Search

Coral reefs are an essential source of marine biodiversity, but they are declining at an alarming rate under the combined effects of global change and human pressure. A precise mapping of coral reef habitat with high spatial and time resolutions has become a necessary step for monitoring their health and evolution. This mapping can be achieved remotely thanks to satellite imagery coupled with machine-learning algorithms. In this paper, we review the different satellites used in recent literature, as well as the most common and efficient machine-learning methods. To account for the recent explosion of published research on coral reel mapping, we especially focus on the papers published between 2018 and 2020. Our review study indicates that object-based methods provide more accurate results than pixel-based ones, and that the most accurate methods are Support Vector Machine and Random Forest. We emphasize that the satellites with the highest spatial resolution provide the best images for benthic habitat mapping. We also highlight that preprocessing steps (water column correction, sunglint removal, etc.) and additional inputs (bathymetry data, aerial photographs, etc.) can significantly improve the mapping accuracy. [ABSTRACT FROM AUTHOR]

High dynamic range imaging (HDRI) is an essential task in remote sensing, enhancing low dynamic range (LDR) remote sensing images and benefiting downstream tasks, such as object detection and image segmentation. However, conventional frame-based HDRI methods may encounter challenges in real-world scenarios due to the limited information inherent in a single image captured by conventional cameras. In this paper, an event-based remote sensing HDR imaging framework is proposed to address this problem, denoted as ERS-HDRI, which reconstructs the remote sensing HDR image from a single-exposure LDR image and its concurrent event streams. The proposed ERS-HDRI leverages a coarse-to-fine framework, incorporating the event-based dynamic range enhancement (E-DRE) network and the gradient-enhanced HDR reconstruction (G-HDRR) network. Specifically, to efficiently achieve dynamic range fusion from different domains, the E-DRE network is designed to extract the dynamic range features from LDR frames and events and perform intra- and cross-attention operations to adaptively fuse multi-modal data. A denoise network and a dense feature fusion network are then employed for the generation of the coarse, clean HDR image. Then, the G-HDRR network, with its gradient enhancement module and multiscale fusion module, performs structure enforcement on the coarse HDR image and generates a fine informative HDR image. In addition, this work introduces a specialized hybrid imaging system and a novel, real-world event-based remote sensing HDRI dataset that contains aligned remote sensing LDR images, remote sensing HDR images, and concurrent event streams for evaluation. Comprehensive experiments have demonstrated the effectiveness of the proposed method. Specifically, it improves state-of-the-art PSNR by about 30% and the SSIM score by about 9% on the real-world dataset. [ABSTRACT FROM AUTHOR]

Mine pollution from mining activities is often widely recognised as a serious threat to public health, with mine solid waste causing problems such as tailings pond accumulation, which is considered the biggest hidden danger. The construction of tailings ponds not only causes land occupation and vegetation damage but also brings about potential environmental pollution, such as water and dust pollution, posing a health risk to nearby residents. If remote sensing images and machine learning techniques could be used to determine whether a tailings pond might have potential pollution and safety hazards, mainly monitoring tailings ponds that may have potential hazards, it would save a lot of effort in tailings ponds monitoring. Therefore, based on this background, this paper proposes to classify tailings ponds into two categories according to whether they are potentially risky or generally safe and to classify tailings ponds with remote sensing satellite images of tailings ponds using the DDN + ResNet-50 machine learning model based on ML.Net developed by Microsoft. In the discussion section, the paper introduces the environmental hazards of mine pollution and proposes the concept of "Healthy Mine" to provide development directions for mining companies and solutions to mine pollution and public health crises. Finally, we claim this paper serves as a guide to begin a conversation and to encourage experts, researchers and scholars to engage in the research field of mine solid waste pollution monitoring, assessment and treatment. [ABSTRACT FROM AUTHOR]

Optical satellite image change detection has attracted extensive research due to its comprehensive application in earth observation. Recently, deep learning (DL)-based methods have become dominant in change detection due to their outstanding performance. Remote sensing (RS) images contain different sizes of ground objects, so the information at different scales is crucial for change detection. However, the existing DL-based methods only employ summation or concatenation to aggregate several layers of features, lacking the semantic association of different layers. On the other hand, the UNet-like backbone is favored by deep learning algorithms, but the gradual downscaling and upscaling operation in the mainstream UNet-like backbone has the problem of misalignment, which further affects the accuracy of change detection. In this paper, we innovatively propose a hierarchical feature association and global correction network (HFA-GCN) for change detection. Specifically, a hierarchical feature association module is meticulously designed to model the correlation relationship among different scale features due to the redundant but complementary information among them. Moreover, a global correction module on Transformer is proposed to alleviate the feature misalignment in the UNet-like backbone, which, through feature reuse, extracts global information to reduce false alarms and missed alarms. Experiments were conducted on several publicly available databases, and the experimental results show the proposed method is superior to the existing state-of-the-art change detection models. [ABSTRACT FROM AUTHOR]

Motion blur recovery is a common method in the field of remote sensing image processing that can effectively improve the accuracy of detection and recognition. Among the existing motion blur recovery methods, the algorithms based on deep learning do not rely on a priori knowledge and, thus, have better generalizability. However, the existing deep learning algorithms usually suffer from feature misalignment, resulting in a high probability of missing details or errors in the recovered images. This paper proposes an end-to-end generative adversarial network (SDD-GAN) for single-image motion deblurring to address this problem and to optimize the recovery of blurred remote sensing images. Firstly, this paper applies a feature alignment module (FAFM) in the generator to learn the offset between feature maps to adjust the position of each sample in the convolution kernel and to align the feature maps according to the context; secondly, a feature importance selection module is introduced in the generator to adaptively filter the feature maps in the spatial and channel domains, preserving reliable details in the feature maps and improving the performance of the algorithm. In addition, this paper constructs a self-constructed remote sensing dataset (RSDATA) based on the mechanism of image blurring caused by the high-speed orbital motion of satellites. Comparative experiments are conducted on self-built remote sensing datasets and public datasets as well as on real remote sensing blurred images taken by an in-orbit satellite (CX-6(02)). The results show that the algorithm in this paper outperforms the comparison algorithm in terms of both quantitative evaluation and visual effects. [ABSTRACT FROM AUTHOR]

Introduction: Object detection in remotely sensed satellite images is critical to socio-economic, bio-physical, and environmental monitoring, necessary for the prevention of natural disasters such as flooding and fires, socio-economic service delivery, and general urban and rural planning and management. Whereas deep learning approaches have recently gained popularity in remotely sensed image analysis, they have been unable to efficiently detect image objects due to complex landscape heterogeneity, high inter-class similarity and intra-class diversity, and difficulty in acquiring suitable training data that represents the complexities, among others. Methods: To address these challenges, this study employed multi-object detection deep learning algorithms with a transfer learning approach on remotely sensed satellite imagery captured on a heterogeneous landscape. In the study, a new dataset of diverse features with five object classes collected from Google Earth Engine in various locations in southern KwaZulu-Natal province in South Africa was used to evaluate the models. The dataset images were characterized with objects that have varying sizes and resolutions. Five (5) object detection methods based on R-CNN and YOLO architectures were investigated via experiments on our newly created dataset. Conclusions: This paper provides a comprehensive performance evaluation and analysis of the recent deep learning-based object detection methods for detecting objects in high-resolution remote sensing satellite images. The models were also evaluated on two publicly available datasets: Visdron and PASCAL VOC2007. Results showed that the highest detection accuracy of the vegetation and swimming pool instances was more than 90%, and the fastest detection speed 0.2 ms was observed in YOLOv8. [ABSTRACT FROM AUTHOR]

Satellite image time series (SITS) classification is a challenging application concurrently driven by long-term, large-scale, and high spatial-resolution observations acquired by remote sensing satellites. The focus of current SITS classification research is to exploit the richness of temporal information in SITS data. In the literature, self-attention mechanism-based networks, which are capable of capturing global temporal attention, have achieved state-of-the-art results in SITS classification. However, these methods lack attention to local temporal information, which is also significant for SITS classification tasks. To explore the potential of different scales of temporal information in SITS data, a global–local temporal attention encoder (GL-TAE) is proposed in this paper. GL-TAE has two submodules set up in parallel, one of which is a lightweight temporal attention encoder (LTAE) for extracting global temporal attention and the other is lightweight convolution (LConv) for extracting local temporal attention. Compared with methods exploring global-only or local-only temporal features, the proposed GL-TAE can achieve better performance on two public SITS datasets, which proves the effectiveness of hybrid global–local temporal attention features. The experiments also demonstrate that GL-TAE is a lightweight model, which achieves the same performance as other models but with fewer parameters. [ABSTRACT FROM AUTHOR]

Global solar photovoltaic capacity has consistently doubled every 18 months over the last two decades, going from 0.3 GW in 2000 to 643 GW in 2019, and is forecast to reach 4240 GW by 2040. However, these numbers are uncertain, and virtually all reporting on deployments lacks a unified source of either information or validation. In this paper, we propose, optimize, and validate a deep learning framework to detect and map solar farms using a state-of-the-art semantic segmentation convolutional neural network applied to satellite imagery. As a final step in the pipeline, we propose a model to estimate the energy generation capacity of the detected solar energy facilities. Objectively, the deep learning model achieved highly competitive performance indicators, including a mean accuracy of 96.87%, and a Jaccard Index (intersection over union of classified pixels) score of 95.5%. Subjectively, it was found to detect spaces between panels producing a segmentation output at a sub-farm level that was better than human labeling. Finally, the detected areas and predicted generation capacities were validated against publicly available data to within an average error of 4.5% Deep learning applied specifically for the detection and mapping of solar farms is an active area of research, and this deep learning capacity evaluation pipeline is one of the first of its kind. We also share an original dataset of overhead solar farm satellite imagery comprising 23,000 images (256 × 256 pixels each), and the corresponding labels upon which the machine learning model was trained. [ABSTRACT FROM AUTHOR]

Grassland pastures are crucial for the global food supply through their milk and meat production; hence, forage species monitoring is essential for cattle feed. Therefore, knowledge of pasture above-ground canopy features help understand the crop status. This paper finds how to construct machine learning models to predict above-ground canopy features in Brachiaria pasture from ground truth data (GTD) and remote sensing at larger (satellite data on the cloud) and smaller (unmanned aerial vehicles (UAV)) scales. First, we used above-ground biomass (AGB) data obtained from Brachiaria to evaluate the relationship between vegetation indices (VIs) with the dry matter (DM). Next, the performance of machine learning algorithms was used for predicting AGB based on VIs obtained from ground truth and satellite and UAV imagery. When comparing more than twenty-five machine learning models using an Auto Machine Learning Python API, the results show that the best algorithms were the Huber with R2 = 0.60, Linear with R2 = 0.54, and Extra Trees with R2 = 0.45 to large scales using satellite. On the other hand, short-scale best regressions are K Neighbors with an R2 of 0.76, Extra Trees with an R2 of 0.75, and Bayesian Ridge with an R2 of 0.70, demonstrating a high potential to predict AGB and DM. This study is the first prediction model approach that assesses the rotational grazing system and pasture above-ground canopy features to predict the quality and quantity of cattle feed to support pasture management in Colombia. [ABSTRACT FROM AUTHOR]

Satellite-derived Chlorophyll-a (Chl-a) is an important environmental evaluation indicator for monitoring water environments. However, the available satellite images either have a coarse spatial or low spectral resolution, which restricts the applicability of Chl-a retrieval in coastal water (e.g., less than 1 km from the shoreline) for large- and medium-sized lakes/oceans. Considering Lake Chaohu as the study area, this paper proposes a physical-based spatial-spectral deep fusion network (PSSDFN) for Chl-a retrieval using Moderate Resolution Imaging Spectroradiometer (MODIS) and Sentinel-2 Multispectral Instrument (MSI) reflectance data. The PSSDFN combines residual connectivity and attention mechanisms to extract effective features, and introduces physical constraints, including spectral response functions and the physical degradation model, to reconcile spatial and spectral information. The fused and MSI data were used as input variables for collaborative retrieval, while only the MSI data were used as input variables for MSI retrieval. Combined with the Chl-a field data, a comparison between MSI and collaborative retrieval was conducted using four machine learning models. The results showed that collaborative retrieval can greatly improve the accuracy compared with MSI retrieval. This research illustrates that the PSSDFN can improve the estimated accuracy of Chl-a for coastal water (less than 1 km from the shoreline) in large- and medium-sized lakes/oceans. [ABSTRACT FROM AUTHOR]

At present, with the advance of satellite image processing technology, remote sensing images are becoming more widely used in real scenes. However, due to the limitations of current remote sensing imaging technology and the influence of the external environment, the resolution of remote sensing images often struggles to meet application requirements. In order to obtain high-resolution remote sensing images, image super-resolution methods are gradually being applied to the recovery and reconstruction of remote sensing images. The use of image super-resolution methods can overcome the current limitations of remote sensing image acquisition systems and acquisition environments, solving the problems of poor-quality remote sensing images, blurred regions of interest, and the requirement for high-efficiency image reconstruction, a research topic that is of significant relevance to image processing. In recent years, there has been tremendous progress made in image super-resolution methods, driven by the continuous development of deep learning algorithms. In this paper, we provide a comprehensive overview and analysis of deep-learning-based image super-resolution methods. Specifically, we first introduce the research background and details of image super-resolution techniques. Second, we present some important works on remote sensing image super-resolution, such as training and testing datasets, image quality and model performance evaluation methods, model design principles, related applications, etc. Finally, we point out some existing problems and future directions in the field of remote sensing image super-resolution. [ABSTRACT FROM AUTHOR]

To date, no method utilizing satellite imagery exists for detailing the locations and functions of critical infrastructure across the United States, making response to natural disasters and other events challenging due to complex infrastructural interdependencies. This paper presents a repeatable, transferable, and explainable method for critical infrastructure analysis and implementation of a robust model for critical infrastructure detection in satellite imagery. This model consists of a DenseNet-161 convolutional neural network, pretrained with the ImageNet database. The model was provided additional training with a custom dataset, containing nine infrastructure classes. The resultant analysis achieved an overall accuracy of 90%, with the highest accuracy for airports (97%), hydroelectric dams (96%), solar farms (94%), substations (91%), potable water tanks (93%), and hospitals (93%). Critical infrastructure types with relatively low accuracy are likely influenced by data commonality between similar infrastructure components for petroleum terminals (86%), water treatment plants (78%), and natural gas generation (78%). Local interpretable model-agnostic explanations (LIME) was integrated into the overall modeling pipeline to establish trust for users in critical infrastructure applications. The results demonstrate the effectiveness of a convolutional neural network approach for critical infrastructure identification, with higher than 90% accuracy in identifying six of the critical infrastructure facility types. [ABSTRACT FROM AUTHOR]

The COVID-19 pandemic has affected many aspects of human life around the world, due to its tremendous outcomes on public health and socio-economic activities. Policy makers have tried to develop efficient responses based on technologies and advanced pandemic control methodologies, to limit the wide spreading of the virus in urban areas. However, techniques such as social isolation and lockdown are short-term solutions that minimize the spread of the pandemic in cities and do not invert long-term issues that derive from climate change, air pollution and urban planning challenges that enhance the spreading ability. Thus, it seems crucial to understand what kind of factors assist or prevent the wide spreading of the virus. Although AI frameworks have a very efficient predictive ability as data-driven procedures, they often struggle to identify strong correlations among multidimensional data and provide robust explanations. In this paper, we propose the fusion of a heterogeneous, spatio-temporal dataset that combine data from eight European cities spanning from 1 January 2020 to 31 December 2021 and describe atmospheric, socio-economic, health, mobility and environmental factors all related to potential links with COVID-19. Remote sensing data are the key solution to monitor the availability on public green spaces between cities in the study period. So, we evaluate the benefits of NIR and RED bands of satellite images to calculate the NDVI and locate the percentage in vegetation cover on each city for each week of our 2-year study. This novel dataset is evaluated by a tree-based machine learning algorithm that utilizes ensemble learning and is trained to make robust predictions on daily cases and deaths. Comparisons with other machine learning techniques justify its robustness on the regression metrics RMSE and MAE. Furthermore, the explainable frameworks SHAP and LIME are utilized to locate potential positive or negative influence of the factors on global and local level, with respect to our model's predictive ability. A variation of SHAP, namely treeSHAP, is utilized for our tree-based algorithm to make fast and accurate explanations. [ABSTRACT FROM AUTHOR]

A variety of remote sensing applications call for automatic optical classification of satellite images. Recently, satellite missions, such as Sentinel-2, allow us to capture images in real-time of the Earth's scenario. The classification of this large amount of data requires increasingly precise and fast methods, which must take into account not only the spectral features dependence of each individual image but also that of the temporal ones. Copulas are an excellent statistical tool, able to model joint distributions between even random variables. In this paper, we propose a new approach for Satellite Image Time-Series (SITS) land cover classification, which combines the matrix factorization to reduce the dimensionality of the data and the use of copulas distribution to model the dependencies. We will show how the use of particular copulas can improve the accuracy of classification compared to the latest methodologies used for the classification task, such as those using Neural Networks. Experiments were conducted at a study site located on Reunion Island, using Sentinel-2 SITS data. Results are compared to those achieved by several approaches commonly used to address SITS-based land cover mapping and show that the use of copulas, in combination with the matrix factorization, achieved the highest classification yield compared to competing approaches. [ABSTRACT FROM AUTHOR]

Due to the country's rapid economic growth, the problem of air pollution in China is becoming increasingly serious. In order to achieve a win-win situation for the environment and urban development, the government has issued many policies to strengthen environmental protection. PM2.5 is the primary particulate matter in air pollution, so an accurate estimation of PM2.5 distribution is of great significance. Although previous studies have attempted to retrieve PM2.5 using geostatistical or aerosol remote sensing retrieval methods, the current rough resolution and accuracy remain as limitations of such methods. This paper proposes a fine-grained spatiotemporal PM2.5 retrieval method that comprehensively considers various datasets, such as Landsat 8 satellite images, ground monitoring station data, and socio-economic data, to explore the applicability of different machine learning algorithms in PM2.5 retrieval. Six typical algorithms were used to train the multi-dimensional elements in a series of experiments. The characteristics of retrieval accuracy in different scenarios were clarified mainly according to the validation index, R2. The random forest algorithm was shown to have the best numerical and PM2.5-based air-quality-category accuracy, with a cross-validated R2 of 0.86 and a category retrieval accuracy of 0.83, while both maintained excellent retrieval accuracy and achieved a high spatiotemporal resolution. Based on this retrieval model, we evaluated the PM2.5 distribution characteristics and hourly variation in the sample area, as well as the functions of different input variables in the model. The PM2.5 retrieval method proposed in this paper provides a new model for fine-grained PM2.5 concentration estimation to determine the distribution laws of air pollutants and thereby specify more effective measures to realize the high-quality development of the city. [ABSTRACT FROM AUTHOR]

In an effort to detect the region-of-interest (ROI) of remote sensing images with complex data distributions, sparse representation based on dictionary learning has been utilized, and has proved able to process high dimensional data adaptively and efficiently. In this paper, a visual attention model uniting hyperparameter sparse representation with energy distribution optimization is proposed for analyzing saliency and detecting ROIs in remote sensing images. A dictionary learning algorithm based on biological plausibility is adopted to generate the sparse feature space. This method only focuses on finite features, instead of various considerations of feature complexity and massive parameter tuning in other dictionary learning algorithms. In another portion of the model, aimed at obtaining the saliency map, the contribution of each feature is evaluated in a sparse feature space and the coding length of each feature is accumulated. Finally, we calculate the segmentation threshold using the saliency map and obtain the binary mask to separate the ROI from the original images. Experimental results show that the proposed model achieves better performance in saliency analysis and ROI detection for remote sensing images. [ABSTRACT FROM AUTHOR]

Given the continuous increase in the global population, the food manufacturers are advocated to either intensify the use of cropland or expand the farmland, making land cover and land usage dynamics mapping vital in the area of remote sensing. In this regard, identifying and classifying a high-resolution satellite imagery scene is a prime challenge. Several approaches have been proposed either by using static rule-based thresholds (with limitation of diversity) or neural network (with data-dependent limitations). This paper adopts the inductive approach to learning from surface reflectances. A manually labeled Sentinel-2 dataset was used to build a Machine Learning (ML) model for scene classification, distinguishing six classes (Water, Shadow, Cirrus, Cloud, Snow, and Other). This models was accessed and further compared to the European Space Agency (ESA) Sen2Cor package. The proposed ML model presents a Micro-F1 value of 0.84 , a considerable improvement when compared to the Sen2Cor corresponding performance of 0.59 . Focusing on the problem of optical satellite image scene classification, the main research contributions of this paper are: (a) an extended manually labeled Sentinel-2 database adding surface reflectance values to an existing dataset; (b) an ensemble-based and a Neural-Network-based ML models; (c) an evaluation of model sensitivity, biasness, and diverse ability in classifying multiple classes over different geographic Sentinel-2 imagery, and finally, (d) the benchmarking of the ML approach against the Sen2Cor package. [ABSTRACT FROM AUTHOR]

Urban functional area (UFA) recognition is one of the most important strategies for achieving sustainable city development. As remote-sensing and social-sensing data sources have increasingly become available, UFA recognition has received a significant amount of attention. Research on UFA recognition that uses a single dataset suffers from a low update frequency or low spatial resolution, while data fusion-based methods are limited in efficiency and accuracy. This paper proposes an integrated model to identify UFA using satellite images and taxi global positioning system (GPS) trajectories in four steps. First, blocks were generated as spatial units in the study area, and the spatiotemporal information entropy of the taxi GPS trajectory (STET) for each block was calculated. Second, a 24-hour time-frequency series was formed based on the pick-up and drop-off points extracted from taxi trajectories and used as the interpretation indicator of the blocks. The K-Means++ and k-Nearest Neighbor (kNN) algorithm were used to identify their social functions. Third, a multilabel classification method based on the residual neural network (MLC-ResNets) and "You Only Look Once" (YOLO) target detection algorithms were used to identify the features of the typical and atypical spatial textures, respectively, of the satellite images in the blocks. The confidence scores of the features of the blocks were categorized by the decision tree algorithm. Fourth, to find the best way to integrate the two sub-models for UFA identification, the 10-fold cross-validation method based on stratified random sampling was applied to determine the most optimal STET thresholds. The results showed that the average accuracy reached 82.0%, with an average kappa of 73.5%—significant improvements over most existing studies. This paper provides new insights into how the advantages of satellite images and taxi trajectories in UFA identification can be fully exploited to support sustainable city management. [ABSTRACT FROM AUTHOR]

Satellite image time series (SITS) is a sequence of satellite images that record a given area at several consecutive times. The aim of such sequences is to use not only spatial information but also the temporal dimension of the data, which is used for multiple real-world applications, such as classification, segmentation, anomaly detection, and prediction. Several traditional machine learning algorithms have been developed and successfully applied to time series for predictions. However, these methods have limitations in some situations, thus deep learning (DL) techniques have been introduced to achieve the best performance. Reviews of machine learning and DL methods for time series prediction problems have been conducted in previous studies. However, to the best of our knowledge, none of these surveys have addressed the specific case of works using DL techniques and satellite images as datasets for predictions. Therefore, this paper concentrates on the DL applications for SITS prediction, giving an overview of the main elements used to design and evaluate the predictive models, namely the architectures, data, optimization functions, and evaluation metrics. The reviewed DL-based models are divided into three categories, namely recurrent neural network-based models, hybrid models, and feed-forward-based models (convolutional neural networks and multi-layer perceptron). The main characteristics of satellite images and the major existing applications in the field of SITS prediction are also presented in this article. These applications include weather forecasting, precipitation nowcasting, spatio-temporal analysis, and missing data reconstruction. Finally, current limitations and proposed workable solutions related to the use of DL for SITS prediction are also highlighted. [ABSTRACT FROM AUTHOR]

Remote sensing satellite images in the optical domain often contain missing or misleading data due to overcast conditions or sensor malfunctioning, concealing potentially important information. In this paper, we apply expectation maximization (EM) Tucker to NDVI satellite data from the Iberian Peninsula in order to gap-fill missing information. EM Tucker belongs to a family of tensor decomposition methods that are known to offer a number of interesting properties, including the ability to directly analyze data stored in multidimensional arrays and to explicitly exploit their multiway structure, which is lost when traditional spatial-, temporal- and spectral-based methods are used. In order to evaluate the gap-filling accuracy of EM Tucker for NDVI images, we used three data sets based on advanced very-high resolution radiometer (AVHRR) imagery over the Iberian Peninsula with artificially added missing data as well as a data set originating from the Iberian Peninsula with natural missing data. The performance of EM Tucker was compared to a simple mean imputation, a spatio-temporal hybrid method, and an iterative method based on principal component analysis (PCA). In comparison, imputation of the missing data using EM Tucker consistently yielded the most accurate results across the three simulated data sets, with levels of missing data ranging from 10 to 90%. [ABSTRACT FROM AUTHOR]

Small mammals, and particularly rodents, are common inhabitants of farmlands, where they play key roles in the ecosystem, but when overabundant, they can be major pests, able to reduce crop production and farmers' incomes, with tangible effects on the achievement of Sustainable Development Goals no 2 (SDG2, Zero Hunger) of the United Nations. Farmers do not currently have a standardized, accurate method of detecting the presence, abundance, and locations of rodents in their fields, and hence do not have environmentally efficient methods of rodent control able to promote sustainable agriculture oriented to reduce the environmental impacts of cultivation. New developments in unmanned aerial system (UAS) platforms and sensor technology facilitate cost-effective data collection through simultaneous multimodal data collection approaches at very high spatial resolutions in environmental and agricultural contexts. Object detection from remote-sensing images has been an active research topic over the last decade. With recent increases in computational resources and data availability, deep learning-based object detection methods are beginning to play an important role in advancing remote-sensing commercial and scientific applications. However, the performance of current detectors on various UAS-based datasets, including multimodal spatial and physical datasets, remains limited in terms of small object detection. In particular, the ability to quickly detect small objects from a large observed scene (at field scale) is still an open question. In this paper, we compare the efficiencies of applying one- and two-stage detector models to a single UAS-based image and a processed (via Pix4D mapper photogrammetric program) UAS-based orthophoto product to detect rodent burrows, for agriculture/environmental applications as to support farmer activities in the achievements of SDG2. Our results indicate that the use of multimodal data from low-cost UASs within a self-training YOLOv3 model can provide relatively accurate and robust detection for small objects (mAP of 0.86 and an F1-score of 93.39%), and can deliver valuable insights for field management with high spatial precision able to reduce the environmental costs of crop production in the direction of precision agriculture management. [ABSTRACT FROM AUTHOR]

Building-damage mapping using remote sensing images plays a critical role in providing quick and accurate information for the first responders after major earthquakes. In recent years, there has been an increasing interest in generating post-earthquake building-damage maps automatically using different artificial intelligence (AI)-based frameworks. These frameworks in this domain are promising, yet not reliable for several reasons, including but not limited to the site-specific design of the methods, the lack of transparency in the AI-model, the lack of quality in the labelled image, and the use of irrelevant descriptor features in building the AI-model. Using explainable AI (XAI) can lead us to gain insight into identifying these limitations and therefore, to modify the training dataset and the model accordingly. This paper proposes the use of SHAP (Shapley additive explanation) to interpret the outputs of a multilayer perceptron (MLP)—a machine learning model—and analyse the impact of each feature descriptor included in the model for building-damage assessment to examine the reliability of the model. In this study, a post-event satellite image from the 2018 Palu earthquake was used. The results show that MLP can classify the collapsed and non-collapsed buildings with an overall accuracy of 84% after removing the redundant features. Further, spectral features are found to be more important than texture features in distinguishing the collapsed and non-collapsed buildings. Finally, we argue that constructing an explainable model would help to understand the model's decision to classify the buildings as collapsed and non-collapsed and open avenues to build a transferable AI model. [ABSTRACT FROM AUTHOR]

Usage of multispectral satellite imaging data opens vast possibilities for monitoring and quantitatively assessing properties or objects of interest on a global scale. Machine learning and computer vision (CV) approaches show themselves as promising tools for automatizing satellite image analysis. However, there are limitations in using CV for satellite data. Mainly, the crucial one is the amount of data available for model training. This paper presents a novel image augmentation approach called MixChannel that helps to address this limitation and improve the accuracy of solving segmentation and classification tasks with multispectral satellite images. The core idea is to utilize the fact that there is usually more than one image for each location in remote sensing tasks, and this extra data can be mixed to achieve the more robust performance of the trained models. The proposed approach substitutes some channels of the original training image with channels from other images of the exact location to mix auxiliary data. This augmentation technique preserves the spatial features of the original image and adds natural color variability with some probability. We also show an efficient algorithm to tune channel substitution probabilities. We report that the MixChannel image augmentation method provides a noticeable increase in performance of all the considered models in the studied forest types classification problem. [ABSTRACT FROM AUTHOR]

Every year, wildfires strike the Portuguese territory and are a concern for public entities and the population. To prevent a wildfire progression and minimize its impact, Fuel Management Zones (FMZs) have been stipulated, by law, around buildings, settlements, along national roads, and other infrastructures. FMZs require monitoring of the vegetation condition to promptly proceed with the maintenance and cleaning of these zones. To improve FMZ monitoring, this paper proposes the use of satellite images, such as the Sentinel-1 and Sentinel-2, along with vegetation indices and extracted temporal characteristics (max, min, mean and standard deviation) associated with the vegetation within and outside the FMZs and to determine if they were treated. These characteristics feed machine-learning algorithms, such as XGBoost, Support Vector Machines, K-nearest neighbors and Random Forest. The results show that it is possible to detect an intervention in an FMZ with high accuracy, namely with an F1-score ranging from 90% up to 94% and a Kappa ranging from 0.80 up to 0.89. [ABSTRACT FROM AUTHOR]

Modern convolutional neural networks (CNNs) are often trained on pre-set data sets with a fixed size. As for the large-scale applications of satellite images, for example, global or regional mappings, these images are collected incrementally by multiple stages in general. In other words, the sizes of training datasets might be increased for the tasks of mapping rather than be fixed beforehand. In this paper, we present a novel algorithm, called GeoBoost, for the incremental-learning tasks of semantic segmentation via convolutional neural networks. Specifically, the GeoBoost algorithm is trained in an end-to-end manner on the newly available data, and it does not decrease the performance of previously trained models. The effectiveness of the GeoBoost algorithm is verified on the large-scale data set of DREAM-B. This method avoids the need for training on the enlarged data set from scratch and would become more effective along with more available data. [ABSTRACT FROM AUTHOR]

Precision agriculture is considered to be a fundamental approach in pursuing a low-input, high-efficiency, and sustainable kind of agriculture when performing site-specific management practices. To achieve this objective, a reliable and updated description of the local status of crops is required. Remote sensing, and in particular satellite-based imagery, proved to be a valuable tool in crop mapping, monitoring, and diseases assessment. However, freely available satellite imagery with low or moderate resolutions showed some limits in specific agricultural applications, e.g., where crops are grown by rows. Indeed, in this framework, the satellite's output could be biased by intra-row covering, giving inaccurate information about crop status. This paper presents a novel satellite imagery refinement framework, based on a deep learning technique which exploits information properly derived from high resolution images acquired by unmanned aerial vehicle (UAV) airborne multispectral sensors. To train the convolutional neural network, only a single UAV-driven dataset is required, making the proposed approach simple and cost-effective. A vineyard in Serralunga d'Alba (Northern Italy) was chosen as a case study for validation purposes. Refined satellite-driven normalized difference vegetation index (NDVI) maps, acquired in four different periods during the vine growing season, were shown to better describe crop status with respect to raw datasets by correlation analysis and ANOVA. In addition, using a K-means based classifier, 3-class vineyard vigor maps were profitably derived from the NDVI maps, which are a valuable tool for growers. [ABSTRACT FROM AUTHOR]

In this paper, we present a state-of-the-art precipitation estimation framework which leverages advances in satellite remote sensing as well as Deep Learning (DL). The framework takes advantage of the improvements in spatial, spectral and temporal resolutions of the Advanced Baseline Imager (ABI) onboard the GOES-16 platform along with elevation information to improve the precipitation estimates. The procedure begins by first deriving a Rain/No Rain (R/NR) binary mask through classification of the pixels and then applying regression to estimate the amount of rainfall for rainy pixels. A Fully Convolutional Network is used as a regressor to predict precipitation estimates. The network is trained using the non-saturating conditional Generative Adversarial Network (cGAN) and Mean Squared Error (MSE) loss terms to generate results that better learn the complex distribution of precipitation in the observed data. Common verification metrics such as Probability Of Detection (POD), False Alarm Ratio (FAR), Critical Success Index (CSI), Bias, Correlation and MSE are used to evaluate the accuracy of both R/NR classification and real-valued precipitation estimates. Statistics and visualizations of the evaluation measures show improvements in the precipitation retrieval accuracy in the proposed framework compared to the baseline models trained using conventional MSE loss terms. This framework is proposed as an augmentation for PERSIANN-CCS (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Network- Cloud Classification System) algorithm for estimating global precipitation. [ABSTRACT FROM AUTHOR]

Road network extraction from remote sensing images has played an important role in various areas. However, due to complex imaging conditions and terrain factors, such as occlusion and shades, it is very challenging to extract road networks with complete topology structures. In this paper, we propose a learning-based road network extraction framework via a Multi-supervised Generative Adversarial Network (MsGAN), which is jointly trained by the spectral and topology features of the road network. Such a design makes the network capable of learning how to "guess" the aberrant road cases, which is caused by occlusion and shadow, based on the relationship between the road region and centerline; thus, it is able to provide a road network with integrated topology. Additionally, we also present a sample quality measurement to efficiently generate a large number of training samples with a little human interaction. Through the experiments on images from various satellites and the comprehensive comparisons to state-of-the-art approaches on the public datasets, it is demonstrated that the proposed method is able to provide high-quality results, especially for the completeness of the road network. [ABSTRACT FROM AUTHOR]