Search Results (57)

Anthropogenic and climatic pressures can transform contiguous forests into smaller, less connected fragments. Forest biodiversity and ecosystem functioning can furthermore be compromised or enhanced. We present a descriptive analysis of forest fragmentation in Bavaria, the largest federal state in Germany. We calculated 22 metrics of fragmentation using forest polygons, aggregated within administrative units and with respect to both elevation and aspect orientation. Using a forest mask from September 2024, we found 2.384 million hectares of forest across Bavaria, distributed amongst 83,253 forest polygons 0.1 hectare and larger. The smallest patch category (XS, <25 ha) outnumbered all other size classes by nearly 13 to 1. Edge zones accounted for more than 1.68 million hectares, leaving less than 703,000 hectares as core forest. Although south-facing slopes dominated the state, the highest forest cover (~36%) was found on the least abundant east-oriented slopes. Most of the area is located at 400–600 m.a.s.l., with around 30% of this area covered by forests; however, XL forest patches (>3594 ha) dominated higher elevations, covering 30–60% of land surface area between 600 and 1400 m.a.s.l. The distribution of the largest patches follows the higher terrain and corresponds well to protected areas. K-means clustering delineated 3 clusters, which corresponded well with the predominance of patchiness, aggregation, and edginess within districts. Full article

Surface cracks are key indicators of landslide deformation, crucial for early landslide identification and deformation pattern analysis. However, due to the complex terrain and landslide extent, manual surveys or traditional digital image processing often face challenges with efficiency, precision, and interference susceptibility in detecting these cracks. Therefore, this study proposes a comprehensive automated pipeline to enhance the efficiency and accuracy of landslide surface crack detection. First, high-resolution images of landslide areas are collected using unmanned aerial vehicles (UAVs) to generate a digital orthophoto map (DOM). Subsequently, building upon the U-Net architecture, an improved encoder–decoder semantic segmentation network (IEDSSNet) was proposed to segment surface cracks from the images with complex backgrounds. The model enhances the extraction of crack features by integrating residual blocks and attention mechanisms within the encoder. Additionally, it incorporates multi-scale skip connections and channel-wise cross attention modules in the decoder to improve feature reconstruction capabilities. Finally, post-processing techniques such as morphological operations and dimension measurements were applied to crack masks to generate crack inventories. The proposed method was validated using data from the Heifangtai loess landslide in Gansu Province. Results demonstrate its superiority over current state-of-the-art semantic segmentation networks and open-source crack detection networks, achieving F1 scores and IOU of 82.11% and 69.65%, respectively—representing improvements of 3.31% and 4.63% over the baseline U-Net model. Furthermore, it maintained optimal performance with demonstrated generalization capability under varying illumination conditions. In this area, a total of 1658 surface cracks were detected and cataloged, achieving an accuracy of 85.22%. The method proposed in this study demonstrates strong performance in detecting surface cracks in landslide areas, providing essential data for landslide monitoring, early warning systems, and mitigation strategies. Full article

The pre-training and fine-tuning paradigm has significantly advanced satellite remote sensing applications. However, its potential remains largely underexplored for airborne laser scanning (ALS), a key technology in domains such as forest management and urban planning. In this study, we address this gap by constructing a large-scale ALS point cloud dataset and evaluating its effectiveness in downstream applications. We first propose a simple, generalizable framework for dataset construction, designed to maximize land cover and terrain diversity while allowing flexible control over dataset size. We instantiate this framework using ALS, land cover, and terrain data collected across the contiguous United States, resulting in a dataset geographically covering 17,000 + ${\mathrm{km}}^2$ (184 billion points) with diverse land cover and terrain types included. As a baseline self-supervised learning model, we adopt BEV-MAE, a state-of-the-art masked autoencoder for 3D outdoor point clouds, and pre-train it on the constructed dataset. The resulting models are fine-tuned for several downstream tasks, including tree species classification, terrain scene recognition, and point cloud semantic segmentation. Our results show that pre-trained models consistently outperform their counterparts trained from scratch across all downstream tasks, demonstrating the strong transferability of the learned representations. Additionally, we find that scaling the dataset using the proposed framework leads to consistent performance improvements, whereas datasets constructed via random sampling fail to achieve comparable gains. Full article

Terrain conductivity meters (TCMs) are efficient devices for different sorts of subsurface investigations, including detecting and tracing buried utilities, such as metallic pipes and cables. However, data collected using TCMs are usually ambiguous and hard to interpret. This ambiguity originates from the complex shape of apparent conductivity anomalies, the influence of irrelevant conductive bodies, and the interference of random noise with the collected data. To overcome this ambiguity and produce more interpretable apparent conductivity maps, a three-step filtering routine is proposed and tested using different real datasets. The filtering routine begins with applying a Savitzky–Golay (SG) filter to reduce the effect of random noise. This is followed by a modified rolling ball (MRB) filter to convert the complex M-shape of the anomaly into a single trough pointing to the underground utility. Finally, a virtual resolution enhancement (VRE) filter is applied to enhance the pinpointing apex of the trough. The application of the proposed filtering routine to apparent conductivity data collected using different terrain conductivity meters over different utilities in different urban environments shows a significant improvement of the data and an effective ability to reveal masked underground utilities. The proposed triple filtering routine can be a starting point for a new generation of TCMs with a built-in operation mode for instantaneous delineation and characterization of underground utilities in real time. Full article

Shadow in remote sensing images can obscure important details of land features, making shadow detection crucial for enhancing the accuracy of subsequent analyses and applications. Current shadow detection methods primarily rely on the spectral information of images, which can often result in shadow misdetection due to the phenomenon of spectral confusion of different objects. To mitigate this issue, we propose a method that combines topography and spectra (CTS). Firstly, we introduce a new DEM-based shadow coarse detection method to obtain the DEM rough shadow mask, which uses a relationship between the magnitude of terrain height angle and solar elevation angle to determine shadow properties. Then, we use the MC3 (modified C3 component) index-based shadow fine detection method to obtain an MC3 mean map, which includes image enhancement with a stretching process and multi-scale superpixel segmentation. We then derive the Shadow pixel Proportion Map (SPM) by counting the DEM rough shadow mask in terms of superpixels. The Joint Shadow probability Map (JSM) is obtained by combining the SPM and the MC3 mean map with specific weights. Finally, a multi-level Otsu threshold method is applied to the JSM to generate the shadow mask. We compare the proposed CTS method against several state-of-the-art algorithms through both qualitative assessments and quantitative metrics. The results show that the CTS method demonstrates superior accuracy and consistency in detecting true shadows, achieving an average overall accuracy of 95.81% on mountainous remote sensing images. Full article

This paper proposes a novel SDA-Mask R-CNN framework for precise seabed terrain edge feature extraction from Side-Scan Sonar (SSS) images to enhance Unmanned Underwater Vehicle (UUV) perception and navigation. The developed architecture addresses critical challenges in underwater image analysis, including low segmentation accuracy and ambiguous edge delineation, through three principal innovations. First, we introduce a Structural Synergistic Group-Attention Residual Network (SSGAR-Net) that integrates group convolution with an enhanced convolutional block attention mechanism, complemented by a layer-skipping architecture for optimized information flow and redundancy verification for computational efficiency. Second, a Depth-Weighted Hierarchical Fusion Network (DWHF-Net) incorporates depthwise separable convolution to minimize computational complexity while preserving model performance, which is particularly effective for high-resolution SSS image processing. This module further employs a weighted pyramid architecture to achieve multi-scale feature fusion, significantly improving adaptability to diverse object scales in dynamic underwater environments. Third, an Adaptive Synergistic Mask Optimization (ASMO) strategy systematically enhances mask generation through classification head refinement, adaptive post-processing, and progressive training protocols. Comprehensive experiments demonstrate that our method achieves 0.695 (IoU) segmentation accuracy and 1.0 (AP) edge localization accuracy. The proposed framework shows notable superiority in preserving topological consistency of seabed features, offering a reliable technical framework for underwater navigation and seabed mapping in marine engineering applications. Full article

Assessing vines’ vigour is essential for vineyard management and automatization of viticulture machines, including shaking adjustments of berry harvesters during grape harvest or leaf pruning applications. To address these problems, based on a standardized growth class assessment, labeled ground truth data of precisely located grapevines were predicted with specifically selected Machine Learning (ML) classifiers (Random Forest Classifier (RFC), Support Vector Machines (SVM)), utilizing multispectral UAV (Unmanned Aerial Vehicle) sensor data. The input features for ML model training comprise spectral, structural, and texture feature types generated from multispectral orthomosaics (spectral features), Digital Terrain and Surface Models (DTM/DSM- structural features), and Gray-Level Co-occurrence Matrix (GLCM) calculations (texture features). The specific features were selected based on extensive literature research, including especially the fields of precision agri- and viticulture. To integrate only vine canopy-exclusive features into ML classifications, different feature types were extracted and spatially aggregated (zonal statistics), based on a combined pixel- and object-based image-segmentation-technique-created vine row mask around each single grapevine position. The extracted canopy features were progressively grouped into seven input feature groups for model training. Model overall performance metrics were optimized with grid search-based hyperparameter tuning and repeated-k-fold-cross-validation. Finally, ML-based growth class prediction results were extensively discussed and evaluated for overall (accuracy, f1-weighted) and growth class specific- classification metrics (accuracy, user- and producer accuracy). Full article

Optical remote sensing images have a wide range of applications but are often affected by cloud cover, which interferes with subsequent analysis. Therefore, cloud removal has become indispensable in remote sensing data processing. The Tibetan Plateau, as a sensitive region to climate change, plays a crucial role in the East Asian water cycle and regional climate due to its snow cover. However, the rich ice and snow resources, rapid snow condition changes, and active atmospheric convection in the plateau as well as its surrounding mountainous areas, make optical remote sensing prone to cloud interference. This is particularly significant when monitoring snow cover changes, where cloud removal becomes essential considering the complex terrain and unique snow characteristics of the Tibetan Plateau. This paper proposes a novel Multi-Scale Attention-based Cloud Removal Model (MATT). The model integrates global and local information by incorporating multi-scale attention mechanisms and local interaction modules, enhancing the contextual semantic relationships and improving the robustness of feature representation. To improve the segmentation accuracy of cloud- and snow-covered regions, a cloud mask is introduced in the local-attention module, combined with the local interaction module to modulate and reconstruct fine-grained details. This enables the simultaneous representation of both fine-grained and coarse-grained features at the same level. With the help of multi-scale fusion modules and selective attention modules, MATT demonstrates excellent performance on both the Sen2_MTC_New and XZ_Sen2_Dataset datasets. Particularly on the XZ_Sen2_Dataset, it achieves outstanding results: PSNR = 29.095, SSIM = 0.897, FID = 125.328, and LPIPS = 0.356. The model shows strong cloud removal capabilities in cloud- and snow-covered areas in mountainous regions while effectively preserving snow information, and providing significant support for snow cover change studies. Full article

Instance segmentation of remote sensing images can not only provide object-level positioning information but also provide pixel-level positioning information. This pixel-level information annotation has a wide range of uses in the field of remote sensing, and it is of great value for environmental detection and resource management. Because optical images generally have complex terrain environments and changeable object shapes, SAR images are affected by complex scattering phenomena, and the mask quality obtained by the traditional instance segmentation method used in remote sensing images is not high. Therefore, it is a challenging task to improve the mask quality of instance segmentation in remote sensing images. Since the traditional two-stage instance segmentation method consists of backbone, neck, bbox head, and mask head, the final mask quality depends on the product of all front-end work quality. Therefore, we consider the difficulty of optical and SAR images to bring instance segmentation to the targeted improvement of the neck, bbox head, and mask head, and we propose the Context-Aggregated and SAM-Guided Network (CSNet). In this network, the plain feature fusion pyramid network (PFFPN) can generate a pyramid for the plain feature and provide a feature map of the appropriate instance scale for detection and segmentation. The network also includes a context aggregation bbox head (CABH), which uses the context information and instance information around the instance to solve the problem of missed detection and false detection in detection. The network also has a SAM-Guided mask head (SGMH), which learns by using SAM as a teacher, and uses the knowledge learned to improve the edge of the mask. Experimental results show that CSNet significantly improves the quality of masks generated under optical and SAR images, and CSNet achieves 5.1% and 3.2% AP increments compared with other SOTA models. Full article

Unmanned aerial vehicle (UAV) photogrammetry allows the generation of orthophoto and digital surface model (DSM) rasters of terrain. However, DSMs of water bodies mapped using this technique often reveal distortions in the water surface, thereby impeding the accurate sampling of water surface elevation (WSE) from DSMs. This study investigates the capability of deep neural networks to accommodate the aforementioned perturbations and effectively estimate WSE from photogrammetric rasters. Convolutional neural networks (CNNs) were employed for this purpose. Two regression approaches utilizing CNNs were explored: direct regression employing an encoder and a solution based on prediction of the weight mask by an autoencoder architecture, subsequently used to sample values from the photogrammetric DSM. The dataset employed in this study comprises data collected from five case studies of small lowland streams in Poland and Denmark, consisting of 322 DSM and orthophoto raster samples. A grid search was employed to identify the optimal combination of encoder, mask generation architecture, and batch size among multiple candidates. Solutions were evaluated using two cross-validation methods: stratified k-fold cross-validation, where validation subsets maintained the same proportion of samples from all case studies, and leave-one-case-out cross-validation, where the validation dataset originates entirely from a single case study, and the training set consists of samples from other case studies. Depending on the case study and the level of validation strictness, the proposed solution achieved a root mean square error (RMSE) ranging between 2 cm and 16 cm. The proposed method outperforms methods based on the straightforward sampling of photogrammetric DSM, achieving, on average, an 84% lower RMSE for stratified cross-validation and a 62% lower RMSE for all-in-case-out cross-validation. By utilizing data from other research, the proposed solution was compared on the same case study with other UAV-based methods. For that benchmark case study, the proposed solution achieved an RMSE score of 5.9 cm for all-in-case-out cross-validation and 3.5 cm for stratified cross-validation, which is close to the result achieved by the radar-based method (RMSE of 3 cm), which is considered the most accurate method available. The proposed solution is characterized by a high degree of explainability and generalization. Full article

In recent years, the three-dimensional (3D) radar detection range has played an essential role in the layout of devices such as aircraft and drones. To compensate for the shortcomings of three-dimensional calculations for radar terrain masking, a new calculation method is proposed for assessing the terrain occlusion of radar detection range. First, the high-dimensional electromagnetic data after discretization are modeled based on the tensor data structure, and the tensor grid dilation operator is constructed. Then, the dilation process begins from the overlapping section of the radar detection range and terrain, and it is adjusted by the terrain occlusion judgment factor and the dilation judgment factor to obtain the obstructed part due to the terrain. Finally, the actual radar detection range under terrain occlusion is obtained. The simulation results show that the method proposed in this paper can adapt to different grid sizes and terrain shapes, significantly enhancing computational efficiency while maintaining internal features. Full article

This is an exercise to explore the concentration of lithium, lithium-7 isotope and the possible presence of black dirty ice on the lunar surface using spectral data obtained from the Clementine mission. The main interest in tracing the lithium and presence of dark ice on the lunar surface is closely related to future human settlement missions on the moon. We investigate the distribution of lithium and 7 Li isotope on the lunar surface by employing spectral data from the Clementine images. We utilized visible (VIS–NIR) imagery at wavelengths of 450, 750, 900, 950 and 1000 nm, along with near-infrared (NIR–SWIR) at 1100, 1250, 1500, 2000, 2600 and 2780 nm, encompassing 11 bands in total. This dataset offers a comprehensive coverage of about 80% of the lunar surface, with resolutions ranging from 100 to 500 m, spanning latitudes from 80°S to 80°N. In order to extract quantitative abundance of lithium, ground-truth sites were used to calibrate the Clementine images. Samples (specifically, 12045, 15058, 15475, 15555, 62255, 70035, 74220 and 75075) returned from Apollo missions 12, 15, 16 and 17 have been correlated to the Clementine VIS–NIR bands and five spectral ratios. The five spectral ratios calculated synthesize the main spectral features of sample spectra that were grouped by their lithium and 7 Li content using Principal Component Analysis. The ratios spectrally characterize substrates of anorthosite, silica-rich basalts, olivine-rich basalts, high-Ti mare basalts and Orange and Glasses soils. Our findings reveal a strong linear correlation between the spectral parameters and the lithium content in the eight Apollo samples. With the values of the 11 Clementine bands and the 5 spectral ratios, we performed linear regression models to estimate the concentration of lithium and 7 Li. Also, we calculated Digital Terrain Models (Altitude, Slope, Aspect, DirectInsolation and WindExposition) from LOLA-DTM to discover relations between relief and spatial distribution of the extended models of lithium and 7 Li. The analysis was conducted in a mask polygon around the Apollo 15 landing site. This analysis seeks to uncover potential 7 Li enrichment through spallation processes, influenced by varying exposure to solar wind. To explore the possibility of finding ice mixed with regolith (often referred to as `black ice’), we extended results to the entire Clementine coverage spectral indices, calculated with a library (350–2500 nm) of ice samples contaminated with various concentrations of volcanic particles. Full article

Sentinel-1-based flood mapping works well but with well-known issues over rugged terrain. Applying exclusion masks to improve the results is common practice in unsupervised and global applications. One such mask is the height above the nearest drainage (HAND), which uses terrain information to reduce flood lookalikes in SAR images. The TU Wien flood mapping algorithm is one operational workflow using this mask. Being a Bayesian method, this algorithm can integrate auxiliary information as prior probabilities to improve classifications. This study improves the TU Wien flood mapping algorithm by introducing a HAND prior function instead of using it as a mask. We estimate the optimal function parameters and observe the performance in flooded and non-flooded scenarios in six study sites. We compare the flood maps generated with HAND and (baseline) non-informed priors with reference CEMS rapid mapping flood extents. Our results show enhanced performance by decreasing false negatives at the cost of slightly increasing false positives. In utilizing a single parametrization, the improved algorithm shows potential for global implementation. Full article

Geothermal energy is an eco-friendly, renewable source of underground thermal energy that exists in the interior of the earth. By tapping into these formations, fluids can be channeled to heat the rock formations above, resulting in a significantly higher land surface temperature (LST). However, LST readings are influenced by various factors such as sun radiation, cyclical variations, and precipitation, which can mask the temperature anomalies caused by geothermal heat. To address these issues and highlight the LST anomalies caused by geothermal heat, this paper proposes a methodology to efficiently and quickly calculate the multi-temporal LST leveraging of the Google Earth Engine (GEE) in the Damxung–Yangbajain basin, Qinghai–Tibet Plateau. This method incorporates terrain correction, altitude correction, and multi-temporal series comparison to extract thermal anomaly signals. The existing geothermal manifestations are used as a benchmark to further refine the methodology. The results indicate that the annual mean winter LST is a sensitive indicator of geothermal anomaly signals. The annual mean winter LST between 2015 and 2020 varied from −14.7 °C to 26.7 °C, with an average of 8.6 °C in the study area. After altitude correction and water body removal, the annual mean winter LST varied from −22.1 °C to 23.3 °C, with an average of 6.2 °C. When combining the distribution of faults with the results of the annual mean winter LST, this study delineated the geothermal potential areas that are located predominantly around the fault zone at the southern foot of the Nyainqentanglha Mountains. Geothermal potential areas exhibited a higher LST, ranging from 12.6 °C to 23.3 °C. These potential areas extend to the northeast, and the thermal anomaly range reaches as high as 19.6%. The geothermal potential area makes up 8.2% of the entire study area. The results demonstrate that the approach successfully identified parts of known geothermal fields and indicates sweet spots for future research. This study highlights that utilizing the multi-temporal winter LST is an efficient and cost-effective method for prospecting geothermal resources in plateau environments. Full article

Radiometric Terrain Corrected (RTC) gamma nought backscatter, which was introduced around a decade ago, has evolved into the standard for analysis-ready Synthetic Aperture Radar (SAR) data. While working with RTC backscatter data is particularly advantageous over undulated terrain, it requires substantial computing resources given that the terrain flattening is more computationally demanding than simple orthorectification. The extra computation may become problematic when working with large SAR datasets such as the one provided by the Sentinel-1 mission. In this study, we examine existing Sentinel-1 RTC pre-processing workflows and assess ways to reduce processing and storage overheads by considering the satellite’s high orbital stability. By propagating Sentinel-1’s orbital deviations through the complete pre-processing chain, we show that the local contributing area and the shadow mask can be assumed to be static for each relative orbit. Providing them as a combined external static layer to the pre-processing workflow, and streamlining the transformations between ground and orbit geometry, reduces the overall processing times by half. We conducted our experiments with our in-house developed toolbox named wizsard, which allowed us to analyse various aspects of RTC, specifically run time performance, oversampling, and radiometric quality. Compared to the Sentinel Application Platform (SNAP) this implementation allowed speeding up processing by factors of 10–50. The findings of this study are not just relevant for Sentinel-1 but for all SAR missions with high spatio-temporal coverage and orbital stability. Full article