Reliable detection of clouds is a critical pre-processing step in optical satellite based remote sensing. Currently, most methods are based on classifying invidual pixels from their spectral ...signatures, therefore they do not incorporate the spatial patterns. This often leads to misclassifications of highly reflective surfaces, such as human made structures or snow/ice. Multi-temporal methods can be used to alleviate this problem, but these methods introduce new problems, such as the need of a cloud-free image of the scene. In this paper, we introduce the Remote Sensing Network (RS-Net), a deep learning model for detection of clouds in optical satellite imagery, based on the U-net architecture. The model is trained and evaluated using the Landsat 8 Biome and SPARCS datasets, and it shows state-of-the-art performance, especially over biomes with hardly distinguishable scenery, such as clouds over snowy and icy regions. In particular, the performance of the model that uses only the RGB bands is significantly improved, showing promising results for cloud detection with smaller satellites with limited multi-spectral capabilities. Furthermore, we show how training the RS-Net models on data from an existing cloud masking method, which are treated as noisy data, leads to increased performance compared to the original method. This is validated by using the Fmask algorithm to annotate the Landsat 8 datasets, and then use these annotations as training data for regularized RS-Net models, which then show improved performance compared to the Fmask algorithm. Finally, the classification time of a full Landsat 8 product is 18.0 ± 2.4 s for the largest RS-Net model, thereby making it suitable for production environments.
•This algorithm show high performance, even when limited to RGB bands.•The processing time is low due to GPU based calculations.•Cloud detection performance over snow/ice is significantly improved.•This algorithm can be trained on the output of earlier cloud detection algorithms.
Recent technological advances in remote sensing sensors and platforms, such as high-resolution satellite imagers or unmanned aerial vehicles (UAV), facilitate the availability of fine-grained earth ...observation data. Such data reveal vegetation canopies in high spatial detail. Efficient methods are needed to fully harness this unpreceded source of information for vegetation mapping. Deep learning algorithms such as Convolutional Neural Networks (CNN) are currently paving new avenues in the field of image analysis and computer vision. Using multiple datasets, we test a CNN-based segmentation approach (U-net) in combination with training data directly derived from visual interpretation of UAV-based high-resolution RGB imagery for fine-grained mapping of vegetation species and communities. We demonstrate that this approach indeed accurately segments and maps vegetation species and communities (at least 84% accuracy). The fact that we only used RGB imagery suggests that plant identification at very high spatial resolutions is facilitated through spatial patterns rather than spectral information. Accordingly, the presented approach is compatible with low-cost UAV systems that are easy to operate and thus applicable to a wide range of users.
This exploration primarily aims to jointly apply the local FCN (fully convolution neural network) and YOLO-v5 (You Only Look Once-v5) to the detection of small targets in remote sensing images. ...Firstly, the application effects of R-CNN (Region-Convolutional Neural Network), FRCN (Fast Region-Convolutional Neural Network), and R-FCN (Region-Based-Fully Convolutional Network) in image feature extraction are analyzed after introducing the relevant region proposal network. Secondly, YOLO-v5 algorithm is established on the basis of YOLO algorithm. Besides, the multi-scale anchor mechanism of Faster R-CNN is utilized to improve the detection ability of YOLO-v5 algorithm for small targets in the image in the process of image detection, and realize the high adaptability of YOLO-v5 algorithm to different sizes of images. Finally, the proposed detection method YOLO-v5 algorithm + R-FCN is compared with other algorithms in NWPU VHR-10 data set and Vaihingen data set. The experimental results show that the YOLO-v5 + R-FCN detection method has the optimal detection ability among many algorithms, especially for small targets in remote sensing images such as tennis courts, vehicles, and storage tanks. Moreover, the YOLO-v5 + R-FCN detection method can achieve high recall rates for different types of small targets. Furthermore, due to the deeper network architecture, the YOL v5 + R-FCN detection method has a stronger ability to extract the characteristics of image targets in the detection of remote sensing images. Meanwhile, it can achieve more accurate feature recognition and detection performance for the densely arranged target images in remote sensing images. This research can provide reference for the application of remote sensing technology in China, and promote the application of satellites for target detection tasks in related fields.
Terrestrial ecosystem and carbon cycle feedbacks will significantly impact future climate, but their responses are highly uncertain. Models and tipping point analyses suggest the tropics and ...arctic/boreal zone carbon–climate feedbacks could be disproportionately large. In situ observations in those regions are sparse, resulting in high uncertainties in carbon fluxes and fluxes. Key parameters controlling ecosystem carbon responses, such as plant traits, are also sparsely observed in the tropics, with the most diverse biome on the planet treated as a single type in models. We analyzed the spatial distribution of in situ data for carbon fluxes, stocks and plant traits globally and also evaluated the potential of remote sensing to observe these quantities. New satellite data products go beyond indices of greenness and can address spatial sampling gaps for specific ecosystem properties and parameters. Because environmental conditions and access limit in situ observations in tropical and arctic/boreal environments, use of space‐based techniques can reduce sampling bias and uncertainty about tipping point feedbacks to climate. To reliably detect change and develop the understanding of ecosystems needed for prediction, significantly, more data are required in critical regions. This need can best be met with a strategic combination of remote and in situ data, with satellite observations providing the dense sampling in space and time required to characterize the heterogeneity of ecosystem structure and function.
Recently proposed methods for weakly-supervised semantic segmentation have achieved impressive performance in predicting pixel classes despite being trained with only image labels which lack ...positional information. Because image annotations are cheaper and quicker to generate, weak supervision is more practical than full supervision for training segmentation algorithms. These methods have been predominantly developed to solve the background separation and partial segmentation problems presented by natural scene images and it is unclear whether they can be simply transferred to other domains with different characteristics, such as histopathology and satellite images, and still perform well. This paper evaluates state-of-the-art weakly-supervised semantic segmentation methods on natural scene, histopathology, and satellite image datasets and analyzes how to determine which method is most suitable for a given dataset. Our experiments indicate that histopathology and satellite images present a different set of problems for weakly-supervised semantic segmentation than natural scene images, such as ambiguous boundaries and class co-occurrence. Methods perform well for datasets they were developed on, but tend to perform poorly on other datasets. We present some practical techniques for these methods on unseen datasets and argue that more work is needed for a generalizable approach to weakly-supervised semantic segmentation. Our full code implementation is available on GitHub:
https://github.com/lyndonchan/wsss-analysis
.
The past and future of global river ice Yang, Xiao; Pavelsky, Tamlin M; Allen, George H
Nature (London),
01/2020, Letnik:
577, Številka:
7788
Journal Article
Recenzirano
Odprti dostop
More than one-third of Earth's landmass is drained by rivers that seasonally freeze over. Ice transforms the hydrologic
, ecologic
, climatic
and socio-economic
functions of river corridors. Although ...river ice extent has been shown to be declining in many regions of the world
, the seasonality, historical change and predicted future changes in river ice extent and duration have not yet been quantified globally. Previous studies of river ice, which suggested that declines in extent and duration could be attributed to warming temperatures
, were based on data from sparse locations. Furthermore, existing projections of future ice extent are based solely on the location of the 0-°C isotherm
. Here, using satellite observations, we show that the global extent of river ice is declining, and we project a mean decrease in seasonal ice duration of 6.10 ± 0.08 days per 1-°C increase in global mean surface air temperature. We tracked the extent of river ice using over 400,000 clear-sky Landsat images spanning 1984-2018 and observed a mean decline of 2.5 percentage points globally in the past three decades. To project future changes in river ice extent, we developed an observationally calibrated and validated model, based on temperature and season, which reduced the mean bias by 87 per cent compared with the 0-degree-Celsius isotherm approach. We applied this model to future climate projections for 2080-2100: compared with 2009-2029, the average river ice duration declines by 16.7 days under Representative Concentration Pathway (RCP) 8.5, whereas under RCP 4.5 it declines on average by 7.3 days. Our results show that, globally, river ice is measurably declining and will continue to decline linearly with projected increases in surface air temperature towards the end of this century.