Methane emissions are a major contributor to global warming, making the detection and monitoring of anthropogenic methane sources, such as oil and gas operations and landfills, a critical task. Remote sensing using hyperspectral sensors has proven to be a valuable tool for this purpose. However, traditional methane detection algorithms based on matched filters often produce spurious results and require laborious post-processing to accurately identify actual plumes. In this paper, we propose a novel approach for estimating the target signature of a matched filter as a learnable layer integrated into a user-selected segmentation model, trainable in an end-to-end manner. By considering the signature as a parameter of the segmentation model, our methodology enables the signature to adapt both to the task and the user-selected segmentation approach. We use hyperspectral images from the EMIT instrument and their associated publicly available methane plumes to train deep learning segmentation models. Our results show consistent improvements in both convolutional neural networks (CNNs) and transformer architectures, as demonstrated by the F1 segmentation score increasing by approximately 18% compared to the baseline values. This allows for more accurate and automated plume segmentation, ultimately aiding in the identification of methane leaks from point source emitters.

Maximizing carbon sequestration in trees across different ecoregions has the potential to support carbon markets while improving forest restoration and preserving biodiversity. Tree height is a predictor of carbon stored in its biomass, but estimating tree height using publicly available remote sensing datasets remains challenging. Artificial intelligence can play an important role in improving estimates of vegetation canopy height, especially in regions with limited local measurements. This study compares a transformer-based Geospatial Foundation Model (GFM) and a baseline deep learning model (U-Net) for predicting tree canopy height across Kenya’s diverse ecoregions. The models use cloudfree mosaics from the Harmonized Landsat and Sentinel-2 (HLS) product as predictors and space-borne GEDI laser data for canopy height reference. Both models had similar root mean square error (RMSE) scores: GFM at 6.05 m and UNet at 5.80 m for the most prevalent small to medium-sized trees. In a second experiment, the models trained in Kenya were applied to a Mozambique study area. In this challenging set-up, GFM generalized better to the different ecoregions.

Methane emissions from oil and gas infrastructure, wetlands, and livestock contribute to the greenhouse gas inventory. The analysis of satellite short-wave infrared imagery offers opportunities for screening large areas to detect methane leaks. Deep learning algorithms excel at analyzing these data, however, they require large annotated datasets for model calibration that are difficult to get. To overcome this limitation, we explore a methodology to spot methane plumes using deep binary classifiers trained on a large dataset of synthetically created methane plumes, customized for this specific task, using publicly available images of the Sentine1-2 satellites. To build the database, we simulate plume patterns using the Hybrid Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT) and use a simple stochastic model to account for reflectance attenuation due to methane in band 12 centered at 2190 nm. To help distinguish methane plumes from the image background, we compute a methane signature image based on a background subtraction technique. Once calibrated, the classification model is applied to image patches centered in the local minima of the methane signature within the satellite image, scoring a value ranging from 0 to 1 associated with the presence of a methane plume. We compare experimentally the general-purpose ResNet architecture and MethaNet, a domain-specific convolutional neural network, using simulated data. Then, we evaluate the feasibility of our approach in detecting large methane leaks at two study sites located in the Hassi Messaoud oil field in Algeria and the Permian Basin in the US, each covering an area of 0.25$\times$ 0.25 degrees. We found that ResNet is effective in identifying large, known methane plumes that were set aside for testing purposes. This method could be considered as a component of a solution for planning mitigation activities.

Maintaining and, ultimately, increasing vegetation coverage is likely the most impactful approach to globally capture carbon. Biomass is a crucial parameter for quantifying carbon stored in vegetation, and estimating it poses challenges as statistical models need to be customized to specific biomes. This study compares the prediction of aboveground biomass using various regression methods that were locally fitted in three distinct study sites located in Texas and Louisiana, USA. These sites (biomes) had average aboveground biomass densities of 4.1, 17.3, and 94.6 Mg/ha. The predictions obtained from these localized models were then compared to those derived from a general model that pooled data from all three sites together. Optical and radar imagery acquired from Sentinel satellites were used as predictors, while biomass density from GEDI served as the reference. In most experiments, Random Forest scored best, and the results indicate that the biome-specific models exhibited slightly higher accuracy. Specifically, the root mean square error (RMSE) values for the biome-specific models were 8.8, 16.8, and 54.8 Mg/ha, respectively. In comparison, the general model exhibited approximately 1 Mg/ha higher RMSE. The results indicate that the locally fitted models tailored to specific biomes generally outperformed the general model tested.

Floods cause loss of lives and multi-billion dollar damages every year. When these events strike, automated tools using remote sensing for quick mapping of the affected areas are critical for planning rescue activities and assessing impact. The most recent techniques to map floods are based on semantic segmentation and deep learning, which require large datasets and ground truth for training, that are difficult to get. To overcome this challenge, we propose an effective method for synthesizing patches of synthetic aperture radar images containing open-land flooded areas. With bi-temporal image acquisitions, we replace portions of land areas in the second acquisition with water pixels borrowed from permanent water bodies. Spatial patterns derived from elevation data help guide the process. With this approach, we build a large dataset with pre and post-event Sentinel-1-based VH-polarized radar intensity images to train deep neural networks to map real-life floods. In a case study, we employ an established U-Net architecture and show that a model version with less than 2% of the number of original parameters achieves almost identical flood detection accuracy. This allows for faster processing and is a clear advantage from an operational perspective. For comparison, we provide empirical segmentation results of four flood cases. F1 scores agree between 0.80 and 0.90 when compared to reference flood maps from the Copernicus service. This confirms that the approach is valuable for mapping areas affected by floods, e.g. during or immediately after catastrophic events, even in areas that were not included during the model training.

Floods are among the most costly weather hazards for societies and businesses globally. With increasing global warming, these events have become even more frequent and more devastating. Thus, accurate flood mapping has become critical for disaster relief, risk management and mitigation. Current flood segmentation methods use either threshold-based approaches or deep-learning schemes, e.g. using the U-Net architecture, to differentiate between water-covered bodies or dry land on Earth observation images. Many schemes are exploiting imagery from synthetic aperture radar (e.g. Sentinel 1 satellites) or visual bands of satellites such as the Sentinel 2, but often restrict themselves to using one or very few modalities, i.e. spectral wavelengths, despite the availability of many more wavelengths or pre-processed indices with potential value to the challenge. In support of operationalizing flood segmentation on a global scale using deep learning, we propose semantic flood segmentation exploiting optionally many different modalities (i.e. multimodal flood segmentation), making the approach largely immune to geographic differences across the globe. Using U-Net at the core of our work, we observe very good generalisation of our segmentation model to unseen flood events in our holdout set at the level of 0.95 F1 Score (0.92 IoU) for both no water and water class, and 0.53 F1 Score (0.43 IoU) for water class, respectively.

Surface water mapping in satellite images enables flood monitoring, a task with increasing importance under changing climate conditions. Current segmentation methods based on Deep Learning require large, curated datasets for training’ which are difficult to obtain. In this paper, we present a probabilistic approach to water segmentation based on established computer vision methods that requires little training and is easy to interpret. We use prior knowledge of typical backscatter intensity to locate seed pixels likely to be in water and land. A preliminary rough segmentation using thresholding guides the selection of two image patches that will be fully labeled using seeded region growing segmentation. Then, we sample small patches within the automatically labeled regions to train a fully connected neural network that, running in sliding windows, scores for the presence of water in the entire image. The approach is tested for mapping surface water during a large flood event in Aude, France. Outputs of the proposed approach are compared to the reference flood delineation map provided online by the Copernicus service. A F1 score of 0.67 suggests that performance of the proposed approach is similar or better than classic thresholding methods used as benchmark.

Unmanned aerial vehicles (UAVs) allow on-demand imaging of orchards at an unprecedented level of detail. The automated detection of plantation rows in the images helps in the successive analysis steps, such as the detection of individual fruit trees and planting gaps, aiding producers with inventory and planting operations. Citrus trees can be planted in curved rows that form intricate geometric patterns in aerial images, requiring robust detection approaches. While deep learning methods rank among state-of-the-art methods for segmenting images with particular geometrical patterns, they struggle to hold their performance when testing data differs much from training data (e.g., image intensity differences, image artifacts, vegetation characteristics, and landscape conditions). In this letter, we propose a method to learn geometric features of orchards in UAV images and use them to improve the detection of plantation rows. First, we train a detection encoder–decoder network (DetED) to segment planting rows in RGB images. Then, with labeled data, we train an encoder–decoder correction network (CorrED) that learns to map binary masks with spurious row segmentation geometries into corrected ones. Finally, we use the CorrED network to fix geometric inconsistencies in DetED outcome. Our experiments with commercial plantations of orange trees show that the proposed CorrED postprocessing can restore missing segments of plantation rows and improve detection accuracy in testing data.

This work addresses the problem of artificially extending satellite-derived soil moisture data using soil moisture estimates generated by a land surface model. We calibrate regression algorithms in a set of spatially and temporally coincident surface soil moisture estimates derived from the coarse Soil Moisture Active Passive (SMAP) radiometer and soil moisture simulated by the Global Land Data Assimilation System (GLDAS) Noah model, which assimilates atmospheric forcing and ancillary data. Once calibrated, we apply the regression model to the GLDAS-Noah soil moisture, and ancillary data, to estimate the soil moisture that would have been observed if SMAP had been available on a given past date. We explore the feasibility of the approach in a study area of size$12\times 12$degrees located in Southern Brazil. The Random Forests and XGBoost regression algorithms show reasonable reconstruction skills over a 642-day hindcast period$(r^{2}=0.84,\text{RMSE}=0.051\mathrm{m}^{3}/\mathrm{m}^{3})$. These results suggest the approach is worth further investigation.

Palm oil is the largest vegetable oil in the world in terms of produced volume, and 75% of global production is used for food and cooking purposes. Sustainable management of the producing areas calls for the frequent assessment of field conditions. In this paper, we investigate an automatic algorithm based on deep learning that is capable to build an inventory of individual oil-palm trees using aereal color images collected by unmanned aerial vehicles. The idea consists of combining the outputs of two independent convolutional neural networks, trained on partially distinct subsets of samples and different spatial scales to capture coarse and fine details of image patches. The estimated posterior probabilities are combined by simple averaging as to improve detection accuracy and estimate the confidence for each individual detection. Non-maxima suppression removes weak detections. Experiments at three commercial oil-palm tree plantations sites aged two, four, and 16 years in Northern Brazil revealed overall detection accuracies in the range 91.2-98.8% using orthomosaics of decimeter spatial resolution. The proposed approach can be a useful component of a forest monitoring system based on remote sensing.

This paper presents a new paradigm for object based classification of multispectral images. Instead of classifying objects only after the segmentation process is completed, it is proposed to intercept the early stages of the segmentation by iteratively performing classification tests to under growing regions. By applying this simultaneous analysis, mislabeling of objects considered only after segmentation is completely done can be avoided. The proposed technique assumes that some growing regions can present higher membership to a particular class when comparing to the final object in which it is included. A Bayesian framework was applied in classification tests performed by pixel based, traditional object based, and the proposed technique were performed. The results show the soundness of the proposed method when comparing overall accuracies with a reference map.

This paper aims to use unique features of hyperspectral data on an automatic process for outlining individual tree crowns (ITCs) in a tropical forest area, with special focus on semi-deciduous species. In order to enhance biophysical and biochemical properties of canopy species, a set of vegetation indices were computed. These indices served as input for a region growing segmentation algorithm that takes into account mutual similarity of pixels and spectral separability between neighbor segments. Segmentation output was evaluated on the basis of a score computed with the proportion of the area of the segments located within manually delineated ITCs. Results show that the segmentation approach is able to automatically delineate up to 70% of the control ITCs.

This paper presents an alternative object based classification for multispectral remote sensing images. Instead of classifying the images after the segmentation process, it is suggested to involve some steps of objects recognition during the segmentation process in order to improve the final classification results. The methodology is based on the statistical distribution of object classes. Experiments were performed with a TM-Landsat image and the results were compared with a reference data. The results indicate the soundness of the proposed methodology.

This paper develops a noise-robust spatial preprocessing module which can be used prior to spectral unmixing of remotely sensed hyperspectral images. The method first derives a spatial homogeneity index which is relatively insensitive to the noise present in the original hyperspectral data. Then, it fuses this index with a spectral-based classification, obtaining a set of pure regions which are used to guide the unmixing process. An experimental comparison of the proposed method with other spatial-spectral unmixing approaches is conducted using both synthetic and real hyperspectral data collected by the Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS). Our experiments indicate that spectral unmixing can benefit from the proposed pre-processing approach, in particular, when the noise level present in the original hypespectral scene is relatively high.

Supervised classification of hyperspectral images is a very challenging task due to the generally unfavorable ratio between the number of spectral bands and the number of training samples available a priori, which results in the Hughes phenomenon. For this purpose, several feature extraction methods have been investigated in order to reduce the dimensionality of the data to the right subspace without significant loss of the original information that allows for the separation of classes. In this letter, we explore the use of spectral unmixing for feature extraction prior to supervised classification of hyperspectral data using support vector machines. The proposed feature extraction strategy has been implemented in the form of four different unmixing chains and evaluated using two different scenes collected by National Aeronautics and Space Administration Jet Propulsion Laboratory's Airborne Visible/Infrared Imaging Spectrometer. The experiments suggest competitive results but also show that the definition of the unmixing chains plays an important role in the final classification accuracy. Moreover, differently from most feature extraction techniques available in the literature, the features obtained using linear spectral unmixing are potentially easier to interpret due to their physical meaning.

Hyperspectral data lossy compression has not yet achieved global acceptance in the remote sensing community, mainly because it is generally perceived that using compressed images may affect the results of posterior processing stages. This possible negative effect, however, has not been accurately characterized so far. In this letter, we quantify the impact of lossy compression on two standard approaches for hyperspectral data exploitation: spectral unmixing, and supervised classification using support vector machines. Our experimental assessment reveals that different stages of the linear spectral unmixing chain exhibit different sensitivities to lossy data compression. We have also observed that, for certain compression techniques, a higher compression ratio may lead to more accurate classification results. Even though these results may seem counterintuitive, this work explains these observations in light of the spatial regularization and/or whitening that most compression techniques perform and further provides recommendations on best practices when applying lossy compression prior to hyperspectral data classification and/or unmixing.

In this paper, we investigate the use of the watershed transformation for integrating spatial and spectral information in the process of endmember extraction for spectral unmixing of hyperspectral images. The proposed approach is presented as a preprocessing module designed to automatically select a small subset of pixels containing potentially relevant candidates from both spatial and spectral point of view. Dimensionality reduction is required. The idea is to use the morphological watershed transformation to guide the endmember searching process to spatially homogeneous and spectrally “purer” areas. Here the main assumption is that such areas can be located at the local minima of the catchment basins, and far away from watershed lines that define the transition areas between different regions, expected to contain mixed pixels. Experimental results, conducted using a database of 28 simulated hyperspectral data sets obtained through manipulation of a real hyperspectral image acquired by the Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) over a mixed scenario including agricultural, vegetation, and urban areas, suggests a promising trade-off between percentage of endmember candidates retained and degree of spectral purity of predominant endmembers.

Over the last decade, several algorithms have been developed for automatic or semi-automatic extraction of spectral endmembers from hyperspectral image data. In this paper, we present a thorough analytical comparison of endmember extraction methods which include spatial information in the search of spectral endmembers versus a few algorithms which are exclusively based on spectral information. Our quantitative and comparative assessment of algorithm accuracy and computational performance, conducted using both synthetic and real hyperspectral data, provides interesting findings about the potential benefits that can be obtained after incorporating spatial information into the design of endmember extraction algorithms.

The N-FINDR algorithm is one of the most widely used and successfully applied methods for automatically determining endmembers in hyperspectral image data without usinga prioriinformation. The algorithm attempts to automatically find the simplex of maximum volume that can be inscribed within the hyperspectral data set. Due to the intrinsic complexity of remotely sensed scenes, the final volume-based solution provided by N-FINDR may be not the global maximum. In addition, the final results provided by the algorithm are typically dependent of its initialization. In this letter, we explore the aforementioned issues and conduct a quantitative and comparative analysis of different (available and new) strategies for the implementation of N-FINDR. Our experimental evaluation and comparison are conducted using two well-known hyperspectral scenes collected by the National Aeronautics and Space Administration Jet Propulsion Laboratory's Airborne Visible Infrared Imaging Spectrometer.

Endmember extraction is the process of selecting a collection of pure signature spectra of the materials present in a remotely sensed hyperspectral scene. These pure signatures are then used to decompose the scene into abundance fractions by means of a spectral unmixing algorithm. Most techniques available in the endmember extraction literature rely on exploiting the spectral properties of the data alone. As a result, the search for endmembers in a scene is conducted by treating the data as a collection of spectral measurements with no spatial arrangement. In this paper, we propose a novel strategy to incorporate spatial information into the traditional spectral-based endmember search process. Specifically, we propose to estimate, for each pixel vector, a scalar spatially derived factor that relates to the spectral similarity of pixels lying within a certain spatial neighborhood. This scalar value is then used to weigh the importance of the spectral information associated to each pixel in terms of its spatial context. Two key aspects of the proposed methodology are given as follows: 1) No modification of existing image spectral-based endmember extraction methods is necessary in order to apply the proposed approach. 2) The proposed preprocessing method enhances the search for image spectral endmembers in spatially homogeneous areas. Our experimental results, which were obtained using both synthetic and real hyperspectral data sets, indicate that the spectral endmembers obtained after spatial preprocessing can be used to accurately model the original hyperspectral scene using a linear mixture model. The proposed approach is suitable for jointly combining spectral and spatial information when searching for image-derived endmembers in highly representative hyperspectral image data sets.

We study a novel ensemble method as a supervised tool for the accurate classification of optical high-resolution imagery. The method uses partially optimized Support Vector Machines as basis classifier and a simple random mechanism, inspired on Random Forests, to promote diversity and include spatial information into the ensemble. Experimental results on an IKONOS image are compared with those from well-known classification methods, including spectral, contextual, and ensemble based techniques. The best results have been achieved, in both the classification accuracy and visual quality of the classification map, with the use of the proposed ensemble method.

High-dimensional image data open new possibilities in remote sensing digital image classification, particularly when dealing with classes that are spectrally very similar. The main problem refers to the estimation of a large number of classifier's parameters. One possible solution to this problem consists in reducing the dimensionality of the original data without a significant loss of information. In this letter, a new approach to reduce data dimensionality is proposed. In the proposed methodology, each pixel's curve of spectral response is initially segmented, and the digital numbers (DNs) at each segment are replaced by a smaller number of statistics. In this letter, the proposed statistics are the mean and variance of the segment's DNs, which are supposed to carry information about the segment's position and shape, respectively. Tests were performed by using Airborne Visible/Infrared Imaging Spectrometer hyperspectral image data. The experiments have shown that this methodology is capable of providing very acceptable results, in addition of being computationally efficient

In this paper we focus on the algorithm development allowing an improvement of land surface temperature (LST) estimates obtained from passive remote sensing images. An innovative regression algorithm is presented and results are compared with the main classical algorithms for LST estimation: the Split Window techniques. Specifically, a functional approximation scheme based on Support Vector Machines (SVMs) will be adopted and applied to AVHRR data. Particular attention will be devoted to the optimization of the SVM estimator, focusing on the selection of the employed kernel functions (linear, RBF gaussian, and hyperbolic tangent "tanh"). Results, suggesting the effectiveness of the proposed SVM-based approach, are presented and discussed.

In the present study, we propose a new simple approach to reduce the data dimensionality in hyperspectral image data. The basic assumption here consists in assuming that a pixel's curve of spectral response, as defined in the spectral space by the recorded digital numbers (DNs) at the available spectral bands, can be segmented, and each segment can be replaced by a smaller number of statistics: mean and variance, describing the main characteristics of a pixel's spectral response. It is expected that this procedure can be accomplished without significant loss of information. The DNs at even spectral band are used to calculate a few statistics that would be used instead of the DNs themselves in the classification process. For the pixel's spectral curve segmentation, we propose tree sub-optimal algorithms that are easy to implement and also computationally efficient. Using a top-down strategy, the original pixel's spectral curve is sequentially segmented. Experiments using a parametric classifier are performed on an AVIRIS data set. Encouraging results have been obtained in terms of classification accuracy and execution time, suggesting the effectiveness of the proposed algorithms. The results suggest that the proposed algorithms can be faster and achieve a better accuracy than the classical Sequential Forward Selection (SFS) technique, known from literature as one of the simplest and fastest techniques for data dimensionality reduction using the feature selection approach.