Large-area Mapping Research Articles

Band Weight-Optimized BiGRU Model for Large-Area Bathymetry Inversion Using Satellite Images

Currently, using satellite images combined with deep learning models has become an efficient approach for bathymetry inversion. However, only limited bands are usually used for bathymetry inversion in most methods, and they rarely applied for large-area bathymetry inversion (it is important for methods to be used in operational environments). Aiming to utilize all band information of satellite optical image data, this paper first proposes the Band Weight-Optimized Bidirectional Gated Recurrent Unit (BWO_BiGRU) model for bathymetry inversion. To further improve the accuracy, the Stumpf model is incorporated into the BWO_BiGRU model to form another new model—Band Weight-Optimized and Stumpf’s Bidirectional Gated Recurrent Unit (BWOS_BiGRU). In addition, using RANSAC to accurately extract in situ water depth points from the ICESat-2 dataset can accelerate computation speed and improve convergence efficiency compared to DBSCAN. This study was conducted in the eastern bay of Shark Bay, Australia, covering an extensive shallow-water area of 1725 km2. A series of experiments were performed using Stumpf, Band-Optimized Bidirectional LSTM (BoBiLSTM), BWO_BiGRU, and BWOS_BiGRU models to infer bathymetry from EnMAP, Sentinel-2, and Landsat 9 satellite images. The results show that when using EnMAP hyperspectral images, the bathymetry inversion of BWO_BiGRU and BWOS_BiGRU models outperform Stumpf and BoBiLSTM models, with RMSEs of 0.64 m and 0.63 m, respectively. Additionally, the BWOS_BiGRU model is particularly effective in nearshore water areas (depth between 0 and 5 m) of multispectral images. In general, comparing to multispectral satellite images, using the proposed BWO_BiGRU model to infer hyperspectral satellite images can achieve better bathymetry inversion results for large-area bathymetry maps.

The early development of a combined micro- and full-field X-ray fluorescence analysis system using white X-rays at PLS-II.

X-ray fluorescence (XRF) is widely used to analyze elemental distributions in samples. Micro-XRF (µ-XRF), the most basic conventional XRF technique, offers good spatial resolution through precise 2D scanning with a micrometre-sized X-ray source. Recently, synchrotron based XRF analysis platforms have achieved nano-XRF with highly focused X-rays using polycapillary optics or mirrors, leveraging the excellent coherence of synchrotron radiation. However, XRF techniques are hindered by long data acquisition times (exceeding several hours) due to their point-by-point scanning approach, impeding large-area elemental mapping. Full-field XRF (FF-XRF), developed in the 2010s and based on the high brilliance of synchrotron X-rays, enables significantly shorter (less than a few minutes) data acquisition times via single-exposure imaging using a 2D X-ray detector. Nevertheless, it is constrained by relatively low spatial resolution and sensitivity. Hence, a new XRF platform is required to accommodate resolution demands to cover diverse experimental purposes. In this study, we developed a preliminary model of a novel XRF system that combines micro- and full-field XRF setups to address these limitations. This system allows easy mode switching while maintaining the region of interest of the imaging system within a single apparatus, simply by rotating the sample to face either detector depending on research purposes. We anticipate that this new XRF system will be widely utilized in various research fields as the initial XRF setup at Pohang Light Source-II.

Synergistic image and point cloud processing of UAV data for urban flood modeling: point cloud smart thinning and curb mapping

Abstract. We propose an integrated approach for automatic point cloud thinning and curb mapping in Uncrewed Aerial Vehicle - Structure from Motion (UAV-SfM) point clouds to enhance hydrological modeling in flood-prone urban areas. UAV flights were conducted to generate an initial orthoimage, which was used to train a convolutional neural network (CNN) segmentation model. The trained model was then applied to the UAV images to produce two binary mask sets: one for vegetation and one for streets and sidewalks. These masks were incorporated during photogrammetric 3D reconstruction to estimate camera geometry and generate a dense point cloud. Our results show that vegetation masks did not improve camera geometry estimation. However, by applying UAV masks, we achieved a 15% reduction in total processing time and decreased the number of points by a factor of 2.7. This targeted approach enabled curb detection by focusing on expected curb locations. Curb candidate points were proposed using geometric characteristics of the point cloud, including normal values, linearity, and verticality. Our rule-based method effectively mapped even subtle curb features, providing a rapid, cost-effective solution for large-area curb mapping. Further, we explored the potential of random forest for curb mapping, with promising results. Our approach can support urban flood modeling efforts and strengthen urban resilience for flood-prone communities.

Rapid and precise large area mapping of rare-earth doping homogeneity in luminescent materials

Doping of luminescent materials by rare-earth ions is common practice to achieve desired emission properties for a large variety of applications. As several rare-earths ions are frequently combined, it is subsequently difficult to effectively detect and control their homogeneous distribution within the host material. Here, we present a simple, rapid, large scale and precise method of rare-earth mapping using a commercial UV-Vis scanner. We discuss the influence of rare-earth distribution on the physical, optical and luminescent properties with no observable qualitative effect on photoluminescent properties and optical anisotropy. On the contrary, rare-earth-rich areas exhibit significantly higher values of refractive index and optical absorption, which allowed for their identification by the commercial scanner device. The presented method thus provides fast and accurate information about the rare-earth distribution in the material volume with high resolution (≈2.7 µm) and low limit of concentration difference detection (≈0.014 at.%) compared to other techniques, which makes it a promising candidate for high throughput measurements.

Prediction and spatial–temporal changes of soil organic matter in the Huanghuaihai Plain by combining legacy and recent data

Soil organic matter (SOM) is critical for soil fertility, crop growth, and plays an important role in the global carbon cycle and climate change. Therefore, spatial prediction of SOM is important to rational soil resource utilization, agricultural production, and ecological environment management. However, large-area SOM mapping research heavily relies on legacy soil data, and large-scale recent SOM mapping may not be possible or have lower accuracy due to limited or less recent data availability. In this study, we aimed to improve SOM prediction and mapping accuracy by combining legacy data with limited recent data. Three models, namely, partial least squares regression (PLSR), random forest (RF), and one-dimensional convolutional neural network (1D-CNN), were applied and compared. The results showed that combining legacy and recent data effectively improved SOM prediction accuracy compared to using only recent data. Among the three modeling methods, 1D-CNN exhibited superior performance, with an averaged determination coefficient of the prediction (R2) of 0.58, a root mean square error (RMSE) of 4.56 g/kg, and a ratio of performance to interquartile distance (RPIQ) of 2.05. The predicted SOM content for both legacy (1980 s) and recent (2010 s) periods showed similar spatial distribution patterns throughout the Huanghuaihai Plain. Generally, there was a noticeable trend of increasing SOM content from northwest to southeast, with higher values observed in Jiangsu and lower values concentrated in Henan, Hebei, and Shandong regions within the study area. Over time, SOM contents in the Huanghuaihai Plain showed an increasing trend, with an average increase of 5.90 g/kg from legacy to recent period. This study provides a promising approach for improving SOM prediction and mapping accuracy at large scales, particularly when recent data availability is limited.

Detection and automatic identification of loess sinkholes from the perspective of LiDAR point clouds and deep learning algorithm

Nowadays, the detection and automatic identification of the three-dimensional structure of sinkholes is extremely lacking, which has resulted in significant gaps in sinkholes mapping, soil erosion estimation and morphological studies. In this study, we discovered 249 sinkholes on a river terrace (about 2050 m long and 100 m wide) in a small watershed of Chinese Loess Plateau. Subsequently, we used the unmanned aircraft systems (UAS) and handheld laser scanner (HLS) to investigate these loess sinkholes in detail. We introduced the PointNet ++ deep learning model to train the point cloud dataset for 50 epochs and then selected the best model. In order to evaluate the identification accuracy and transferability of the model, we input point clouds of the unknown prediction area into the trained model to predict the sinkhole point clouds. The trained model exhibits excellent transferability and can effectively identify the sinkhole point clouds in the predicted area (OA = 0.935, IoU (Sinkhole) = 0.662, mIoU = 0.794, AUC = 0.966, Recognition rate = 82.46 %), and even sinkholes with complex connected structures can be accurately identified. This study provides a new perspective for future large-area LiDAR surveys, mapping, and assessment of sinkholes.

A New Constraint on the Relative Disorder of Magnetic Fields between Neutral Interstellar Medium Phases

Utilizing Planck polarized dust emission maps at 353 GHz and large-area maps of the neutral hydrogen (H i) cold neutral medium (CNM) fraction (f CNM), we investigate the relationship between dust polarization fraction (p 353) and f CNM in the diffuse high latitude ( b>30° ) sky. We find that the correlation between p 353 and f CNM is qualitatively distinct from the p 353–H i column density (N H i ) relationship. At low column densities (N H i < 4 × 1020 cm−2) where p 353 and N H i are uncorrelated, there is a strong positive p 353–f CNM correlation. We fit the p 353–f CNM correlation with data-driven models to constrain the degree of magnetic field disorder between phases along the line of sight. We argue that an increased magnetic field disorder in the warm neutral medium (WNM) relative to the CNM best explains the positive p 353–f CNM correlation in diffuse regions. Modeling the CNM-associated dust column as being maximally polarized, with a polarization fraction p CNM ∼ 0.2, we find that the best-fit mean polarization fraction in the WNM-associated dust column is 0.22p CNM. The model further suggests that a significant f CNM-correlated fraction of the non-CNM column (an additional 18.4% of the H i mass on average) is also more magnetically ordered, and we speculate that the additional column is associated with the unstable medium. Our results constitute a new large-area constraint on the average relative disorder of magnetic fields between the neutral phases of the interstellar medium, and are consistent with the physical picture of a more magnetically aligned CNM column forming out of a disordered WNM.

In Vivo Tracking and 3D Mapping of Cell Death in Regeneration and Cancer Using Trypan Blue.

Tracking cell death in vivo can enable a better understanding of the biological mechanisms underlying tissue homeostasis and disease. Unfortunately, existing cell death labeling methods lack compatibility with in vivo applications or suffer from low sensitivity, poor tissue penetration, and limited temporal resolution. Here, we fluorescently labeled dead cells in vivo with Trypan Blue (TBlue) to detect single scattered dead cells or to generate whole-mount three-dimensional maps of large areas of necrotic tissue during organ regeneration. TBlue effectively marked different types of cell death, including necrosis induced by CCl4 intoxication in the liver, necrosis caused by ischemia-reperfusion in the skin, and apoptosis triggered by BAX overexpression in hepatocytes. Moreover, due to its short circulating lifespan in blood, TBlue labeling allowed in vivo "pulse and chase" tracking of two temporally spaced populations of dying hepatocytes in regenerating mouse livers. Additionally, upon treatment with cisplatin, TBlue labeled dead cancer cells in livers with cholangiocarcinoma and dead thymocytes due to chemotherapy-induced toxicity, showcasing its utility in assessing anticancer therapies in preclinical models. Thus, TBlue is a sensitive and selective cell death marker for in vivo applications, facilitating the understanding of the fundamental role of cell death in normal biological processes and its implications in disease.

Integrating geodiversity in animal spatial ecology: microhabitat selection of Eurasian lynx (Lynx lynx) and European wildcat (Felis silvestris) in a karst landscape

Geodiversity, encompassing various geophysical elements, can have an important impact on species distribution and affect animal behaviour patterns. Although many wild felids are attracted to rugged terrain and conspicuous relief features, most previous research was limited to general topographical characteristics (e.g., slope or terrain ruggedness) and rarely considered the effects of specific microhabitat characteristics. This gap is primarily due to the limited availability of high-resolution digital terrain models (DTMs) and relief features data at larger scales. However, LiDAR DTMs can be used in combination with various automatic methods to detect relief features, enabling non-contact and accurate mapping of large, remote and densely-forested areas. Here, we investigated the selection patterns of various karstic relief features, as well as topographic, anthropogenic and vegetation characteristics, by two sympatric felids, the Eurasian lynx (Lynx lynx) and the European wildcat (Felis silvestris), in the Dinaric Mountains, Slovenia. We used LiDAR DTM to calculate topographic characteristics and detect karst relief features based on automatic methods. We compared the selection of these features between the GPS-collared lynx and wildcats under a use-availability approach. We also investigated the differences in the selection of these features by lynx based on their origin and experience (remnant vs. translocated and naive vs. experienced, respectively). We observed significant impact of relief features on space use by both felids and detected distinct selection patterns between the two species. Lynx selected rugged terrain and proximity of caves, cliffs, karst depressions, ridges, small rocky outcrops, and roads, but avoided human settlements and forest edges. Wildcats selected areas with lower surface slope, closer to main roads, forest edges, caves and ridges, but avoided cliffs, forest roads and human settlements. We observed stronger selection/avoidance patterns among the translocated compared to the remnant lynx, while the differences in experience levels were less important. Our study demonstrates the potential of integrating remote sensing techniques and information on geodiversity into the study of animal spatial ecology. Furthermore, our results indicate that specific relief features provide important abiotic microhabitats for felids and may influence habitat segregation between sympatric species. Our findings provide further evidence for the importance of geodiversity conservation and the need to incorporate abiotic microhabitat features in wildlife habitat selection studies.

Ensemble Machine Learning on the Fusion of Sentinel Time Series Imagery with High-Resolution Orthoimagery for Improved Land Use/Land Cover Mapping

In the United States, several land use and land cover (LULC) data sets are available based on satellite data, but these data sets often fail to accurately represent features on the ground. Alternatively, detailed mapping of heterogeneous landscapes for informed decision-making is possible using high spatial resolution orthoimagery from the National Agricultural Imagery Program (NAIP). However, large-area mapping at this resolution remains challenging due to radiometric differences among scenes, landscape heterogeneity, and computational limitations. Various machine learning (ML) techniques have shown promise in improving LULC maps. The primary purposes of this study were to evaluate bagging (Random Forest, RF), boosting (Gradient Boosting Machines [GBM] and extreme gradient boosting [XGB]), and stacking ensemble ML models. We used these techniques on a time series of Sentinel 2A data and NAIP orthoimagery to create a LULC map of a portion of Irion and Tom Green counties in Texas (USA). We created several spectral indices, structural variables, and geometry-based variables, reducing the dimensionality of features generated on Sentinel and NAIP data. We then compared accuracy based on random cross-validation without accounting for spatial autocorrelation and target-oriented cross-validation accounting for spatial structures of the training data set. Comparison of random and target-oriented cross-validation results showed that autocorrelation in the training data offered overestimation ranging from 2% to 3.5%. The XGB-boosted stacking ensemble on-base learners (RF, XGB, and GBM) improved model performance over individual base learners. We show that meta-learners are just as sensitive to overfitting as base models, as these algorithms are not designed to account for spatial information. Finally, we show that the fusion of Sentinel 2A data with NAIP data improves land use/land cover classification using geographic object-based image analysis.

Automated transient grating spectroscopy mapping and signal control for large samples.

We present developments for the mapping of large areas using transient grating spectroscopy (TGS) that allow for smoother, larger, autonomous measurements of material samples. The addition of a precise linear stage in the direction parallel to laser sampling coupled with signal optimizing control allows for hands free, self-correcting measurements. In addition, the simplification of the sample holding design to a form that is small enough to mount directly to the linear stage exhibits a straightforward, low-cost solution for automated TGS applications. This capability is demonstrated by taking large uninterrupted maps of gradient wafers, and the results are validated on calibrated tungsten samples and control TGS samples from gradient wafers.

Analysis of the temporal and spatial characteristics of lunar reconnaissance orbiter’s orbit error based on multi-coverage narrow angle camera images

ABSTRACT Photogrammetric mapping of large area of the lunar surface using high-resolution orbiter images is very challenging due to increasing spatial resolution and lack of a comprehensive understanding of orbit error characteristics. The Lunar Reconnaissance Orbiter (LRO) Narrow Angle Camera (NAC) images are currently one of the highest-resolution images of the Moon available. The study of LRO orbit errors from the perspective of mapping applications is important for high accuracy mapping of large areas using NAC images. This paper proposes a method for orbit error estimation using multi-coverage NAC images. First, available NAC images covering the study regions are orthorectified using ISIS. Then, the impact crater feature matching algorithm is used to match the multi-coverage images. Finally, estimation of the orbit error is achieved through statistical analysis of a large number of homologous points obtained from the matching. The orbit error analyses of nine representative regions on the Moon show that in most cases the refined ephemeris provided by the LOLA/GRAIL team can significantly reduce the orbit error. The orbit errors show a sawtooth variation with a period of about one year, which have been stable for a long time and tend to increase slightly in recent years. There is a slight variation of the orbit error in space, which is even less obvious when using the refined ephemeris. The general tendency is for the far side to be larger than the near side at the same latitude, and for the higher latitudes to be larger than the lower latitudes. The proposed method and analysis results provide a reference for selection and processing of multi-coverage LROC NAC images for high-resolution large-area mapping.

Raman spectroscopy for esophageal tumor diagnosis and delineation using machine learning and the portable Raman spectrometer

Esophageal cancer is one of the leading causes of cancer-related deaths worldwide. The identification of residual tumor tissues in the surgical margin of esophageal cancer is essential for the treatment and prognosis of cancer patients. But the current diagnostic methods, either pathological frozen section or paraffin section examination, are laborious, time-consuming, and inconvenient. Raman spectroscopy is a label-free and non-invasive analytical technique that provides molecular information with high specificity. Here, we report the use of a portable Raman system and machine learning algorithms to achieve accurate diagnosis of esophageal tumor tissue in surgically resected specimens. We tested five machine learning-based classification methods, including k-Nearest Neighbors, Adaptive Boosting, Random Forest, Principal Component Analysis-Linear Discriminant Analysis, and Support Vector Machine (SVM). Among them, SVM shows the highest accuracy (88.61 %) in classifying the esophageal tumor and normal tissues. The portable Raman system demonstrates robust measurements with an acceptable focal plane shift of up to 3 mm, which enables large-area Raman mapping on resected tissues. Based on this, we finally achieve successful Raman visualization of tumor boundaries on surgical margin specimens, and the Raman measurement time is less than 5 min. This work provides a robust, convenient, accurate, and cost-effective tool for the diagnosis of esophageal cancer tumors, advancing toward Raman-based clinical intraoperative applications.

Three-Dimensional Graphene Sheet-Carbon Veil Thermoelectric Composite with Microinterfaces for Energy Applications.

Over the years, various processing techniques have been explored to synthesize three-dimensional graphene (3DG) composites with tunable properties for advanced applications. In this work, we have demonstrated a new procedure to join a 3D graphene sheet (3DGS) synthesized by chemical vapor deposition (CVD) with a commercially available carbon veil (CV) via cold rolling to create 3DGS-CV composites. Characterization techniques such as scanning electron microscopy (SEM), Raman mapping, X-ray diffraction (XRD), electrical resistance, tensile strength, and Seebeck coefficient measurements were performed to understand various properties of the 3DGS-CV composite. Extrusion of 3DGS into the pores of CV with multiple microinterfaces between 3DGS and the graphitic fibers of CV was observed, which was facilitated by cold rolling. The extruded 3D graphene revealed pristine-like behavior with no change in the shape of the Raman 2D peak and Seebeck coefficient. Thermoelectric (TE) power generation and photothermoelectric responses have been demonstrated with in-plane TE devices of various designs made of p-type 3DGS and n-type CV couples yielding a Seebeck coefficient of 32.5 μV K-1. Unlike various TE materials, 3DGS, CV, and the 3DGS-CV composite were very stable at high relative humidity. The 3DGS-CV composite revealed a thin, flexible profile, good moisture and thermal stability, and scalability for fabrication. These qualities allowed it to be successfully tested for temperature monitoring of a Li-ion battery during charging cycles and for large-area temperature mapping.

An innovative methodology for the adjustable use of energy line angle for susceptibility mapping by using cone propagation approach

Rockfall frequently occurs in the mountainous areas and threatens structures such as settlement areas, transportation lines, and agricultural field. The empirical approaches for rockfall mapping have been an attractive research topic in rock mechanics in the recent years, because producing rockfall maps of large areas by using the deterministic and the probabilistic analysis seems difficult due to the necessity of numerous inputs. The cone propagation approach is preferred as a practical tool particularly in the regional scale. The digital elevation model (DEM) of a region prone to rockfall is used for determination of possible propagation zones based on a simple geometric rule known as the energy line angle (reach angle). As a new term, ELAmax_stop was defined to represent the energy line angle that extends to the border of the propagation zone as the maximum run-out distance that is obtained from application of the cone approach to all points (pixels) in source area. The angle denoted as α refers to the threshold slope angle of the steep areas utilized to identify potential source areas by using DEM. Conceptually, the fallen rock blocks within a rockfall-prone region stop within the cone propagation zone, which is bounded by the energy line angles α and ELAmax_stop. While the value of α is susceptible to the resolution of DEM, ELAmax_stop, which exhibits a wide angle range as documented in the literature, is controlled by rock block features together with slope surface properties of the propagation zone. Due to the variability of ELAmax_stop and α depending on the studied region and the resolution of the DEM, the boundary value of the energy line angles between different susceptibility classes need to be adjusted by considering α and ELAmax_stop. By adopting the cone propagation approach to enable adjustable use of energy line angles for rockfall susceptibility mapping, a series of graphical presentations was prepared. These graphical presentations allowed for the prediction of energy line angles corresponding to various rockfall susceptibility classes including very low, low, medium, high, and very high. In addition to the graphical presentations, a series of practical equations were derived for the same purpose. In the final part of the study, a new rating system, namely the run-out distance rating (RDR), was introduced for the preliminary determination of ELAmax_stop. Due to the empirical structure of the methodology, the suggested supportive approach to the practitioners for determining ELAmax_stop should be considered as an initial step that opens to improvement. The proposed methodology in this study was implemented in the regions of Kargabedir Hill and Sivrihisar residential areas in Turkey to prepare rockfall susceptibility maps.

Assessing the accuracy of georeferenced landcover data derived from oblique imagery using machine learning

AbstractRepeat photography offers distinctive insights into ecological change, with ground‐based oblique photographs often predating early aerial images by decades. However, the oblique angle of the photographs presents challenges for extracting and analyzing ecological information using traditional remote sensing approaches. Several innovative methods have been developed for analyzing repeat photographs, but none offer a comprehensive end‐to‐end workflow incorporating image classification and georeferencing to produce quantifiable landcover data. In this paper, we provide an overview of two new tools, an automated deep learning classifier and intuitive georeferencing tool, and describe how they are used to derive landcover data from 19 images associated with the Mountain Legacy Project, a research team that works with the world's largest collection of systematic high‐resolution historic mountain photographs. We then combined these data to produce a contemporary landcover map for a study area in Jasper National Park, Canada. We assessed georeferencing accuracy by calculating the root‐mean‐square error and mean displacement for a subset of the images, which was 4.6 and 3.7 m, respectively. Overall classification accuracy of the landcover map produced from oblique images was 68%, which was comparable to landcover data produced from aerial imagery using a conventional classification method. The new workflow advances the use of repeat photographs for yielding quantitative landcover data. It has several advantages over existing methods including the ability to produce quick and consistent image classifications with little human input, and accurately georeference and combine these data to generate landcover maps for large areas.

Fast thickness mapping of large-area exfoliated two-dimensional transition metal dichalcogenides by imaging spectroscopic ellipsometry

Two-dimensional transition metal dichalcogenides (2D TMDCs) have gained significant attention from the scientific community due to their exceptional properties, making them extremely attractive for optoelectronic and photonic applications. However, many exfoliation or synthesis techniques yield 2D crystals with limited crystalline quality and/or small lateral size. Here, we report a facile Au-assisted exfoliation method, yielding high-quality, large-area monolayers with lateral sizes of hundreds of micrometers. A self-assembled monolayer of (3-aminopropyl)triethoxysilane (APTES) is employed to improve the adhesion between the 2D material and the target substrate, dramatically improving the yield and reliability of the exfoliation process. The monolayer nature of the final sample is then assessed by means of Imaging Spectroscopic Ellipsometry (iSE), which enables a quick and reliable thickness mapping over millimeter-sized areas.

Spatial–Spectral–Temporal Deep Regression Model With Convolutional Long Short-Term Memory and Transformer for the Large-Area Mapping of Mangrove Canopy Height by Using Sentinel-1 and Sentinel-2 Data

Spatial–Spectral–Temporal Deep Regression Model With Convolutional Long Short-Term Memory and Transformer for the Large-Area Mapping of Mangrove Canopy Height by Using Sentinel-1 and Sentinel-2 Data

Probing the Intrinsic Strain in Suspended Graphene Films Using Electron and Optical Microscopy.

Quantifying the intrinsic properties of 2D materials is of paramount importance for advancing their applications. Large-scale production of 2D materials merits the need for approaches that provide direct information about the role of growth substrate on 2D material properties. Transferring the 2D material from its growth substrates can modify the intrinsic properties of the asgrown 2D material. In this study, suspended chemical vapor deposition (CVD) graphene films are prepared directly on their growth substrates in a high-density grid array. The approach facilitates the quantification of intrinsic strain and doping in suspended CVD graphene films. To achieve this, transmission electron microscopy and large-area Raman mapping are employed. Remarkably, the analysis reveals consistent patterns of compressive strain (≈-0.2%) both in the diffraction patterns and Raman maps obtained from these suspended graphene films. By conducting investigations directly on the growth substrates, the potential influences introduced during the transfer process are circumvented effectively. Consequently, the methodology offers a robust and reliable means of studying the intrinsic properties of 2D materials in their authentic form, uninfluenced by the transfer-induced alterations that may skew the interpretation of their properties.

Spatio-temporal multi-level attention crop mapping method using time-series SAR imagery

Accurate crop mapping is of great significance for crop yield forecasting, agricultural productivity development and agricultural management. Thanks to its all-time and all-weather capability, integrating multi-temporal synthetic aperture radar (SAR) for crop mapping has become essential and challenging task in remote sensing. In recent years, deep learning (DL) has demonstrated excellent crop mapping accuracy to interpret crop dynamics. However, existing DL-based methods tend to be incapable of capturing spatial and temporal features at different scales simultaneously, and this often leads to severe mis-classification due to the complex and heterogeneous distribution of crops and diverse phenological patterns. In this paper, we propose a novel spatio-temporal multi-level attention method, named as STMA, for crop mapping using time-series SAR imagery in an end-to-end fashion to increase the capability of crop phenology retrieval. Specifically, the multi-level attention mechanism is designed to aggregate multi-scale spatio-temporal representations on crops via cascaded spatio-temporal self-attention (STSA) and multi-scale cross-attention (MCA) modalities. To ensure a fine extraction of multi-granularity features, a learnable spatial attention position encoding is proposed to adaptively generate the position priors to facilitate multi-level attention learning. Experimental results on Brandenburg Sentinel-1 dataset, public PASTIS-R dataset and South Africa dataset demonstrated that STMA can achieve state-of-the-art performance in crop mapping tasks, with the accuracy of 96.54% in the Brandenburg Sentinel-1 dataset, 86.77% in the PASTIS-R dataset and 83.37% in the South Africa dataset, validating its effectiveness and superiority. Further comparison of spatio-temporal generalization capability reflected its excellent performance in spatio-temporal modeling on different crops and scenarios. This research provides a viable and intelligent spatio-temporal framework for large-area crop mapping using time-series SAR imagery in complex agricultural systems. The Brandenburg Sentinel-1 dataset and the STMA code will be publicly available at https://github.com/hanzhu97702/ISPRS_STMA.