ISPRS Journal of Photogrammetry and Remote Sensing最新文献

Accurate and detailed spatial information on rice cultivation is essential to developing agricultural policy and reducing the negative impacts of agriculture. However, the dependence of most traditional methods on samples severely limits the feasibility of large-scale rice cultivation mapping. This study proposes a robust large-scale sample-free monitoring method for single-season rice in northern China. A new rice phenology index, quantifying dynamic phenological features of rice (i.e., the occurrence of flooding during transplanting and the growth of rice after transplanting), was generated to highlight rice. Subsequently, a constrained cyclic threshold classification strategy was designed to obtain plausible rice maps using statistical data. Innovatively combining rice mapping with statistical data, the most detailed (10 m) single-season rice map in northern China to date was created. Compared with three other high-precision rice map products, the resulting rice map has high accuracy and good local details. The results indicate that the rice phenology index has excellent and robust performance in identifying rice cultivation locations in northern China. Moreover, the proposed mapping method exhibits clear advantages in the tracking of large-scale and historical rice cultivation. As a whole, this study provides a paradigm of using statistical data instead of samples for crop mapping.

Synthetic aperture radar (SAR) is an efficient and widely used remote sensing tool. However, data extracted from SAR images are contaminated with speckle, which precludes the application of techniques based on the assumption of additive and normally distributed noise. One of the most successful approaches to describing such data is the multiplicative model, where intensities can follow a variety of distributions with positive support. The $G_I^0$ model is among the most successful ones. Although several estimation methods for the $G_I^0$ parameters have been proposed, there is no work exploring a regression structure for this model. Such a structure could allow us to infer unobserved values from available ones. In this work, we propose a $G_I^0$ regression model and use it to describe the influence of intensities from other polarimetric channels. We derive some theoretical properties for the new model: Fisher information matrix, residual measures, and influential tools. Maximum likelihood point and interval estimation methods are proposed and evaluated by Monte Carlo experiments. Results from simulated and actual data show that the new model can be helpful for SAR image analysis.

Cropland non-agriculturalization (CNA) refers to the conversion of cropland into construction land, woodland/garden/grassland, water body, or other non-agricultural land, which ultimately disrupts local agroecosystems and the cultivation and production of crops. Remote sensing technology is an important tool for large-area CNA detection, and remote sensing based methods that can be used for this task include the time-series analysis method and change detection from bi-temporal images. In particular, change detection methods using high-resolution remote sensing imagery have great potential for CNA detection, but enormous challenges do still remain. The large intra-class variance of cropland with different phenological stages and planting patterns leads to cropland areas being difficult to identify effectively, while certain features can be misidentified because they are similar to cropland, resulting in false alarms and missed detections in the results. There is also a lack of large-scale CNA datasets covering multiple change scenarios as data support. To address these problems, a lightweight model focused on CNA detection (CNANet) is proposed in this paper. Specifically, the uniquely crafted represent-consist-enhance (RCE) module is seamlessly integrated between the encoder and decoder components of CNANet to perform a contrast operation on the deep features extracted by the feature extractor. The RCE module is specifically designed to aggregate multiple cropland representations and extend the cropland representations from the confusing background, to achieve the purpose of reducing the intra-class reflectance differences and enhancing the model’s perception of cropland. In addition, a large-scale high-resolution cropland non-agriculturalization (Hi-CNA) dataset was built for the CNA identification task, with a total of 6797 pairs of 512 × 512 images with semantic annotations. Compared to the existing datasets, the Hi-CNA dataset has the advantages of multiple phenological stages, multiple change scenarios, and multiple annotation types, in addition to the large data volume. The experimental results obtained in this study show that the benchmark methods tested on the Hi-CNA dataset can all achieve a good accuracy, proving the high-quality annotation of the dataset. The overall accuracy and F1-score of CNANet with the default settings reach 93.81 % and 78.9 %, respectively, achieving a superior accuracy, compared to the other benchmark methods, and demonstrating stronger perception of cropland changes. In addition, in two selected verification regions within the large-scale real-world CNA mapping results, the F1-score is 83.61 % and 50.87 %. The Hi-CNA can be downloaded from http://rsidea.whu.edu.cn/Hi-CNA_dataset.htm.

Flood detection from synthetic aperture radar (SAR) imagery plays an important role in crisis and disaster management. Based on pre- and post-flood SAR images, flooded areas can be extracted by detecting changes of water bodies. Existing state-of-the-art change detection methods primarily target optical image pairs. The nature of SAR images, such as scarce visual information, similar backscatter signals, and ubiquitous speckle noise, pose great challenges to identifying water bodies and mining change features, thus resulting in unsatisfactory performance. Besides, the lack of large-scale annotated datasets hinders the development of accurate flood detection methods. In this paper, we focus on the difference between SAR image pairs and present a differential attention metric-based network (DAM-Net), to achieve flood detection. By introducing feature interaction during temporal-wise feature representation, we guide the model to focus on changes of interest rather than fully understanding the scene of the image. On the other hand, we devise a class token to capture high-level semantic information about water body changes, increasing the ability to distinguish water body changes and pseudo changes caused by similar signals or speckle noise. To better train and evaluate DAM-Net, we create a large-scale flood detection dataset using Sentinel-1 SAR imagery, namely S1GFloods. This dataset consists of 5,360 image pairs, covering 46 flood events during 2015–2022, and spanning 6 continents of the world. The experimental results on this dataset demonstrate that our method outperforms several advanced change detection methods. DAM-Net achieves 97.8% overall accuracy, 96.5% F1, and 93.2% IoU on the test set. Our dataset and code are available at https://github.com/Tamer-Saleh/S1GFlood-Detection.

We developed a new algorithm for computing radiometrically and spectrally consistent surface reflectances from multiple sensors. The algorithm approximates surface reflectances of reference sensors directly from top-of-atmosphere (TOA) reflectances of sensors-to-be-transformed. A unique characteristic of the algorithm is that coefficients in the algorithm are computed independently using statistics of time-series reflectance data for each sensor; thus, no regressions or optimizations using pairs of data from different sensors are required. This characteristic can lead to a substantial reduction in the number of computational tasks required for calibrating models when numerous satellite sensors or datasets are used. First, a system of equations relating TOA reflectances of one sensor and surface reflectances of another sensor in the red and near-infrared bands was analytically approximated using a linear mixture model of three land-cover types and radiative transfer in the atmosphere. The equations were subsequently used to develop an unmixing-based algorithm for radiometric corrections and spectral transformations. The algorithm was evaluated using synchronous observation data and long-term time-series data with middle spatial resolution, which were obtained from the Landsat 4–5 Multispectral Scanner (MSS) and Thematic Mapper (TM) sensors. Results obtained using contemporaneous data from the two sensors indicated that cross-sensor differences in reflectances and in a spectral index, the normalized difference vegetation index (NDVI), between the MSS and TM sensors were reduced to reasonable levels after the algorithm was applied; the magnitudes of remaining biases were less than 0.005 in reflectance units and less than 0.03 in NDVI units. Results obtained using time-series data for four regions of interest with different land-cover types indicated that the transformed MSS time-series data well synchronized with the TM data used as a reference. Reflectance differences remaining after implementation of the algorithm were possibly due to instability of the algorithm for computing parameters, sensor-dependent quality assurance (QA) data and QA accuracy, and geolocation errors, among others. The concept of the developed algorithm might be applicable universally to various combinations of spectral bands and sensors/missions, which should be further evaluated for cross-sensor radiometric and spectral harmonization with the aim of multi-sensor analysis.

Monitoring and managing Earth’s surface resources is critical to human settlements, encompassing essential tasks such as city planning, disaster assessment, etc. To accurately recognize the categories and locations of geographical objects and reason about their spatial or semantic relations , we propose a multi-task framework named EarthVQANet, which jointly addresses segmentation and visual question answering (VQA) tasks. EarthVQANet contains a hierarchical pyramid network for segmentation and semantic-guided attention for VQA, in which the segmentation network aims to generate pixel-level visual features and high-level object semantics, and semantic-guided attention performs effective interactions between visual features and language features for relational modeling. For accurate relational reasoning, we design an adaptive numerical loss that incorporates distance sensitivity for counting questions and mines hard-easy samples for classification questions, balancing the optimization. Experimental results on the EarthVQA dataset (city planning for Wuhan, Changzhou, and Nanjing in China), RSVQA dataset (basic statistics for general objects), and FloodNet dataset (disaster assessment for Texas in America attacked by Hurricane Harvey) show that EarthVQANet surpasses 11 general and remote sensing VQA methods. EarthVQANet simultaneously achieves segmentation and reasoning, providing a solid benchmark for various remote sensing applications. Data is available at http://rsidea.whu.edu.cn/EarthVQA.htm

To tackle the issue of lack of semantic consistency between ground and non-ground points, as well as the damage to the integrity of terrain boundary information during network downsampling, we developed a Semantic Consistency-Convolutional Neural Network (SC-CNN) to improve the precision of point cloud filtering under complex terrain conditions. The novel aspects include: (1) farthest point sampling (FPS) with slope constraints, which enhances terrain contour preservation through adaptive subblock partitioning and slope-based sampling; (2) intra-class feature enhancement via copula correlation and attention mechanisms, improving the network’s ability to distinguish between ground and non-ground points by focusing on intra-class feature consistency and inter-class differences; and (3) filter error correction using copula correlation and confidence intervals. refining filtering accuracy by adjusting for negatively correlated point sets. Tested on the ISPRS and 3D Vaihingen datasets, SC-CNN notably outperformed existing methods, reducing the mean total error (MT.E) by 0.17% and 1.93%, respectively, thereby significantly enhancing point-cloud filtering accuracy under complex terrain conditions.

Extensive mortality of susceptible spruce can be caused by spruce beetles at epidemic population levels, as in the ongoing outbreak in Southcentral Alaska. Although information on outbreak extent and severity underpins forest management and research, the data products available in Alaska have substantial gaps. Widely available high-resolution satellite imagery are a promising data source for detecting beetle kill because it is possible, though challenging, to identify individual trees. However, the applicability of automated deep-learning approaches for regional-scale mapping has not been evaluated. Here, we assess a deep convolutional network for mapping dead spruce in high-resolution ($\sim$2 m) satellite imagery of Southcentral Alaska. The network identified dead spruce pixels across stand characteristics, achieving an average accuracy of 95%. To upscale to the stand scale, we mitigated overestimation of dead tree pixels at elevated severity by calibration. Stand-scale areal severity, the fraction of dead spruce pixels within a stand, was mapped with an RMSE of 0.02 at 90 m scale. The estimated severity exceeded 0.05 in fewer than 4% of the landscape, and approximately 90% of dead trees pixels were found in low-severity stands. Severity was weakly associated with stand-scale Landsat reflectance changes, a clear relation between SWIR reflectance change and severity only emerging above 0.1 severity. In conclusion, high-resolution satellite imagery are suited to automated mapping of beetle-associated kill at tree and stand scale across the severity spectrum. Such data products support forest and fire management and further understanding of the dynamics and consequences of beetle outbreaks.

Phenolic compounds constitute an essential part of the plant’s secondary metabolites and play a crucial role in ecosystem functioning, including nutrient cycling and plant defence against biotic and abiotic stressors. Quantifying the phenolic compounds across global biomes is important for monitoring the biological diversity and ecosystem processes. However, our understanding of foliar phenolic compounds remains limited, particularly regarding how they vary among temperate tree species and whether their variation and absorption features can be assessed using spectroscopy at the leaf level. In this study, we examined the relationships between the spectral properties of fresh leaves from temperate tree species and two ecologically important phenolic compounds (i.e., total phenol and tannin). We sampled the leaves of four dominant tree species (i.e., English oak, European beech, Norway spruce, and Scots pine) across two European temperate forest sites. Continuum removal was applied to the leaf spectra to enhance the assessment of the subtle absorption features that correlate with the phenolic content. Total phenol and tannin concentrations were estimated by comparing the performance of two empirical methods, namely partial least squares regression (PLSR) and Gaussian processes regression (GPR). Our results showed a large range of variation in total phenol and tannin between temperate tree species (p < 0.05). Spectral analysis revealed persistent and distinct phenolic absorption features near 1666 nm in the spectra of English oak, Norway spruce and European beech, whereas Scots pine exhibited a weaker absorption feature near 1653 nm. Regression results showed that both PLSR and GPR accurately estimated total phenol and tannin across temperate tree species, with informative bands for predicting these two traits well-corresponded between the two models utilised. Our results also suggested that total phenol was overall more accurately predicted than tannin regardless of employed methods. The most accurate estimations were achieved using PLSR with the continuum-removed SWIR spectra (total phenol: R²=0.79, NRMSE=9.95%; tannin: R²=0.59, NRMSE=14.53%). Testing the models established for individual species or forest types revealed variability in their prediction performances, with these specific models demonstrating lower accuracy (R²=0.47–0.69 and 0.34–0.54 for total phenol and tannin, respectively) compared to the cross-species model. Our study extends the understanding of absorption features of phenolic compounds in common temperate tree species and demonstrates the potential for a generalised spectroscopy model to predict foliar phenolic compounds across temperate forests. These findings provide a foundation for mapping and monitoring phenolic compounds in temperate forests at the canopy level using airborne and spaceborne imaging spectroscopy.

The rapid development of augmented reality (AR), 3D reconstruction, simultaneous localization and mapping (SLAM), and autonomous driving requires off-the-shelf camera calibration solutions that are adaptable to cameras of different configurations in different complex scenarios. To this end, we propose a generic, robust, and accurate camera calibration framework, called Meta-Calib, by using single or multiple novel designed ArUco-encoded meta-board(s), which is dedicated to estimate accurate camera intrinsic parameters and extrinsic transformations of different multi-camera configurations. The ArUco calibration board has been redesigned to facilitate learning-based robust detection and obtain higher precision control point coordinates, which is termed the meta-board. This completely replaces the widely-used chessboard based on the corner extraction scheme to greatly alleviate the impact of image distortion on control points, especially when it is located at the boundary area of the fish-eye camera. A robust two-stage deep learning detection strategy is applied to reliably localize the ArUco-encoded inner coding region of the meta-board followed by identifying two categories of circular shapes representing “0” and “1” encoded in the ArUco pattern for decoding and orientation determination. The center points of circular shapes on the meta-board in the distorted image taken under the perspective view can be approximated through elliptical fitting with contour edges. The deviation between the fitting center points and ground-truth can be greatly suppressed when the refined sub-pixel contour edges extracted on the original image are projected to the orthographic projection view based on the camera intrinsic parameters, distortion coefficients and the prior information of the meta-board. Based on this observation, we propose a systematic iterative refinement approach to achieve the high-precision intrinsic calibration of a camera. This process involves improving the estimation of camera intrinsic parameters and fitting the center control points of circular shapes on the meta-boards in an iterative manner. The progressive nature of our approach permits reliably calibrate large distortion camera models under the presence of noisy measurements, which ensures good convergence. In addition, we also propose a graph-based multi-camera extrinsic calibration method via the corrected control points to reliably estimate both the relative poses of the meta-boards and cameras in the multi-camera system. The proposed method is not constrained by the number of cameras and meta-boards used, which makes our strategy accessible even with inflexible computer vision experts. Furthermore, we have derived the mathematical form for computing the covariance of the extrinsic transformation, which makes it possible to evaluate the uncertainty of the calibration results. Extensive experiments on a large number of both real and synthetic datasets, including perspective, fi