ISPRS Journal of Photogrammetry and Remote Sensing最新文献

Rapid changes in the densely distributed lakes on the Tibetan Plateau (TP) reflect the responses of terrestrial water resources to climate change. Timely and accurate monitoring of lake dynamics is essential for formulating adaptation strategies to manage water and protect public facility safety sustainably. Interfered by the numerous glaciers and snow mountains and limited by the acquisition and computing capacities of massive satellite data, annual inventories of all the lakes ranging from mini to large on the TP are still lacking. Here, we annually mapped these lake areas using all the Landsat imagery, a robust algorithm for detecting surface water according to multiple spectral indices, and Google Earth Engine. We further proposed an effective approach for accurately identifying the glaciers, snow, and mountain shadows in satellite imagery by introducing the characteristics of image luminosity and terrain slope, and removing their data noise remained in the lake maps to generate an annual precise dataset (Lake_TP) of the approximately 9,000 lakes over 0.1 km² on the TP during 1991–2023. We revealed a rapid expansion of lakes with significant spatial heterogeneity, with 6,590 newly increased and 2,851 disappeared lakes found. The total lake areas (554.1 km²/yr) and numbers (77.9/yr) continuously and significantly increased in the period. The growth in lake numbers dominated by small lakes mainly happened before 2005, while the increases in lake areas dominated by large lakes lasted the whole period after 1995. The most significant increases in lake areas and numbers happened in the north of the Inner Basin and Yangtze, the hotspot of lake changes identified in the study. The dataset is expected to promote our understanding of the complete lake evolution process and the dynamic response of the cryosphere to the changing climate. The method proposed is also applicable to continuously monitoring the dynamics of lakes with higher accuracies in other alpine regions around the world. The Lake_TP dataset is publicly available at https://doi.org/10.5281/zenodo.10686952 (Zhou et al. 2024).

The frequent occurrence of El Niño events, in the context of climate change, brings heavy precipitation and extreme heat, severely disrupting agricultural production. Previous efforts have focused on monitoring crop planting areas and evaluating affected crops during disasters. Nevertheless, a comprehensive analysis, including crop planting area mapping, crop damage assessment, and mitigation effectiveness throughout the entire course of a disaster, has been seldom addressed. In this study, we built a comprehensive framework to rapidly investigate the areas of early rice, the extent of flooding impacts, and the post-flood mitigations of early rice during the El Niño flooding event in a typical rice production region – Jiangxi Province in 2023. Early rice planting areas were first mapped by integrating 15-day time series gap-filled Sentinel-1/2 datasets using the Google Earth Engine (GEE) platform, based on a random forest classifier built with the 55 optimized training features. Then the flood-affected early rice map was produced by integrating the early rice planting areas and the Sentinel-1 images-based flood map. Finally, the post-flood newly planted rice fields were identified using the random forest algorithm and classification features from the Sentinel-1/2 images composited during four phenology phases of newly planted rice. The results showed the early rice planting area map, the flooding map, and the newly planted early rice map have overall accuracies of over 90 %. The early rice planting areas reached 120 × 10⁴ ha, and an area of 3.60 × 10⁴ ha (3 %) was flooded due to the heavy rain, and 3.43 × 10⁴ ha flooded areas were newly planted, eventually mitigating the flooding impacts on the production of early rice. This study showcases the potential of all the available Sentinel-1/2 data, cloud computing, and well-established mapping algorithms for tracking rice areas, flooding impacts, and mitigations (i.e., after-flooding replanting) during extreme climate events. The established framework is expected to serve as an early warning system for agricultural adaptation to extreme climate events.

Efficient soybean mapping is critical for agricultural production and yield prediction. However, current sample-driven soybean mapping methods heavily rely on large representative sample datasets, limiting the interpretability of physical mechanisms. Besides, sample-free methods failed to exploit key features that differentiate soybean from other crops, especially Chlorophyll content. Misclassification errors persist and spatiotemporal generalization remains limited. Therefore, this study develops a novel Soybean Mapping Composite Index (SMCI) within a precise Global Optimal Time Window (GOTW). It integrates unique features of soybean Chlorophyll content, canopy water content, and canopy greenness by coupling three red-edge bands (RE2, RE3, and RE4), one near-infrared band, one shortwave infrared band, and two feature indices (Enhanced Vegetation Index and Green Chlorophyll Vegetation Index). The novel index was applied to soybean mapping at six sites in four major soybean producing countries (China, Argentina, Brazil, and the United States) from 2019 to 2021, using an optimal threshold of 3.25. Within the GOTW, the index responds better to spectral features and improves soybean separability. The average overall accuracy (OA: 91%) and average Kappa coefficient (Kappa: 0.83) for the novel index at all sites outperformed the traditional sample-driven Random Forest (RF) method (OA: 84%, Kappa: 0.70) and the existing sample-free index-based Greenness and Water Content Composite Index (GWCCI) (OA: 81%, Kappa: 0.64). Furthermore, interannual transfer experiments consistently showed high accuracy, demonstrating robust spatiotemporal transferability. The proposed SMCI index meets the need for a lightweight and stable soybean mapping tool and serves as a valuable reference for efficient global crop mapping.

Approximately 70%–75% of people worldwide have no formally registered land rights. Fit-For-Purpose Land Administration was introduced to address this problem and focuses on delineating visible cadastral boundaries from earth observation imagery. Recent studies have shown the potential of deep learning models to extract these visible cadastral boundaries automatically. However, studies are limited by the small size and geographical coverage of available datasets and by the lack of information about which cadastral boundaries are visible, i.e., associated with a physical object boundary. To overcome these problems, we present CadastreVision, a benchmark dataset containing cadastral reference data and corresponding multi-resolution earth observation imagery from The Netherlands, with a spatial resolution ranging from 0.1 m to 10 m. The ratio between visible and non-visible cadastral boundaries is essential to evaluate the potential automation level in cadastral boundary extraction from earth observation images and interpret results obtained by deep learning models. We investigate this ratio using a novel analysis pipeline that overlays cadastral reference data with visible topographic object boundaries. Our results show that approximately 72% of the total length of cadastral boundaries in The Netherlands are visible. CadastreVision will enable new developments in cadastral boundary delineation and future endeavours to investigate knowledge transfer to data-scarce areas. Our data and code is available at https://github.com/jeroengrift/cadastrevision.

Accurate and robust simultaneous localization and mapping (SLAM) is essential for autonomous exploration, unmanned transportation, and emergency rescue operations in complex underground spaces. However, the demanding conditions of underground spaces, characterized by poor lighting, weak textures, and high dust levels, pose substantial challenges to SLAM. To address this issue, we propose a graph-based adaptive weighted fusion SLAM (AWF-SLAM) for autonomous robots to achieve accurate and robust SLAM in complex underground spaces. First, a contrast limited adaptive histogram equalization (CLAHE) that combined adaptive gamma correction with weighting distribution (AGCWD) in hue, saturation, and value (HSV) space is proposed to enhance the brightness and contrast of visual images in underground spaces. Then, the performance of each sensor is evaluated using a consistency check based on the Mahalanobis distance to select the optimal configuration for specific conditions. Subsequently, we elaborate an adaptive weighting function model, which leverages the residuals from point cloud matching and the inner point rate of image matching. This model fuses data from light detection and ranging (LiDAR), inertial measurement unit (IMU), and cameras dynamically, enhancing the flexibility of the fusion process. Finally, multiple primitive features are adaptively fused within the factor graph optimization, utilizing a sliding window approach. Extensive experiments were conducted to check the performance of AWF-SLAM using a self-designed mobile robot in underground parking lots, excavated subway tunnels, and complex underground coal mine spaces based on reference trajectories and reconstructions provided by state-of-the-art methods. Satisfactorily, the root mean square error (RMSE) of trajectory translation is only 0.17 m, and the mean relative robustness distance between the point cloud maps reconstructed by AWF-SLAM and the reference point cloud map is lower than 0.09 m. These results indicate a substantial improvement in the accuracy and robustness of SLAM in complex underground spaces.

Fine classification of city-scale buildings using satellite imagery is a crucial research area with significant implications for urban planning, infrastructure development, and population distribution analysis. However, the task faces great challenges due to low-resolution overhead images acquired from high-altitude space-borne platforms and the long-tailed sample distribution of fine-grained urban building categories, leading to a severe class imbalance problem. To address these issues, we propose a deep network approach to the fine-grained classification of urban buildings using open-access satellite images. A Denoising Diffusion Probabilistic Model (DDPM) based super-resolution method is first introduced to enhance the spatial resolution of satellite images, which benefits from domain-adaptive knowledge distillation. Then, a new fine-grained classification network with Category Information Balancing Module (CIBM) and Contrastive Supervision (CS) technique is proposed to mitigate the problem of class imbalance and improve the classification robustness and accuracy. Experiments on Hong Kong data set with 11 distinct building types revealed promising classification results with a mean Top-1 accuracy of 60.45%, which is on par with street-view image based approaches. A comprehensive ablation study demonstrates that the CIBM and CS modules improve Top-1 accuracy by 2.6% and 3.5%, respectively, over the baseline approach. In addition, these modules can be easily integrated into other classification networks, achieving similar performance improvements. This research advances urban analysis by providing an effective solution for detailed classification of buildings in complex mega-city environments using only open-access satellite imagery. The proposed technique can serve as a valuable tool for urban planners, aiding in the understanding of economic, industrial, and population distribution within cities and regions, ultimately facilitating informed decision-making in urban development and infrastructure planning. Data and code will be publicly available at https://github.com/ZhiyiHe1997/UB-FineNet.

The Local Climate Zone (LCZ) scheme representing urban structure and land use pattern is essential for urban heat island (UHI) research. Fine-grained LCZ mapping considering spatial and temporal heterogeneity can provide a more precise characterization of surface thermal properties, thereby enabling a comprehensive analysis and understanding of spatiotemporal trends in climate change research. However, data-driven deep learning-based methods have limitations in coping with the complex urban landscapes of the real-world scenarios, including the spectral similarity of different LCZ and the geospatial heterogeneity of urban LCZ categories. In this study, we constructed a geographic knowledge base for enhanced LCZ characterization with the consideration of prior surface spatial information, including a set of spectral indices and urban morphological parameters (UMPs). Then, we integrated the explicable geographic knowledge base into a learnable deep learning framework in an end-to-end manner for accurate LCZ mapping by fusing multi-source heterogeneous data with a multi-level fusion strategy. The constructed Chinese Climate Zone Time Series LCZ (CClimate-TLCZ) dataset derived from Landsat-8 data with a 30 m spatial resolution, covering 18 representative cities in China, were used to evaluate the proposed framework. The experimental results demonstrate that the proposed framework achieved optimal outcomes across 18 cities, with an average overall accuracy exceeding 94 %, which is more than 20 % higher than that obtained by the standard WUDAPT method. Furthermore, the analysis of LCZ mapping-driven land surface temperature (LST) and surface UHI (SUHI) applications shows that cities within the same climate zone have similar LST distribution patterns, while significant heterogeneity exist between different zones. The annual consistency of LST patterns within each climate zone supports the validity of the LCZ classification scheme and accurate LCZ mapping for urban thermal environment studies. From 2016 to 2022, SUHI intensity initially increases and then decreases, indicating improvements in the urban thermal environment. These findings underscore the critical role of precise LCZ mapping in urban climate resilience and sustainable urban planning.

Accurate extraction of Seasonally Inundated Area (SIA) is pivotal for precise delineation of Flood Inundation Area (FIA). Current methods predominantly rely on Water Inundation Frequency (WIF) to extract SIA, which, due to the lack of analysis of dynamic surface water changes, often yields less accurate and robust results. This significantly hampers the rapid and precise mapping of FIA. In the study, based on the Harmonic Models constructed from Long Time-series Surface Water (LTSW) dynamics, an SIA extraction approach (SHM) was introduced to enhance their accuracy and robustness, thereby improving flood mapping. The experiments were conducted in Poyang Lake, a region characterized by active hydrological phenomena. Sentinel-1/2 remote sensing data were utilized to extract LTSW. Harmonic analysis was applied to the LTSW dataset, using the amplitude terms in the harmonic model to characterise the frequency of variation between land and water for the surface units, thus extracting the SIAs. The results reveal that the harmonic model parameters are capable of portraying SIA. In comparison to the commonly used WIF thresholding method for SIA extraction, the SHM approach demonstrates superior accuracy and robustness. Leveraging the SIA extracted through SHM, a higher level of accuracy in FIA extraction is achieved. Overall, the SHM offers notable advantages, including high accuracy, automation, and robustness. It offers reliable reference water extents for flood mapping, especially in areas with active and complex hydrological dynamics. SHM can play a crucial role in emergency response to flood disasters, providing essential technical support for natural disaster management and related departments.

The olive tree holds great cultural, environmental, and economic significance in the Mediterranean region. In particular, Morocco has been making dedicated investments over $10 billion since 2008 to fuel the transition from cereal to olive production. Understanding the spatial extent of this large-scale land conversion is critical for a variety of socioeconomic purposes. In response to this demand, we conducted a study to map individual olive trees in northern Morocco using satellite imagery and deep learning techniques at a sub-national scale. This study utilized cloud-free, very-high-resolution DigitalGlobe imagery collected between 2018 and 2022 to identify each individual olive tree in six northern Morocco provinces. We compared various deep learning models, including both transformer-based and CNN-based models, to generate patch-level spatial constraints and pixel-level tree identification. We found that transformer-based models outperformed CNN-based models in both tasks. Additionally, spatially constraining the pixel-level results improved olive tree mapping accuracy to varying degrees, depending on the initial performance of the model. The evaluation of the olive map generated from this study shows high accuracy in both surveyed and unsampled regions. This research represents the first-of-its-kind individual olive tree mapping at the sub-national scale that can help monitor the large-scale land conversions such as about 110,000 ha of olive plantings in the six Moroccan provinces studies here. Meanwhile it demonstrates a cost-effective and efficient prototype approach that can be adapted to identify similar tree crop expansion occurring in other parts of the world.

Training deep neural networks for semantic segmentation of aerial images relies heavily on obtaining a large number of precise pixel-level annotations, which can cause significant annotation expenses. Given the fact that acquiring fine-class annotations is considerably more challenging than obtaining coarse-class annotations, we present a novel semi-supervised learning framework, which utilizes high spatial resolution images annotated with coarse-class labels alongside a very small set of fine-grained annotated images as the training set, thereby achieving classification results that are refined in both spatial resolution and categorical granularity. Specifically, this framework adopts Mix Transformer (MiT) as the backbone architecture to accommodate both local feature extraction and long-range dependency modeling capabilities and utilizes multi-prototype learning to model each class as multiple sub-prototypes, preserving the intrinsic variance characteristics within classes. We propose a dedicated co-training approach tailored for extracting fine-grained pseudo-labels from coarse-grained samples. In this approach, a local-softmax pseudo-labeling strategy is developed to ensure a harmonious balance between the efficiency and accuracy of the pseudo-labeling, and four losses are formulated for both single-level class and cross-category granularity supervised learning. We evaluate the proposed framework on the Gaofen Image Dataset (GID) and Five-Billion-Pixels (FBP) dataset, confirming its feasibility and superior results. In particular, based on coarse-class annotations, the performance achieved using only 5% of fine-class labels, in terms of the four metrics, namely mIoU, mean UA, mean F1-score, and OA, reached 91%, 96%, 89%, and 93% of the fully-supervised baseline performance respectively. The code is available at https://github.com/chenhaocs/C2F.