructed using lookup tables computed for different atmospheric parameters. However, these approaches are computationally expensive and rely on estimates of parameters that are difficult to sense accurately. This paper proposes a deep learning model as an alternative to physics-based approaches. We present an end-to-end deep neural network trained on seasonally and spatially rich Landsat 8 satellite images without explicit atmospheric parameterization along with our analysis and its validation. We validate the model's effectiveness vis-a-vis Landsat 8's Land Surface Reflectance Code – LaSRC results in $RMSE \sim 0.042$ R M S E ∼ 0.042 , $SSIM \sim 0.97$ S S I M ∼ 0.97 , and correlation coefficient $r \sim 0.99$ r ∼ 0.99. For ground measurements by RadCalNet, the proposed model has an $RMSE \sim 0.053$ R M S E ∼ 0.053 , $SSIM \sim 0.90$ S S I M ∼ 0.90 , and $r \sim 0.88$ r ∼ 0.88. The results show that the model accurately predicts surface reflectance and correlates highly with reference data. [ABSTRACT FROM AUTHOR]

Domain adaptation is the ability to improve the learning efficiency of the target domain by using the prior knowledge of the source domain. When applied to new tasks in the target domain, the performance of domain adaptation trained in the source domain declines sharply. The purpose is to realize a new retrieval task in the target domain using only a small number of labelled data samples, with the aid of the prior knowledge learned in the source domain. This paper focuses on semi-supervised domain adaptation in remote sensing image retrieval. The contributions of this paper are threefold. First, we construct Gabor-based CNNs to facilitate the networks to effectively capture the texture information of images. Second, we propose a cross-domain knowledge transfer strategy based on dual Gabor neural network learning. Third, we propose an unsupervised random feature mapping method based on probability distance. A large number of experiments have been conducted on UCM, WHU-RS, RSSCN7, AID, and PatternNet datasets. The results show that this method greatly improves the retrieval accuracy on the target domain and obtains state-of-the-art retrieval accuracy. The source code is available at . [ABSTRACT FROM AUTHOR]

Accurate and unified information from the increasingly remote sensing (RS) scenes is important for RS applications in multi-sectoral association services of natural resource management. However, these applications in mountain areas are limited by the challenging issues of random geometric distortions and erroneous spatial associations. The paper introduces digital elevation model (DEM) maps as a unified geographic reference to search and match homonymy ground points (HGPs). The proposed computer-based procedure was tested with Landsat TM, ETM and HJ-1B satellite images using ASTER global DEM in the Longitudinal Rang-gorge Valley Region of Southwest China. 1322, 3551 and 694 pairs of HGPs were identified and acquired the geometric accuracies with 43 m (TM), 14 m (ETM) and 123 m (HJ), respectively. The deviations are significantly reduced and the disjoint ground objects are matched. The study satisfies the application requirement of multispectral satellite imagery with less labour and time costs. [ABSTRACT FROM AUTHOR]

This paper presents an overview of the effectiveness of satellite images with remote sensing applications in determining potential fishing grounds and its applicability for the Indian mackerel (R. kanagurta) in the Malaysian EEZ off the South China Sea. Most oceanic fish species tend to aggregate in large schools that can span tens of kilometers. Thus, finding fish schools and productive fishing areas is the main cause of fuel consumption and ship time expense in many commercial fisheries. In order to lower the cost of fishing operations, there is a need to accurately predict and detect economically fishable aggregations of fish in space and time. Recent studies have shown that the combined use of remote sensing together with statistical analysis and the advancement of geographical information analysis can be used not only to help manage fisheries at sustainable levels but also guide fishing fleets to increase their catch. These modern techniques may facilitate the implementation of an ecosystem-based approach to fisheries management. The assessment of the effectiveness of satellite images in predicting potential fishing grounds and its applicability for the Indian mackerel (R. kanagurta) in Peninsular Malaysia are reviewed. [ABSTRACT FROM AUTHOR]

Throughout the Syrian conflict, Raqqa has suffered untold losses, especially after being claimed as the capital of the so-called Islamic State. Damage to Raqqa's built cultural heritage, a component of the human right to identity and culture, is one small aspect of the city's humanitarian crisis. Raqqa's Abbasid-era wall encircling the city's historic core, a UNESCO tentative world heritage site, was the frontline in the battle for the city in the summer of 2017. This paper seeks to establish the extent of visible United States (US)-led Coalition airstrike damage to Raqqa's historic city wall through a time-series analysis of commercially available high-resolution satellite imagery. Much has been published on damage to cultural heritage in Syria, yet little has focused on damage specifically caused by US-led Coalition airstrikes. When compared with other parts of Raqqa, this paper assesses that the city wall did not suffer the same degree of airstrike damage as did its surroundings, suggesting that effort was made by the US-led Coalition to minimize damage to this heritage site. This paper considers international treaty obligations under the 1954 Hague Convention and recent US policy shifts in light of the results of this assessment. [ABSTRACT FROM AUTHOR]

Microbial biofilms have received great attention in the last few decades from both aquatic ecologists and biogeomorphologists. Most recently, this has focused on mapping biofilms to understand their spatial distributions and ecosystem services. Such studies often involve the use of satellite imagery, which typically provides large temporal and spatial scales and wide-range spectral information. Although satellites have the advantage of multi- and hyper-spectral sensors, images often have low spatial resolution that limits their use in river studies, where both rivers are narrower and stream processes occur at resolutions smaller than the footprint of satellite sensors. Spatial resolution is sensor quality dependent but also controlled by sensor elevation above the ground. Hence, high resolutions can be achieved either by using a very expensive sensor or by decreasing the distance between the target area and the sensor itself. To date, sensor technology has advanced to a point where multi- or even hyper-spectral cameras can be easily carried out by an Uncrewed Aerial Vehicle (UAV) at unprecedented spatial resolutions. Where such sensors have high spectral resolution, they are often prohibitively expensive, especially as their use in extreme environments such as glacial forefields risks UAV damage. In this paper, we test the performance of visible band ratios in mapping biofilms in an Alpine glacier forefield characterized by a well-developed and heterogeneous stream ecosystem but using a low-cost UAV. The paper shows that low-cost and consumer grade UAVs can be easily deployed in such extreme environments, delivering both quality RGB images for photogrammetric (SfM-MVS) processing and sufficient spectral information for benthic biofilm mapping at high temporal and spatial resolution. RGB cameras are an alternative to expensive multi- or hyper-spectral cameras. Phototrophic biofilms can be detected and mapped through visible band ratios. High-temporal and high-resolution imagery can be collected by consumer-grade UAVs. Biofilm presence is restricted to stable and water-fed terraces in summer. [ABSTRACT FROM AUTHOR]

As ship target detection technology has high application value in military and civil fields, it is significant to research ship detection in SAR images. Aiming at the complex and diverse backgrounds, significant differences in ship sizes, and real-time detection problems in the ship target detection task of SAR remote sensing images, a lightweight ship detection network based on the YOLOx-Tiny model is proposed. Firstly, a multi-scale ship feature extraction module is proposed, composed of a parallel multi-branch structure connected by a standard convolution layer, asymmetric convolution layer, and dilatation convolution layer with different expansion rates in turn. It makes better use of local features and global features and effectively improves the detection accuracy of multi-scale ship targets; Secondly, to ensure detection performance and eliminate background interference, we propose a whole SAR remote sensing image detection strategy based on an adaptive threshold, which effectively suppresses false alarms caused by background and improves detection speed. The experimental results on two different SAR ship datasets, SSDD and HRSID, show that, compared with several advanced methods, the effectiveness and superiority of the method in this paper are verified, and excellent results are shown in the detection of the whole SAR remote-sensing image. It can provide effective theoretical and technical support for ship detection on platforms with limited computing resources and has good application prospects. [ABSTRACT FROM AUTHOR]

Proper management of the earth's natural resources is imperative to combat further degradation of the natural environment. However, the environmental datasets necessary for informed resource planning and conservation can be costly to collect and annotate. Consequently, there is a lack of publicly available datasets, particularly annotated image datasets relevant for environmental conservation, that can be used for the evaluation of machine learning algorithms to determine their applicability in real-world scenarios. To address this, the Time-evolving Data Science and Artificial Intelligence for Advanced Open Environmental Science (TAIAO) project in New Zealand aims to provide a collection of datasets and accompanying example notebooks for their analysis. This paper showcases three New Zealand-based annotated image datasets that form part of the collection. The first dataset contains annotated images of various predator species, mainly small invasive mammals, taken using low-light camera traps predominantly at night. The second provides aerial photography of the Waikato region in New Zealand, in which stands of Kahikatea (a native New Zealand tree) have been marked up using manual segmentation. The third is a dataset containing orthorectified high-resolution aerial photography, paired with satellite imagery taken by Sentinel-2. Additionally, the TAIAO web platform also contains a collated list of other datasets provided and licensed by our data partners that may be of interest to other researchers. [ABSTRACT FROM AUTHOR]

Black and odorous water seriously affects the ecological balance of rivers and the health of people living nearby. Satellite remote sensing technology with its advantages of a large range, long-time series, low cost, and high efficiency, has provided a new area for water quality detection. Much archived remote sensing satellite data can be further processed and used as a data source for black and odorous water detection. In this paper, Gaofen-2 remote sensing data with a spatial resolution of 1 m is leveraged as the data source. To enrich the data source in the northern coastal zone of China, we have built a high-quality remote sensing dataset, called the remote sensing images for black and odorous water detection (RSBD) dataset, which is collected from the Gaofen-2 satellite in Yantai, China. In addition, we propose a network with an encoder-decoder discriminant structure for black and odorous water detection. In the network, an augmented attention module is designed to capture a more comprehensive semantic feature representation. Further, the median balancing loss function is adopted to solve the imbalance issues. Experimental results demonstrate that the network is superior to other state-of-the-art semantic segmentation methods on our dataset. [ABSTRACT FROM AUTHOR]

Image semantic segmentation methods based on convolutional neural networks rely on supervised learning with labels, and their performance often drops significantly when applied to unlabelled datasets from different sources. The domain adaptation methods can reduce the inconsistency of feature distribution between the unlabelled target domain data used for testing and the labelled source domain data used for training, thus improve the segmentation performance and have more practical applications. However, in the field of remote sensing image processing, if the spatial resolutions of the source domain and the target domain are different and this problem is not to be solved, the performance of the transferred model will be affected. In this paper, we propose a bidirectional semantic segmentation method based on super-resolution and domain adaption (BSSM-SRDA), which is suitable for the transfer learning task of a semantic segmentation model from a low-resolution source domain data to a high-resolution target domain data. BSSM-SRDA mainly consists of three parts: a shared feature extraction network; a super-resolution image translation module, which incorporates a super-resolution approach to reduce spatial resolution differences and visual style differences of the two domains; a domain-adaptive semantic segmentation module, which combines an adversarial domain adaptation approach to reduce differences at the output level. At the same time, we design a new bidirectional self-supervised learning algorithm for BSSM-SRDA that facilitates mutually beneficial learning of the super-resolution image translation module and the domain-adaptive semantic segmentation module. The experiments demonstrate the superiority of the proposed method over other state-of-the-art methods on two remote-sensing image datasets, with mIoU improvements of 2.5% and 3.2%, respectively. Code: [ABSTRACT FROM AUTHOR]

Modern earth observation sensors have revolutionized the remote sensing community by improving remote sensing image quality. However, Pixel-based image analysis methods have challenges in handling very high-resolution (VHR) imagery. Geographic Based Image Analysis (GEOBIA) yielded promising results, but it is not inflexible in capturing domain experts' expressions, therefore geographic information system professionals shifted to ontologies for remote sensing science. This paper advocates for the adoption of knowledge representation using ontologies in remote sensing. To this end, a survey of GEOBIA studies for image analysis and classification is presented, and the limitations of existing methods in reaching the remote sensing expert-level expectation are clarified. New GEOBIA development techniques as well as opportunities for improving GEOBIA models have been looked into. Recent studies that adopted ontologies in forest image classification are analyzed and recommendations for the remote sensing science community are provided, to highlight the advantages of ontologies in interpreting satellite images. [ABSTRACT FROM AUTHOR]

This paper presents a machine learning method to map groundwater potential in crystalline domains. First, a spatially-distributed set of explanatory variables for groundwater occurrence is compiled into a geographic information system. Twenty machine learning classifiers are subsequently trained on a sample of 488 boreholes and excavated wells for a region of eastern Chad. This process includes collinearity, cross-validation, feature elimination and parameter fitting routines. Random forest and extra trees classifiers outperformed other algorithms (test score > 0.80, balanced score > 0.80, AUC > 0.87). Fracture density, slope, SAR coherence (interferometric correlation), topographic wetness index, basement depth, distance to channels and slope aspect proved the most relevant explanatory variables. Three major conclusions stem from this work: (1) using a large number of supervised classification algorithms is advisable in groundwater potential studies; (2) the choice of performance metrics constrains the relevance of explanatory variables; and (3) seasonal variations from satellite images contribute to successful groundwater potential mapping. [ABSTRACT FROM AUTHOR]

This paper discusses the use of a small ultraportable hyperspectral camera in the VIS-NIR (Visible-Near Infrared) range for archaeological fieldwork and its hardware, data processing workflows, and spectral information that can be used for in situ screening. Hyperspectral imaging is a widespread, non-destructive analytical technique used in various disciplines for highlighting invisible patterns and mapping the spectral signatures of selected targets. In archaeology, it has mostly been applied for remote sensing satellite imagery to uncover information about features that are hidden underground. Targeted applications of hyperspectral imaging have been developed in the last few years, opening new perspectives for material analysis based on spectral mapping. Recent advances in portable instrumentation have led to the development of small and rugged cameras that can be used directly in the field for investigating different materials. [ABSTRACT FROM AUTHOR]

Remote sensing has become an important technique for monitoring vegetation phenology. The quality of remote-sensing images and derived products is key to successful extraction of vegetation phenophases. There is a need to develop quality management methods to evaluate the data uncertainty and assist the removal of the noises. This paper developed a shape quality assurance score threshold (SQAT) method which accounts for the trend in satellite-derived vegetation index associated with the process of vegetation growth. The proposed method was tested on six widely used methods for extracting vegetation phenophases. Results showed that the SQAT method can effectively identify noises in the vegetation index time series and improve the accuracies of estimated start of season (SOS) and end of season (EOS) of the six methods. After removal of identified noises, the Pearson correlation coefficient (r) averagely increased by 8% for SOS, and 11% for EOS. Regression analyses of vegetation phenophases between the PhenoCam observations and MCD12Q2 product showed that the proposed method performs better than the QA score of MCD12Q2 for quality management. This paper provides promising method for quality management; it has the potential to reduce the uncertainty of the vegetation index time series that can support studies of vegetation phenology monitoring. [ABSTRACT FROM AUTHOR]

Multi-temporal remote sensing imagery has the potential to classify river landforms to reconstruct the evolutionary trajectory of river morphologies. Whilst open-access archives of high spatial resolution imagery are increasingly available from satellite sensors, such as Sentinel-2, there remains a fundamental challenge of maximising the utility of information in each band whilst maintaining a sufficiently fine resolution to identify landforms. Although image fusion and downscaling methods on Sentinel-2 imagery have been investigated for many years, there is a need to assess their performance for multi-temporal object-based river landform classification. This investigation first compared three downscaling methods: area to point regression kriging (ATPRK), super-resolution based on Sen2Res, and nearest neighbour resampling. We assessed performance of the three downscaling methods by accuracy, precision, recall and F1-score. ATPRK was the optimal downscaling approach, achieving an overall accuracy of 0.861. We successively engaged a set of experiments to determine an optimal training model, exploring single and multi-date scenarios. We find that not only does remote sensing imagery with better quality improve river landform classification performance, but multi-date datasets for establishing machine learning models should be considered for contributing higher classification accuracy. This paper presents a workflow for automated river landform recognition that could be applied to other tropical rivers with similar hydro-geomorphological characteristics. Choice of downscaling approach influences the performance of river landform classification from satellite imagery and should be considered in river and flood management. An efficient and straightforward operating workflow was developed for automated river landform classification with high accuracy that supports an improved understanding of the use of machine learning approaches in river landform recognition. Freely available and easy-to-access remote sensing datasets can help extend the operating workflow to difficult-to-access or remote regions and allow for complete regional and/or national coverage. [ABSTRACT FROM AUTHOR]

Classification-based methods for estimating water quality parameter (WQP) using remote sensing have shown great application potential in inland waters. Water classification algorithms have seen progress in water remote sensing. In this paper, we conducted the Secchi Depth value (ZSD) estimation based on a global water typology for the Wuhan area. Firstly, we classified the water into seven types using the Sentinel-2 data. The procedure was based on spectral angle mapping (SAM) of the 13 inland water optical types (IWOTs). Afterwards, seven IWOTs were summarized into four categories for Wuhan area. We then developed empirical models for each water category by stepwise multiple linear regression, generalized regression neural network (GRNN), and sparse spectrum Gaussian process regression (SSGPR), and applied the better approaches (GRNN and SSGPR) to three full satellite images (27 October 2018, 10 May 2019, and 29 July 2019). Finally, the retrieved results were validated using in situ-satellite match-ups and compared with the results based on unclassified imagery. With root-mean-square error (RMSE) of three satellite-derived results reduced from 0.32 m (without classification) to 0.16 m (with classification), and mean absolute percentage error (MAPE) reduced from 52% to 18%, from 0.51 m (MAPE = 57%) to 0.19 m (MAPE = 21%), and from 0.18 m (MAPE = 35%) to 0.09 m (MAPE = 17%), ZSD estimations over optically complex waters were improved based on this water classification. Due to its low cost and ease of operation, the SAM – derived classification applied in this paper provides a possibility for dynamic and high-precision monitoring for water management. [ABSTRACT FROM AUTHOR]

Multi-spectral satellite remote sensing imagery have several applications including detection of objects or distinguishing land surface areas based on amount of greenery or water etc. The enhancement of spectral images helps extracting and visualizing spatial and spectral features. This paper identifies some specific regions of interest (RoI) of the earth's surface from the remotely sensed spectral or satellite image. The RoI are extracted and identified as major segments. Trivially, uni-variate histogram thresholding is used for gray images as a tool of segmentation. However, for color images multivariate histogram is effective to get control on color bands. It also helps emphasizing color information for clustering purpose. In this paper, the 2D and 3D histograms are used for clustering pixels in order to extract the RoI. The RGB color bands along with the infrared (IR) band information are used to form the multivariate histogram. Two datasets are used to carry out the experiment. The first one is an artificially designed dataset and the next is Indian Remotely Sensed (IRS-1A) satellite imagery. This paper proves the correctness of the proposed mathematical implication on the artificial dataset and consequently perform the application on LandSat Spectral data. The test result is found to be satisfactory. [ABSTRACT FROM AUTHOR]

Information about the state of the Arctic sea ice is becoming increasingly important. This paper describes an approach for automatic retrieval of daily thin sea ice thickness maps around Svalbard. The algorithm uses thermal satellite imagery from MODIS to estimate the surface temperature of the ice and further uses a thermal model of the ice surface to estimate the thickness of the sea ice. The approach is usable for thin sea ice, up to ca. 50 cm thick, during cold weather (freezing) conditions and without cloud cover present. The algorithm is compared with helicopter-borne electromagnetic ice thickness measurements. The comparison yields increasing root-mean-square deviation (RMSD) for thicker ice. The lowest RMSD found is 8.7 cm for ice thickness in the range 10 cm < hi ≤ 20 cm. The highest RMSD found is 25.2 cm for ice thickness in the range 30 cm < hi ≤ 40 cm. The bias shows no such trend, and the overall bias is found to be −5.5 cm. The results show that this is a promising approach, allowing monitoring of thin sea ice thickness at relatively higher spatial resolution. [ABSTRACT FROM AUTHOR]

Ship detection for SAR imagery plays a crucial rule in the field of remote-sensing image processing. Superpixel methods have attracted enormous interest in the recent years, resulting in superpixel-based CFAR (Constant False Alarm Rate) ship detection methods for SAR images become a hot research issue. However, existing methods take superpixel generation as a preprocessing, and CFAR detection is essentially carried out at the pixel level, and the influence of speckle noise has not been fundamentally overcome. This paper innovatively proposed a real superpixel-level CFAR detection method, which takes superpixels instead of pixels as the minimum processing primitives on the basis of reliable superpixel segmentation. Firstly, in order to obtain the background superpixel region of each under test superpixel more accurately and efficiently, we designed a ring topology, which was specially developed for non-Euclidean structure data like superpixels. Secondly, the Johnsonsu function is used to fit the superpixel clutters from the background ring, and the detection threshold was obtained by a given false alarm rate. Finally, the candidate targets were further optimized to carry out accurate detection results. Experimental results based on multiple sets of real SAR images shown that our method has significant advantages in terms of detection rate and false alarm rate compared to the traditional K-CFAR and the existing superpixel-based CFAR methods. [ABSTRACT FROM AUTHOR]

Many papers have reviewed remote sensing image segmentation (RSIS) algorithms currently. Those existing surveys are insufficiently exhaustive to sort out the various RSIS methods, it is impossible to comprehensively compare characteristics of different RSIS methods. In addition, the segmentation efficiency and accuracy of the RSIS methods cannot always meet the subsequent image analysis requirements. Thus, a clear comparative analysis of various RSIS methods is essential to provide an in-depth understanding of RSIS and theoretical ideas for conducting in-depth research in the future. The goal of this article is to provide readers with the latest information on optical RSIS technology. Comparative measures of these methods are provided in terms of conceptual details, the merits and demerits, and the performance of various RSIS methods. Moreover, various RSIS methods' experiments are carried out on optical images using the NWPU VHR-10 public remote sensing datasets. Through the review of optical RSIS methods, this paper provides data as complete as possible for further related research and development of RSIS methods. [ABSTRACT FROM AUTHOR]

High-resolution remote sensing scene classification is a widely applicable task. Due to the diversity of natural scenes and acquisition methods, in different satellite images, scenes of the same class are often variable in texture, background, illumination and spatial resolution. Thus, it is hard for a remote sensing sense database to contain enough representative images. In this case, the generalization error of traditional supervised classification methods might be too large to generate ideal results. Domain adaptation (DA) has been applied to image classification by reducing feature distribution discrepancy between the source domain (where labels are available) and the target domain (where images need to be classified). In this paper, we propose an unsupervised adversarial domain adaptation method boosted by a domain confusion network (ADA-BDC) which aims at adapting the images from different domains to appear as if drawn from the same domain and improve the transferability of our classifier. For this purpose, the feature extractor in ADA-BDC makes the source and target distributions closer by training a Generative Adversarial Nets (GAN) model. After that, a transferred classifier trained by transferred source domain features is able to acquire a better classification accuracy on the target domain than a non-transferred classifier. In this paper, the experiments are conducted on four remote sensing scene benchmark datasets which are different in spatial scale, resolution, land-cover pattern, etc. Experimental results show that our proposed method is able to improve the transferability across different datasets and improve the classification overall accuracy by 17.18% on average. The comparative experiments demonstrate that the proposed DA network outperforms the compared state-of-the-art domain adaptation methods on remote sensing image scene classification. [ABSTRACT FROM AUTHOR]

Object detection is a key technique for automatic information extraction, analysis and understanding of high spatial resolution remote-sensing images (HSRIs). Object anchor scales are a critical factor for object detection of HSRIs based on convolutional neural networks (CNNs). With respect to adaptively obtaining optimal object anchor scales for object detection of HSRIs, this paper proposes a novel object detection method for HSRIs based on CNNs with optimal object anchor scales. First, optimal object anchor scales for object detection of HSRIs are obtained using an adaptive object-scale learning operator. Then, a CNN object detection framework for HSRIs is designed based on optimal object anchor scales. Using multiple-object detection datasets, the proposed method is compared with some state-of-the-art object detection algorithms. Experimental results show that the proposed method can achieve 90.42%, 91.98% and 85.07% mAP on WHU-RSONE, UCAS-AOD and HSRC2016, respectively, and outperforms state-of-the-art object detection algorithms. [ABSTRACT FROM AUTHOR]

This paper focuses on spatial quality assessment of pan-sharpened imagery that contains valuable information of input images. Its aim is to show that fusion functions respond differently to different types of landscapes. It compares a quality assessment of an object-level procedure with that of a conventional pixel-level-based procedure which assigns uniform quality scores to all image pixels of pan-sharpened images. To do so, after performing a series of pan-sharpening evaluations, a weighted procedure for spatial quality assessments of pan-sharpening products, allocating spatially varying weight factors to the image pixels proportional to their level of spatial information content is proposed. All experiments are performed using five high-resolution image datasets using fusion products produced by three common pan-sharpening algorithms. The datasets are acquired from WorldView-2, QuickBird, and IKONOS. Experimental results show that the spatial distortion of fused images for the class vegetation cover exceeds that of man-made structures, reaching more than 4% in some cases. Our procedure can preclude illogical fidelity estimations occurring when pan-sharpened images contain different land covers. Since particular image structures are of high importance in remote sensing applications, our procedure provides a purpose-oriented estimation of the spatial quality for pan-sharpened images in comparison with conventional procedures. [ABSTRACT FROM AUTHOR]

Building footprint maps are useful for urban planning, infrastructural development, population estimation and disaster management. With the availability of very-high resolution satellite imagery, remote sensing community is pursuing automatic techniques for extracting building footprints for cities with varried building types. Recently, CNNs (Convolutional Neural Network) have been successfully applied for extraction of building footprint from satellite imagery. In this paper, we propose a novel CNN architecture termed UNet-AP inspired by UNet and the concept of Atrous Spatial Pyramid Pooling, for automatic extraction of building footprint from very-high resolution satellite imagery. We demonstrate extraction of building footprints from Cartosat-2 series 4-band (Blue, Green, Red and Near-Infrared) multispectral satellite imagery, pan-sharpened using the panchromatic image with less than 1-meter resolution. We also compare the performance of our proposed architecture with baseline implementation of recently proposed UNet and SegNet architectures. We present a comparative assessment of the architecture performance across different types of urban settlement classes such as dense built-up areas, slums and isolated buildings. We demonstrate that our proposed architecture outperforms SegNet and UNet in terms of overall mean intersection over union (0.75 vs 0.70 and 0.68 for UNet and SegNet respectively) and delivers consistent improvement across all three settlement classes. [ABSTRACT FROM AUTHOR]

Deep convolutional neural networks (CNN) have been widely applied in various fields, especially in the field of object detection. Deep CNN-based models showed great advantages over many traditional methods, even so, there are still many specific problems in the application of certain scenarios. In very high resolution (VHR) remote-sensing image datasets, the uncertainty of the object direction angle causes big trouble to the learning of the detector. Although the pooling operation can slightly alleviate the deviation caused by small angle, the feature learning of the objects with larger angle rotation still relies mainly on the sufficiency of sample data or effective data augmentation, which means the insufficiency of the training instances may cause serious performance degradation of the detector. In this paper, we propose a multi-angle box-based rotation insensitive object detection structure (MRI-CNN), which is an extended exploration for typical region-based CNN methods. On the one hand, we defined a set of directionally rotated bounding boxes before learning, and restricted the classification scene in a small angular range by rotated RoI (Region of Interest) pooling. On the other hand, we proposed a more effective screening method of bounding boxes, enabling the detector to adapt to diverse ground truth annotation methods and learn more accurate object localization. We trained our detector with different datasets containing different amount of training data, and the test results showed that the method proposed in this paper performs better than some mainstream detection methods when limited training data are provided in VHR remote-sensing datasets. [ABSTRACT FROM AUTHOR]

Morphological processing has found several applications in image analysis and pattern recognition. Some of these techniques, known as morphological reconstruction algorithms, have been employed for land cover classification in remote sensing data. In this paper, we analyse the mathematical foundations, applications, and limitations of reconstruction by dilation and by erosion oriented to urban extraction, using Sentinel-2 satellite data. Different techniques oriented to the proper determination of the marker and mask images, the basis for reconstruction, are proposed in this manuscript. In addition, in order to diminish the long computation time required for reconstruction, two parallel implementations using Multi-core and GPU, are proposed. According to our research, these algorithms can be considered as effective and non-supervised solutions for urban extraction applications based on multispectral satellite imagery. [ABSTRACT FROM AUTHOR]

The multiclass change detection task aims to segment an area of land change and to identify the type of ground change exploiting bi-temporal remote-sensing images. In recent years, this technology has become an important means of analysing remote-sensing data and has shown great potential in accurately understanding changes on the earth's surface. However, the existing methods are mainly for binary change detection tasks and cannot detect the specific information of the change category. Hence, in this paper, we propose an end-to-end densely connected network, entitled Y-Net, employing a Siamese structure for multiclass change detection. The proposed network uses a dual-stream DenseNet to extract bi-temporal change features during the encoding stage and introduces the attention fusion mechanism for the decoding stage to enhance the attention to changing features. Add the dice loss to reduce the impact of sample imbalance on multiclass change detection. The suggested network utilizes a deep supervision strategy for each decoding layer to achieve change graph reconstruction of different levels. Y-Net is evaluated on the challenging multiclass change detection data set SECOND, and the ablation study verified the effectiveness of the attention fusion module and dice loss for multiclass change detection, considering both visual interpretation and quantitative evaluation, confirming that the proposed method is superior to current state-of-the-art change detection methods. [ABSTRACT FROM AUTHOR]

In remote sensing, for applications as environment monitoring, change detection based on image processing is one of the most important techniques. To reach high performance various techniques of fusion are exploited using a combination of multi-temporal, multispectral and panchromatic satellite images. A solution for handling such kind of images holds when using some simple statistical methods like the Percent Difference (PD) technique as well as the Principal Component Analysis (PCA) one. In this paper, an automatic change detection method issued from the two previous techniques is proposed and applied on multispectral and panchromatic images captured by a high resolution optical satellite. This approach is characterized by two aspects: the first one consists of the fusion of the different data and the second one performs the detection of the changes for the resulting images. The experimental results show the reasonable quantitative performance and the effectiveness of the proposed method for change detection, consisting of an automatic extraction of most of change information as well as the obtention of better results for most precision metrics consisting of an overall accuracy of up to 91% and a Kappa coefficient of up to 66%, comparing to those obtained using the simple PD and PCA techniques. [ABSTRACT FROM AUTHOR]

The rational function model (RFM) is widely used for high-resolution satellite imagery allowing its use in GIS and remote sensing applications. RFM can be seen as a non-parametric model that establishes a geometric transformation between image and ground coordinate spaces. Unlike the rigorous models, it constitutes an easy and simple approach where any prior knowledge about the sensor's physical configuration and parameters is required. For a precise ortho-rectification, RFM requires a large number of accurate and well-distributed ground control points (GCPs) which is a time-consuming and expensive process. Moreover, adjustment errors may occur due to the over-parameterization resulting from the high number of coefficients in this model. In fact, coefficients in RFM do not have any physical meaning, which makes it impossible to find their best combination. In order to overcome this problem, optimization based on evolutionary algorithms seems to be an appropriate solution. In this paper, a hybrid binary of particle swarm optimization (PSO) method is proposed. It combines the PSO concept with that of genetic algorithms by adding two new operations in the binary version of PSO consisting of crossover and mutation in order to increase the convergence speed and avoid the local optimum phenomenon. The proposed algorithm has been applied on two data sets provided by the Alsat-2 Algerian satellite in different configurations, namely standard scenes and strips which are a group of scenes acquired continuously from the same orbit. To the best of our knowledge, this is the first work dealing with the use of PSO for the RFM optimization aiming at the geometrical modelling of a strip of images. The obtained results have shown that the proposed HPSO-RFM method outperforms other competing methods by achieving, for the best scenario, an accuracy improvement of about 38% and 21% over PSORFO and PSO-KFCV, respectively. [ABSTRACT FROM AUTHOR]

Remote sensing image segmentation has large uncertainty related to the heterogeneity of similar objects and complex spectrum in satellite images, causing the traditional segmentation methods to be greatly limited. Existing semantic segmentation methods represented by deep learning have made breakthrough progress. However, traditional deep learning methods, such as deep convolution neural network, are a completely deterministic model, which cannot describe the uncertainty of remote sensing image well. To solve this problem, a new deep neural network combined with fuzzy logic units is proposed in this paper, called RSFCNN (Remote Sensing image segmentation with Fuzzy Convolutional Neural Network). The network integrates convolution units and fuzzy logic units. Convolution units are used to extract discriminant features with different proportions, thus providing comprehensive information for pixel-level remote sensing image segmentation. Fuzzy logic units are used to deal with various uncertainties and provide more reliable segmentation results, and each unit handles the feature maps at a particular image scale by Gaussian blur function. Experiments were carried out on two data sets, and the results showed that RSFCNN has higher segmentation accuracy and better performance than state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Metadata information and catalogue services are major ways of making satellite images findable and accessible. Spatio-temporal indexing is the key to ensuring efficient searches. Because spatial information and temporal information are usually independently maintained and indexed, the image retrieval process has to include two search steps: a spatial query and a temporal query. As most Earth Observation satellites are specially designed to have repeating sun-synchronous orbits (RSSO), this type of satellite data has a close correlation between its spatial coverage and temporal coverage information. In this paper, an integrated spatio-temporal indexing mechanism is proposed for RSSO satellites. The spatio-temporal Look-Up Table (st-LUT) that serves as the index reflects the coupled correlation between the spatial and temporal coverage information within one orbit revisiting cycle. Image retrieval algorithms are designed based on the st-LUT. In this study, 1,765,797 Landsat 8 scenes collected from 28 June 2013 to 31 December 2019 data are used to establish and validate the proposed indexing mechanism and search algorithms. Because this new method only need to focus on the changes of the spatial and temporal coverage over the time in one orbit revisiting cycle, the spatial search space is limited to the fixed number of grids. Therefore, the search algorithm is at a constant level. Its performance is not related to the volume of the images that need to search. [ABSTRACT FROM AUTHOR]

Over the last two decades, there has been a significant improvement in the quality and volume of the remote-sensing images. With the extensive availability of the images, these are used in many applications such as change detection, image mosaicing, and image fusion where image registration is very necessary. However, image registration is a challenging task in remote sensing due to the following reasons: geometric differences between images, intensity differences, and noise affect. Over the years, wide variety of algorithms have come into existence to handle these issues. In this paper, a comprehensive review is presented on these remote sensing image registration methods. At first, the general information of remote sensing image registration, its classifications, and application areas are provided. Then, an exhaustive review is given on the existing methods and their merits as well as shortcomings are mentioned. Finally, a comparative analysis is provided for the state-of-the-art methods and current challenges of this research are presented. [ABSTRACT FROM AUTHOR]

Ｗith the rapid development of remote sensing, satellite video has become an important data source for vehicle detection, which provides a broader field of surveillance. The achieved work generally focuses on aerial video with moderately sized objects based on feature extraction. However, the moving vehicles in satellite video imagery range from just a few pixels to dozens of pixels and exhibit low contrast with respect to the background, which makes it hard to get available appearance or shape information this paper, a tiny vehicle detection method based on spatio-temporal information is proposed to constrain the significance of the image. Firstly, the background modelling method is used to obtain the motion heat map of the image and constrain the motion region. A significance detection method for small targets was used to obtain the significance mapping of these regions. Finally, the detection results were optimized by combining the significance neighbourhood information and the time information between frames to output the binary target detection map. Finally, taking different urban road scenes in 'Jilin-1'satellite video as examples and compares a variety of existing algorithms. Experiments prove that the proposed algorithm can maintain false alarm rate of less than 10% when the detection accuracy and recall rate reach 85% and has certain anti-interference ability in the image environment with satellite Angle deviation. [ABSTRACT FROM AUTHOR]

This paper proposed a method of extracting the winter wheat area by combining support vector machine (SVM) with variable fuzzy sets. This method mainly aims to deal with mixed pixels in remote sensing images. The SVM classifier can accurately identify pure winter wheat pixels with training samples. However, winter wheat area information in mixed pixels cannot be directly obtained because they contain multiple types of features. In order to estimate the winter wheat area information of mixed pixels in Landsat8 data, this paper introduced the normalized difference vegetation index (NDVI) and the variable fuzzy set method. Finally, this paper reasonably estimated the area of winter wheat from the mixed pixels in Landsat8 images. The accuracy assessment showed that the proposed method could extract winter wheat area information more accurately. [ABSTRACT FROM AUTHOR]

Detecting geospatial objects, especially small, time-sensitive targets such as airplanes and ships in cluttered scenes, is a substantial challenge in large-scale, high-resolution optical satellite images. Directly detecting targets in countless image blocks results in higher false alarms and is also inefficient. In this paper, we introduce a hierarchical architecture to quickly locate related areas and detect these targets effectively. In the coarse layer, we use an improved saliency detection model that utilizes geospatial priors and multi-level saliency features to probe suspected regions in broad and complicated remote sensing images. Then, in the fine layer of each region, an efficacious end-to-end neural network that predicts the categories and locations of the objects is adopted. To improve the detection performance, an enhanced network, adaptive multi-scale anchors, and an improved loss function are designed to overcome the great diversity and complexity of backgrounds and targets. The experimental results obtained for both a public dataset and our collected images validated the effectiveness of our proposed method. In particular, for large-scale images (more than 500 km2), the adopted method far surpasses most unsupervised saliency models in terms of the performance in region saliency detection and can quickly detect targets within 1 minute, with 95.0% recall and 93.2% precision rates on average. [ABSTRACT FROM AUTHOR]

Over the past few years, Convolutional Neural Networks (CNN) have grown in popularity with the remote sensing community due to their relatively easy training process, excellent generalization capacity and state-of-the-art performance. CNN models following an encoder-decoder architecture to perform end-to-end semantic segmentation have been applied to Change Detection (CD) applications with remarkable results. In this paper, we aim to further improve the performance of such models. First, we experiment with the introduction of additional boundary information into an encoder-decoder architecture that performs semantic segmentation for CD. We use the Dense Extreme Inception Network (DexiNeD) to produce the semantically informed edges. Second, we propose a training process that implicitly teaches the model to become more robust to misregistration errors. We evaluate our proposed approaches on a CD dataset, which consists of very high resolution RGB satellite image pairs, using two encoder-decoder models, UNet and UNet++, as our backbone architecture. The evaluation results suggest that both enhancements improve the performance of the CD network, with the average improvement on precision, recall, F1score and IoU ranging between 1% and 2% when incorporating boundary features into our architectures, and up to 2.5% when modeling the misregistration errors in the training process. [ABSTRACT FROM AUTHOR]

Recently, deep learning (DL) techniques including Convolutional neural network (CNN), Recurrent neural network (RNN), and Recurrent-Convolutional neural network (R-CNN) have been extensively used to classify the remotely sensed data. Out of various deep learning algorithms, CNN-based algorithms are most widely used for the satellite image classification. Despite the improved performance of CNN, it also requires various hyper-parameters for training the network architecture to achieve the desired classification accuracy. Keeping in view the fact that the accuracy achieved by any classification algorithms is influenced by a suitable choice and value of hyper-parameter, this paper discusses the influence of several hyper-parameters on the classification accuracy of CNN classifier using three remote sensing datasets. The aim of this study is not to propose a set of values of different hyper-parameters but to study their influence on land cover classification accuracy with remote sensing datasets. Experimental results from the study indicate that various hyper-parameters affect the performance of CNN classifier to different extent suggesting a need to select the optimal value of these hyper-parameters for land cover classification studies using considered datasets. [ABSTRACT FROM AUTHOR]

Due to the different characteristics of satellite imagery, pan-sharpening has become an important field of remote sensing science. It is a reliable method to fuse the high-resolution panchromatic (Pan) image with the low-resolution multispectral (MS) images to generate a composite high-resolution MS (pan-sharpened) image. In this paper, we propose a robust pan-sharpening technique which combines the nonsubsampled contourlet transform (NSCT) with kernel principal component analysis (KPCA). An enhancement method is executed on the source MS image to retain the maximum spatial information present in the MS bands based on the intensity hue saturation (IHS) colour space transform. The proposed pan-sharpening technique is performed as follows: first, the source Pan and MS images are divided into high and low-pass coefficients by NSCT. Second, KPCA is used to propose an effective fusion rule for choosing appropriate low-pass NSCT coefficients for fusion. The high-pass coefficients are fused by proposing an enhanced sum modified Laplacian method (SML). Finally, the final pan-sharpened image is obtained by performing an inverse NSCT and inverse IHS transform on the fused low and high-pass coefficients. Four different groups of satellite image datasets are utilized in the experiments, which show that the proposed technique can both preserve the spatial details of the source images well and avoid spectral distortion. In addition, different image fusion quality metrics are adopted to evaluate the spectral and spatial qualities of the pan-sharpened image. Compared to the state-of-the-art pan-sharpening methods, the proposed technique achieved better performance in balancing spectral and spatial information and improved pan-sharpening results in terms of both visual quality and objective measurements. [ABSTRACT FROM AUTHOR]

Analysing old maps and correlating patterns with remote sensing (RS) images have been a useful technique for identifying features of cultural heritage. The present work has conducted a geospatial analysis of Mughal Agra by comparing historical maps and satellite imagery. This paper make two distinct contributions: (1) it expands existing findings on the riverfront Mughal city by identifying the position and footprint of all the riverfront complexes (by complex we mean property that have or had Mughal mansions, gardens or tombs) marked in a Mughal map of 1720, and (2) it makes original contribution by delineating the two-tiers of city walls and further attempts to comprehend the spatial transformations of the city since the Mughal period. This exploration resulted in recognizing traces of many lost historical built forms like city walls, gateways and riverfront complexes those of which have transformed, integrated, camouflaged or sometimes lost within the urban fabric of the present city of Agra. [ABSTRACT FROM AUTHOR]

Satellite image classification is the core of satellite image interpretation. The traditional shallow learning classification algorithm is difficult to extract remote sensing image features effectively, resulting in low classification accuracy. The deep neural network can automatically realize the feature analysis. However, due to the similarity of the optical properties of the cloud and snow, the existing deep learning model is difficult to obtain satisfactory results for the cloud and snow classification in the plateau. Meanwhile, too many parameters of deep learning lead to slow detection speed, which is difficult to meet the real-time requirements of detection. Furthermore, the deep learning method is prone to model degradation. In view of the existing problems, this paper proposes a dilated multi-scale cascade forest method to realize the classification of satellite cloud images. Multi-scale scanning increases the diversity of feature vector extraction and improves the separability of features. The dilated structure is designed to expand the feature receptive field, it improves the efficiency of feature extraction without losing the texture and spatial correlation information of the original image. In addition, the number of layers of network cascading does not need to be customized, it can be determined automatically according to the performance of the system. The proposed method effectively improves the accuracy of cloud/snow classification. Experiments on the imagery of HuanJing satellites A and B (HJ-1A/1B) in China demonstrate that the proposed model can get better classification results than existing studies. The time complexity of the proposed model is lower than existing works, hence it is practical in real-time and accurate remote sensing image classification. [ABSTRACT FROM AUTHOR]

Updating the physicochemical properties of surface water is important to water quality researchers and decision-makers. Routine surface water quality monitoring methods are inadequate; whereas, remote sensing (RS) is a source of processing data in both spatial and temporal domains. However, RS is a good tool of monitoring freshwater ecosystems, accurate modelling of surface water quality parameters from satellite imagery is still a challenging task. High correlation between the adjacent satellite signals, which creates redundancy, is one of the main challenges that negatively affect the stability of the developed models. Therefore, this paper focuses on the development of a novel approach for modelling concentrations of both turbidity (T) and total suspended solids (TSS) using Landsat-8 tasselled cap transformation (TCT). As a result, robust models were developed between measurements from satellite-derived T and TSS and in situ measurements with coefficient of determination values of 0.854 and 0.812, respectively. These findings confirmed the possibility of using TCT to establish strong models for assessing surface water quality due to its capability of transforming a set of correlated data onto a convenient set of uncorrelated data, which maintain the major information associated with a physical phenomenon. [ABSTRACT FROM AUTHOR]