Search

Landslides, resulting from disturbances in slope equilibrium, pose a significant threat to landscapes, infrastructure, and human life. Triggered by factors such as intense precipitation, seismic activities, or volcanic eruptions, these events can cause extensive damage and endanger nearby communities. A comprehensive understanding of landslide characteristics, including spatio-temporal patterns, dimensions, and morphology, is vital for effective landslide disaster management. Existing remote sensing approaches mostly use either optical or synthetic aperture radar sensors. Integrating information from both these types of sensors promises greater accuracy for identifying and locating landslides. This study proposes a novel approach, the ML-LaDeCORsat (Machine Learning-based coseismic Landslide Detection using Combined Optical and Radar Satellite Imagery), that integrates freely available Sentinel-1, Palsar-2, and Sentinel-2 imagery data in Google Earth Engine (GEE). The approach also integrates relevant spectral indices and suitable bands used in a machine learning-based classification of coseismic landslides. The approach includes a robust and reproducible training and validation strategy and allows one to choose between five classifiers (CART, Random Forest, GTB, SVM, and Naive Bayes). Using landslides from four different earthquake case studies, we demonstrate the superiority of our approach over existing solutions in coseismic landslide identification and localization, providing a GTB-based detection accuracy of 87–92%. ML-LaDeCORsat can be adapted to other landslide events (GEE script is provided). Transfer learning experiments proved that our model can be applied to other coseismic landslide events without the need for additional training data. Our novel approach therefore facilitates quick and reliable identification of coseismic landslides, highlighting its potential to contribute towards more effective disaster management. [ABSTRACT FROM AUTHOR]

With the rapidly growing interest in using structural timber, a need exists to inspect and assess these structures using non-destructive testing (NDT). This review article summarizes NDT methods for wood inspection. After an overview of the most important NDT methods currently used, a detailed review of Ground Penetrating Radar (GPR) and Ultrasonic Testing (UST) is presented. These two techniques can be applied in situ and produce useful visual representations for quantitative assessments and damage detection. With its commercial availability and portability, GPR can help rapidly identify critical features such as moisture, voids, and metal connectors in wood structures. UST, which effectively detects deep cracks, delaminations, and variations in ultrasonic wave velocity related to moisture content, complements GPR's capabilities. The non-destructive nature of both techniques preserves the structural integrity of timber, enabling thorough assessments without compromising integrity and durability. Techniques such as the Synthetic Aperture Focusing Technique (SAFT) and Total Focusing Method (TFM) allow for reconstructing images that an inspector can readily interpret for quantitative assessment. The development of new sensors, instruments, and analysis techniques has continued to improve the application of GPR and UST on wood. However, due to the hon-homogeneous anisotropic properties of this complex material, challenges remain to quantify defects and characterize inclusions reliably and accurately. By integrating advanced imaging algorithms that consider the material's complex properties, combining measurements with simulations, and employing machine learning techniques, the implementation and application of GPR and UST imaging and damage detection for wood structures can be further advanced. [ABSTRACT FROM AUTHOR]

Land cover classification using high-resolution Polarimetric Synthetic Aperture Radar (PolSAR) images obtained from satellites is a challenging task. While deep learning algorithms have been extensively studied for PolSAR image land cover classification, the performance is severely constrained due to the scarcity of labeled PolSAR samples and the limited domain acceptance of models. Recently, the emergence of the Segment Anything Model (SAM) based on the vision transformer (VIT) model has brought about a revolution in the study of specific downstream tasks in computer vision. Benefiting from its millions of parameters and extensive training datasets, SAM demonstrates powerful capabilities in extracting semantic information and generalization. To this end, we propose a dual-stream feature extraction network based on SAM, i.e., BSDSNet. We change the image encoder part of SAM to a dual stream, where the ConvNext image encoder is utilized to extract local information and the VIT image encoder is used to extract global information. BSDSNet achieves an in-depth exploration of semantic and spatial information in PolSAR images. Additionally, to facilitate a fine-grained amalgamation of information, the SA-Gate module is employed to integrate local–global information. Compared to previous deep learning models, BSDSNet's impressive ability to represent features is akin to a versatile receptive field, making it well suited for classifying PolSAR images across various resolutions. Comprehensive evaluations indicate that BSDSNet achieves excellent results in qualitative and quantitative evaluation when performing classification tasks on the AIR-PolSAR-Seg dataset and the WHU-OPT-SAR dataset. Compared to the suboptimal results, our method improves the Kappa metric by 3.68% and 0.44% on the AIR-PolSAR-Seg dataset and the WHU-OPT-SAR dataset, respectively. [ABSTRACT FROM AUTHOR]

Supervised machine learning algorithms usually require huge labeled data sets to produce sufficiently good results. For many applications, these data sets are still not available today, and the reasons for this can be manifold. As a solution, the missing training data can be generated by fast simulators. This procedure is well studied and allows filling possible gaps in the training data, which can further improve the results of a machine learning model. For this reason, this article deals with the development of a two-dimensional electromagnetic field simulator for modeling the response of a radar sensor in an imaging system based on the synthetic aperture radar principle. The creation of completely random scenes is essential to achieve data sets with large variance. Therefore, special emphasis is placed on the development of methods that allow creating random objects, which can then be assembled into an entire scene. In the context of this contribution, we focus on humanitarian demining with regard to improvised explosive devices using a ground-penetrating radar system. This is an area where the use of trained classifiers is of great importance, but in practice, there are little to no labeled datasets for the training process. The simulation results show good agreement with the measurement results obtained in a previous contribution, demonstrating the possibility of enhancing sparse training data sets with synthetic data. [ABSTRACT FROM AUTHOR]

In polarimetric synthetic aperture radar (POLSAR), it is of great significance for civil and military applications to find novel model-based decomposition methods suitable for ship detection in different detection backgrounds. Based on the physical interpretation of polarimetric decomposition theory and the Lasso rule for sparse features, we propose a four-component decomposition model, which is composed of surface scattering (Odd), double-bounce scattering (Dbl), volume scattering (Vol), and ±45° oriented dipole (Od). In principle, the Od component can describe the compounded scattering structure of a ship consisting of odd-bounce and even-bounce reflectors. Moreover, the pocket perceptron learning algorithm (PPLA) and support vector machine (SVM) are utilized to solve the linear inseparable problems in this study. Using large amounts of RADARSAT-2 (RS-2) fully polarized SAR data and AIRSAR data, our experimental results show that the Od component can make a great contribution to ship detection. Compared with other conventional decomposition methods used in the experiments, the proposed four-component decomposition method has better performance and is more effective and feasible to detect ships. [ABSTRACT FROM AUTHOR]

The accurate estimation of forest aboveground biomass (AGB) in areas with complex terrain is very important for quantifying the carbon sequestration capacity of forest ecosystems and studying the regional or global carbon cycle. In our previous research, we proposed the radiometric terrain correction (RTC) process for introducing normalized correction factors, which has strong effectiveness and robustness in terms of the backscattering coefficient of polarimetric synthetic aperture radar (PolSAR) data and the monadic model. However, the impact of RTC on the correctness of feature extraction and the performance of regression models requires further exploration in the retrieval of forest AGB based on a machine learning multiple regression model. In this study, based on PolSAR data provided by ALOS-2, 117 feature variables were accurately extracted using the RTC process, and then Boruta and recursive feature elimination with cross-validation (RFECV) algorithms were used to perform multi-step feature selection. Finally, 10 machine learning regression models and the Optuna algorithm were used to evaluate the effectiveness and robustness of RTC in improving the quality of the PolSAR feature set and the performance of the regression models. The results revealed that, compared with the situation without RTC treatment, RTC can effectively and robustly improve the accuracy of PolSAR features (the Pearson correlation R between the PolSAR features and measured forest AGB increased by 0.26 on average) and the performance of regression models (the coefficient of determination R2 increased by 0.14 on average, and the rRMSE decreased by 4.20% on average), but there is a certain degree of overcorrection in the RTC process. In addition, in situations where the data exhibit linear relationships, linear models remain a powerful and practical choice due to their efficient and stable characteristics. For example, the optimal regression model in this study is the Bayesian Ridge linear regression model (R2 = 0.82, rRMSE = 18.06%). [ABSTRACT FROM AUTHOR]