scholarly journals Innovative methods for earthquake damage detection and classification using airborne observation of critical infrastructures (project LOKI)

2021 ◽  
Author(s):  
Julia Kohns ◽  
Vivien Zahs ◽  
Tahira Ullah ◽  
Danijel Schorlemmer ◽  
Cecilia Nievas ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Resolution ◽  
Damage Detection ◽  
Emergency Response ◽  
Point Clouds ◽  
Mission Planning ◽  
Exposure Model ◽  
Critical Infrastructures ◽  
Innovative Methods ◽  
Building Level

<p>Earthquakes play a major role worldwide regarding economic and social consequences. In the event of an earthquake, many lives are at risk and the impact on the built and natural environment may be significant. Until now, estimations of damage and losses and the assessment of the stability of buildings are, however, only available several days to months after the event and are often based on the subjective assessment of experienced engineers.</p><p>For the effective planning of rescue measures and the best possible use of available resources, a fast, (semi-)automatic and accurate detection of the situation and an objective assessment of damage to critical infrastructures is indispensable. This requires a combination of innovative methods and technologies (UAVs, Machine Learning and Crowdsourcing combined with earthquake engineering knowledge) covering a wide range of spatial and temporal scales.</p><p>The interdisciplinary system LOKI (www.uni-heidelberg.de/loki) consists of the following procedure: After the occurrence of an earthquake, an initial damage forecast is made within a few minutes based on the Global Dynamic Exposure model and integrated vulnerability functions in combination with the ground-motion field to identify areas with potential high/low damage. Missing building footprints and required building information are recorded via a crowdsourcing approach to complete the OpenStreetMap building database, which serves as input to the exposure model. In parallel, mission plans for overview flights are created and transferred to fixed-wing UAVs, which record low to medium-resolution photos and 3D point clouds of the entire affected area. These data are used for damage detection, in which a binary distinction is made at building level between visible and non-visible damage using Machine Learning approaches. Thus, after a few hours, first orthophotos and the location of potentially damaged buildings can already be transmitted to emergency response teams. Thereafter, mission planning focuses on the capture of high-resolution 3D information of individual buildings. Fleets of multicopter drones provide highly detailed 3D imagery following mission plans that can be modified in real time by the emergency response teams. The mission planning algorithms support prioritization of specific areas or buildings for data acquisition, so that rescue measures can be optimally supported. The acquired high-resolution images and point clouds serve as input for damage classification, which is carried out per building using a combination of automatic procedures and Micro-Mapping. This offers the possibility to combine the advantages of fast automated procedures with the human ability to visually interpret details. Potential global and building material-related damage characteristics, which are based on observations of previous earthquakes, are included in a damage catalogue and allow building damage to be classified into five damage grades. In an iterative process, a timely and objective building-level classification of damage with an indication of the reliability of the specified degree of damage is achieved.</p><p>The integration of various disciplines and the combination of different concepts and technologies allows supporting disaster relief in different temporal and spatial resolutions with timely and reliable information on earthquake-induced damage.</p>

scholarly journals Supervised Distributed Multi-Instance and Unsupervised Single-Instance Autoencoder Machine Learning for Damage Diagnostics with High-Dimensional Data—A Hybrid Approach and Comparison Study

Computers ◽  
2021 ◽  
Vol 10 (3) ◽  
pp. 34
Author(s):  
Stefan Bosse ◽  
Dennis Weiss ◽  
Daniel Schmidt
Keyword(s):  
Machine Learning ◽  
High Resolution ◽  
Damage Detection ◽  
High Reliability ◽  
Hybrid Approach ◽  
Sensor Data ◽  
Detection Accuracy ◽  
Low Resource ◽  
Damage Diagnostics ◽  
Single Instance

Structural health monitoring (SHM) is a promising technique for in-service inspection of technical structures in a broad field of applications in order to reduce maintenance efforts as well as the overall structural weight. SHM is basically an inverse problem deriving physical properties such as damages or material inhomogeneity (target features) from sensor data. Often models defining the relationship between predictable features and sensors are required but not available. The main objective of this work is the investigation of model-free distributed machine learning (DML) for damage diagnostics under resource and failure constraints by using multi-instance ensemble and model fusion strategies and featuring improved scaling and stability compared with centralised single-instance approaches. The diagnostic system delivers two features: A binary damage classification (damaged or non-damaged) and an estimation of the spatial damage position in case of a damaged structure. The proposed damage diagnostics architecture should be able to be used in low-resource sensor networks with soft real-time capabilities. Two different machine learning methodologies and architectures are evaluated and compared posing low- and high-resolution sensor processing for low- and high-resolution damage diagnostics, i.e., a dedicated supervised trained low-resource and an unsupervised trained high-resource deep learning approach, respectively. In both architectures state-based recurrent artificial neural networks are used that process spatially and time-resolved sensor data from experimental ultrasonic guided wave measurements of a hybrid material (carbon fibre laminate) plate with pseudo defects. Finally, both architectures can be fused to a hybrid architecture with improved damage detection accuracy and reliability. An extensive evaluation of the damage prediction by both systems shows high reliability and accuracy of damage detection and localisation, even by the distributed multi-instance architecture with a resolution in the order of the sensor distance.

Innovative Methods for High Resolution Imaging

2012 ◽  
Author(s):  
T. C. Torgersen ◽  
V. P. Pauca ◽  
R. J. Plemmons
Keyword(s):  
High Resolution ◽  
High Resolution Imaging ◽  
Innovative Methods ◽  
Resolution Imaging

scholarly journals AI-Driven Framework for Recognition of Guava Plant Diseases through Machine Learning from DSLR Camera Sensor Based High Resolution Imagery

Sensors ◽  
2021 ◽  
Vol 21 (11) ◽  
pp. 3830
Author(s):  
Ahmad Almadhor ◽  
Hafiz Tayyab Rauf ◽  
Muhammad Ikram Ullah Lali ◽  
Robertas Damaševičius ◽  
Bader Alouffi ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Resolution ◽  
Color Difference ◽  
Plant Diseases ◽  
Economic Losses ◽  
Production Loss ◽  
Postharvest Diseases ◽  
Tropical Regions ◽  
Guava Fruit ◽  
Tree Classifier

Plant diseases can cause a considerable reduction in the quality and number of agricultural products. Guava, well known to be the tropics’ apple, is one significant fruit cultivated in tropical regions. It is attacked by 177 pathogens, including 167 fungal and others such as bacterial, algal, and nematodes. In addition, postharvest diseases may cause crucial production loss. Due to minor variations in various guava disease symptoms, an expert opinion is required for disease analysis. Improper diagnosis may cause economic losses to farmers’ improper use of pesticides. Automatic detection of diseases in plants once they emerge on the plants’ leaves and fruit is required to maintain high crop fields. In this paper, an artificial intelligence (AI) driven framework is presented to detect and classify the most common guava plant diseases. The proposed framework employs the ΔE color difference image segmentation to segregate the areas infected by the disease. Furthermore, color (RGB, HSV) histogram and textural (LBP) features are applied to extract rich, informative feature vectors. The combination of color and textural features are used to identify and attain similar outcomes compared to individual channels, while disease recognition is performed by employing advanced machine-learning classifiers (Fine KNN, Complex Tree, Boosted Tree, Bagged Tree, Cubic SVM). The proposed framework is evaluated on a high-resolution (18 MP) image dataset of guava leaves and fruit. The best recognition results were obtained by Bagged Tree classifier on a set of RGB, HSV, and LBP features (99% accuracy in recognizing four guava fruit diseases (Canker, Mummification, Dot, and Rust) against healthy fruit). The proposed framework may help the farmers to avoid possible production loss by taking early precautions.

scholarly journals Identifying Urban Poverty Using High-Resolution Satellite Imagery and Machine Learning Approaches: Implications for Housing Inequality

Land ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 648
Author(s):  
Guie Li ◽  
Zhongliang Cai ◽  
Yun Qian ◽  
Fei Chen
Keyword(s):  
Machine Learning ◽  
High Resolution ◽  
Satellite Imagery ◽  
Low Income ◽  
Urban Poverty ◽  
Image Features ◽  
Housing Affordability ◽  
Learning Approaches ◽  
Housing Inequality ◽  
High Resolution Satellite Imagery

Enriching Asian perspectives on the rapid identification of urban poverty and its implications for housing inequality, this paper contributes empirical evidence about the utility of image features derived from high-resolution satellite imagery and machine learning approaches for identifying urban poverty in China at the community level. For the case of the Jiangxia District and Huangpi District of Wuhan, image features, including perimeter, line segment detector (LSD), Hough transform, gray-level cooccurrence matrix (GLCM), histogram of oriented gradients (HoG), and local binary patterns (LBP), are calculated, and four machine learning approaches and 25 variables are applied to identify urban poverty and relatively important variables. The results show that image features and machine learning approaches can be used to identify urban poverty with the best model performance with a coefficient of determination, R2, of 0.5341 and 0.5324 for Jiangxia and Huangpi, respectively, although some differences exist among the approaches and study areas. The importance of each variable differs for each approach and study area; however, the relatively important variables are similar. In particular, four variables achieved relatively satisfactory prediction results for all models and presented obvious differences in varying communities with different poverty levels. Housing inequality within low-income neighborhoods, which is a response to gaps in wealth, income, and housing affordability among social groups, is an important manifestation of urban poverty. Policy makers can implement these findings to rapidly identify urban poverty, and the findings have potential applications for addressing housing inequality and proving the rationality of urban planning for building a sustainable society.

scholarly journals Using U-Net-Like Deep Convolutional Neural Networks for Precise Tree Recognition in Very High Resolution RGB (Red, Green, Blue) Satellite Images

Forests ◽  
2021 ◽  
Vol 12 (1) ◽  
pp. 66
Author(s):  
Kirill A. Korznikov ◽  
Dmitry E. Kislov ◽  
Jan Altman ◽  
Jiří Doležal ◽  
Anna S. Vozmishcheva ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Resolution ◽  
Satellite Images ◽  
Russian Far East ◽  
Pinus Koraiensis ◽  
Accuracy Score ◽  
Learning Approaches ◽  
Individual Tree ◽  
Populus Suaveolens ◽  
Very High

Very high resolution satellite imageries provide an excellent foundation for precise mapping of plant communities and even single plants. We aim to perform individual tree recognition on the basis of very high resolution RGB (red, green, blue) satellite images using deep learning approaches for northern temperate mixed forests in the Primorsky Region of the Russian Far East. We used a pansharpened satellite RGB image by GeoEye-1 with a spatial resolution of 0.46 m/pixel, obtained in late April 2019. We parametrized the standard U-Net convolutional neural network (CNN) and trained it in manually delineated satellite images to solve the satellite image segmentation problem. For comparison purposes, we also applied standard pixel-based classification algorithms, such as random forest, k-nearest neighbor classifier, naive Bayes classifier, and quadratic discrimination. Pattern-specific features based on grey level co-occurrence matrices (GLCM) were computed to improve the recognition ability of standard machine learning methods. The U-Net-like CNN allowed us to obtain precise recognition of Mongolian poplar (Populus suaveolens Fisch. ex Loudon s.l.) and evergreen coniferous trees (Abies holophylla Maxim., Pinus koraiensis Siebold & Zucc.). We were able to distinguish species belonging to either poplar or coniferous groups but were unable to separate species within the same group (i.e. A. holophylla and P. koraiensis were not distinguishable). The accuracy of recognition was estimated by several metrics and exceeded values obtained for standard machine learning approaches. In contrast to pixel-based recognition algorithms, the U-Net-like CNN does not lead to an increase in false-positive decisions when facing green-colored objects that are similar to trees. By means of U-Net-like CNN, we obtained a mean accuracy score of up to 0.96 in our computational experiments. The U-Net-like CNN recognizes tree crowns not as a set of pixels with known RGB intensities but as spatial objects with a specific geometry and pattern. This CNN’s specific feature excludes misclassifications related to objects of similar colors as objects of interest. We highlight that utilization of satellite images obtained within the suitable phenological season is of high importance for successful tree recognition. The suitability of the phenological season is conceptualized as a group of conditions providing highlighting objects of interest over other components of vegetation cover. In our case, the use of satellite images captured in mid-spring allowed us to recognize evergreen fir and pine trees as the first class of objects (“conifers”) and poplars as the second class, which were in a leafless state among other deciduous tree species.

Interpretable machine learning model to detect chemically adulterated urine samples analyzed by high resolution mass spectrometry

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Gabriel L. Streun ◽  
Andrea E. Steuer ◽  
Lars C. Ebert ◽  
Akos Dobay ◽  
Thomas Kraemer
Keyword(s):  
Machine Learning ◽  
Mass Spectrometry ◽  
Liquid Chromatography ◽  
High Resolution ◽  
Retention Time ◽  
Drug Testing ◽  
High Resolution Mass Spectrometry ◽  
Urine Samples ◽  
Machine Learning Model ◽  
Resolution Mass

Abstract Objectives Urine sample manipulation including substitution, dilution, and chemical adulteration is a continuing challenge for workplace drug testing, abstinence control, and doping control laboratories. The simultaneous detection of sample manipulation and prohibited drugs within one single analytical measurement would be highly advantageous. Machine learning algorithms are able to learn from existing datasets and predict outcomes of new data, which are unknown to the model. Methods Authentic human urine samples were treated with pyridinium chlorochromate, potassium nitrite, hydrogen peroxide, iodine, sodium hypochlorite, and water as control. In total, 702 samples, measured with liquid chromatography coupled to quadrupole time-of-flight mass spectrometry, were used. After retention time alignment within Progenesis QI, an artificial neural network was trained with 500 samples, each featuring 33,448 values. The feature importance was analyzed with the local interpretable model-agnostic explanations approach. Results Following 10-fold cross-validation, the mean sensitivity, specificity, positive predictive value, and negative predictive value was 88.9, 92.0, 91.9, and 89.2%, respectively. A diverse test set (n=202) containing treated and untreated urine samples could be correctly classified with an accuracy of 95.4%. In addition, 14 important features and four potential biomarkers were extracted. Conclusions With interpretable retention time aligned liquid chromatography high-resolution mass spectrometry data, a reliable machine learning model could be established that rapidly uncovers chemical urine manipulation. The incorporation of our model into routine clinical or forensic analysis allows simultaneous LC-MS analysis and sample integrity testing in one run, thus revolutionizing this field of drug testing.

scholarly journals Evaluation of UAV LiDAR for Mapping Coastal Environments

2019 ◽  
Vol 11 (24) ◽  
pp. 2893 ◽  
Author(s):  
Yi-Chun Lin ◽  
Yi-Ting Cheng ◽  
Tian Zhou ◽  
Radhika Ravi ◽  
Seyyed Hasheminasab ◽  
...  
Keyword(s):  
High Resolution ◽  
Lake Michigan ◽  
Point Clouds ◽  
Volume Loss ◽  
Coastal Environments ◽  
Average Volume ◽  
Survey Period ◽  
Shoreline Changes ◽  
Point Density ◽  
One Year

Unmanned Aerial Vehicle (UAV)-based remote sensing techniques have demonstrated great potential for monitoring rapid shoreline changes. With image-based approaches utilizing Structure from Motion (SfM), high-resolution Digital Surface Models (DSM), and orthophotos can be generated efficiently using UAV imagery. However, image-based mapping yields relatively poor results in low textured areas as compared to those from LiDAR. This study demonstrates the applicability of UAV LiDAR for mapping coastal environments. A custom-built UAV-based mobile mapping system is used to simultaneously collect LiDAR and imagery data. The quality of LiDAR, as well as image-based point clouds, are investigated and compared over different geomorphic environments in terms of their point density, relative and absolute accuracy, and area coverage. The results suggest that both UAV LiDAR and image-based techniques provide high-resolution and high-quality topographic data, and the point clouds generated by both techniques are compatible within a 5 to 10 cm range. UAV LiDAR has a clear advantage in terms of large and uniform ground coverage over different geomorphic environments, higher point density, and ability to penetrate through vegetation to capture points below the canopy. Furthermore, UAV LiDAR-based data acquisitions are assessed for their applicability in monitoring shoreline changes over two actively eroding sandy beaches along southern Lake Michigan, Dune Acres, and Beverly Shores, through repeated field surveys. The results indicate a considerable volume loss and ridge point retreat over an extended period of one year (May 2018 to May 2019) as well as a short storm-induced period of one month (November 2018 to December 2018). The foredune ridge recession ranges from 0 m to 9 m. The average volume loss at Dune Acres is 18.2 cubic meters per meter and 12.2 cubic meters per meter within the one-year period and storm-induced period, respectively, highlighting the importance of episodic events in coastline changes. The average volume loss at Beverly Shores is 2.8 cubic meters per meter and 2.6 cubic meters per meter within the survey period and storm-induced period, respectively.

Sign in / Sign up

Export Citation Format

Share Document