Sequential Simulation for Modeling Geological Structures from Training Images

2006 ◽  
pp. 139-149 ◽  
Keyword(s):  
Sequential Simulation ◽  
Geological Structures ◽  
Training Images

CNN performance dependence on linear image processing

2020 ◽  
Vol 2020 (10) ◽  
pp. 310-1-310-7
Author(s):  
Khalid Omer ◽  
Luca Caucci ◽  
Meredith Kupinski
Keyword(s):  
Image Processing ◽  
Texture Classification ◽  
Full Rank ◽  
Detection Performance ◽  
Ideal Observer ◽  
Training Data ◽  
Image Texture ◽  
Training Images ◽  
Analytic Expressions ◽  
Linear Compression

This work reports on convolutional neural network (CNN) performance on an image texture classification task as a function of linear image processing and number of training images. Detection performance of single and multi-layer CNNs (sCNN/mCNN) are compared to optimal observers. Performance is quantified by the area under the receiver operating characteristic (ROC) curve, also known as the AUC. For perfect detection AUC = 1.0 and AUC = 0.5 for guessing. The Ideal Observer (IO) maximizes AUC but is prohibitive in practice because it depends on high-dimensional image likelihoods. The IO performance is invariant to any fullrank, invertible linear image processing. This work demonstrates the existence of full-rank, invertible linear transforms that can degrade both sCNN and mCNN even in the limit of large quantities of training data. A subsequent invertible linear transform changes the images’ correlation structure again and can improve this AUC. Stationary textures sampled from zero mean and unequal covariance Gaussian distributions allow closed-form analytic expressions for the IO and optimal linear compression. Linear compression is a mitigation technique for high-dimension low sample size (HDLSS) applications. By definition, compression strictly decreases or maintains IO detection performance. For small quantities of training data, linear image compression prior to the sCNN architecture can increase AUC from 0.56 to 0.93. Results indicate an optimal compression ratio for CNN based on task difficulty, compression method, and number of training images.

Morphotectonic Analysis of Tanjung Bungo Area Based on Geological Structure Control, Central Sumatera Basin

2020 ◽  
Author(s):  
D., A., L., A. Putri
Keyword(s):  
Natural Resources ◽  
River Flow ◽  
Measurement Data ◽  
River System ◽  
Flow Patterns ◽  
Source Rocks ◽  
Aerial Photographs ◽  
Morphological Data ◽  
Tectonic Activity ◽  
Geological Structures

Tectonic activity in an area could result in various impacts such as changes in elevation, level of slope percentages, river flow patterns and systems, and the formation of geological structures both locally and regionally, which will form a new landscape. The tectonic activity also affects the stratigraphic sequences of the area. Therefore, it is necessary to study morphotectonic or landscape forms that are influenced by active tectonic activities, both those occur recently and in the past. These geological results help provide information of the potential of natural resources in and around Tanjung Bungo area. Morphological data are based on three main aspects including morphogenesis, morphometry, and morphography. The data are collected in two ways, the first is field survey by directly observing and taking field data such as measuring geological structures, rock positions, and outcrop profiles. The second way is to interpret them through Digital Elevation Model (DEM) and aerial photographs by analyzing river flow patterns and lineament analysis. The field measurement data are processed using WinTensor, Dips, and SedLog Software. The supporting data such as Topographic Maps, Morphological Elevation Maps, Slope Maps, Flow Pattern Maps, and Lineament Maps are based on DEM data and are processed using ArcGis Software 10.6.1 and PCI Geomatica. Morphotectonically, the Tanjung Bungo area is at a moderate to high-class level of tectonic activity taken place actively resulted in several joints, faults, and folds. The formation of geological structures has affected the morphological conditions of the area as seen from the development of steep slopes, structural flow patterns such as radial, rectangular, and dendritic, as well as illustrated by rough surface relief in Tanjung Bungo area. This area has the potential for oil and gas resources as indicated by the Telisa Formation, consisting of calcareous silts rich in planktonic and benthonic fossils, which may be source rocks and its contact with the Menggala Formation which is braided river system deposits that could be good reservoirs. Further research needs to be done since current research is only an interpretation of surface data. Current natural resources being exploited in Tanjung Bungo region are coals. The coals have thicknesses of 5-7 cm and are classified as bituminous coals.

scholarly journals Transboundary geophysical mapping of geological elements and salinity distribution critical for the assessment of future sea water intrusion in response to sea level rise

2012 ◽  
Vol 16 (7) ◽  
pp. 1845-1862 ◽  
Author(s):  
F. Jørgensen ◽  
W. Scheer ◽  
S. Thomsen ◽  
T. O. Sonnenborg ◽  
K. Hinsby ◽  
...  
Keyword(s):  
Groundwater Flow ◽  
Sea Level Rise ◽  
Sea Level ◽  
Preferential Flow ◽  
Sea Water ◽  
Saline Water ◽  
Seismic Survey ◽  
Salinity Distribution ◽  
Geological Structures ◽  
The North

Abstract. Geophysical techniques are increasingly being used as tools for characterising the subsurface, and they are generally required to develop subsurface models that properly delineate the distribution of aquifers and aquitards, salt/freshwater interfaces, and geological structures that affect groundwater flow. In a study area covering 730 km2 across the border between Germany and Denmark, a combination of an airborne electromagnetic survey (performed with the SkyTEM system), a high-resolution seismic survey and borehole logging has been used in an integrated mapping of important geological, physical and chemical features of the subsurface. The spacing between flight lines is 200–250 m which gives a total of about 3200 line km. About 38 km of seismic lines have been collected. Faults bordering a graben structure, buried tunnel valleys, glaciotectonic thrust complexes, marine clay units, and sand aquifers are all examples of geological structures mapped by the geophysical data that control groundwater flow and to some extent hydrochemistry. Additionally, the data provide an excellent picture of the salinity distribution in the area and thus provide important information on the salt/freshwater boundary and the chemical status of groundwater. Although the westernmost part of the study area along the North Sea coast is saturated with saline water and the TEM data therefore are strongly influenced by the increased electrical conductivity there, buried valleys and other geological elements are still revealed. The mapped salinity distribution indicates preferential flow paths through and along specific geological structures within the area. The effects of a future sea level rise on the groundwater system and groundwater chemistry are discussed with special emphasis on the importance of knowing the existence, distribution and geometry of the mapped geological elements, and their control on the groundwater salinity distribution is assessed.

scholarly journals Dip angles of active faults from the surface to the seismogenic zone inferred from a 2D numerical analysis of visco-elasto-plastic models: a case study for the Osaka Plain

2021 ◽  
Vol 73 (1) ◽  
Author(s):  
Hayami Nishiwaki ◽  
Takamoto Okudaira ◽  
Kazuhiko Ishii ◽  
Muneki Mitamura
Keyword(s):  
Urban Areas ◽  
Active Faults ◽  
Three Dimensional ◽  
Seismogenic Zone ◽  
Earthquake Ground Motion ◽  
Convex Curve ◽  
Seismic Reflection Profile ◽  
Geological Structures ◽  
Dimensional Distribution ◽  
Dip Angle

AbstractThe geometries (i.e., dip angles) of active faults from the surface to the seismogenic zone are the most important factors used to evaluate earthquake ground motion, which is crucial for seismic hazard assessments in urban areas. In Osaka, a metropolitan city in Japan, there are several active faults (e.g., the Uemachi and Ikoma faults), which are inferred from the topography, the attitude of active faults in surface trenches, the seismic reflection profile at shallow depths (less than 2 km), and the three-dimensional distribution of the Quaternary sedimentary layers. The Uemachi and Ikoma faults are N–S-striking fault systems with total lengths of 42 km and 38 km, respectively, with the former being located ~ 12 km west of the latter; however, the geometries of each of the active faults within the seismogenic zone are not clear. In this study, to examine the geometries of the Uemachi and Ikoma faults from the surface to the seismogenic zone, we analyze the development of the geological structures of sedimentary layers based on numerical simulations of a two-dimensional visco-elasto-plastic body under a horizontal compressive stress field, including preexisting high-strained weak zones (i.e., faults) and surface sedimentation processes, and evaluate the relationship between the observed geological structures of the Quaternary sediments (i.e., the Osaka Group) in the Osaka Plain and the model results. As a result, we propose geometries of the Uemachi and Ikoma faults from the surface to the seismogenic zone. When the friction coefficient of the faults is ~ 0.5, the dip angles of the Uemachi and Ikoma faults near the surface are ~ 30°–40° and the Uemachi fault has a downward convex curve at the bottom of the seismogenic zone, but does not converge to the Ikoma fault. Based on the analysis in this study, the dip angle of the Uemachi fault zone is estimated to be approximately 30°–40°, which is lower than that estimated in the previous studies. If the active fault has a low angle, the width of the fault plane is long, and thus the estimated seismic moment will be large.

scholarly journals Cotton Stand Counting from Unmanned Aerial System Imagery Using MobileNet and CenterNet Deep Learning Models

2021 ◽  
Vol 13 (14) ◽  
pp. 2822
Author(s):  
Zhe Lin ◽  
Wenxuan Guo
Keyword(s):  
Deep Learning ◽  
Cotton Plant ◽  
Unmanned Aerial System ◽  
Learning Models ◽  
Training Images ◽  
Testing Dataset ◽  
Cotton Plants ◽  
Detection And Counting ◽  
Different Dimensions ◽  
The Mean

An accurate stand count is a prerequisite to determining the emergence rate, assessing seedling vigor, and facilitating site-specific management for optimal crop production. Traditional manual counting methods in stand assessment are labor intensive and time consuming for large-scale breeding programs or production field operations. This study aimed to apply two deep learning models, the MobileNet and CenterNet, to detect and count cotton plants at the seedling stage with unmanned aerial system (UAS) images. These models were trained with two datasets containing 400 and 900 images with variations in plant size and soil background brightness. The performance of these models was assessed with two testing datasets of different dimensions, testing dataset 1 with 300 by 400 pixels and testing dataset 2 with 250 by 1200 pixels. The model validation results showed that the mean average precision (mAP) and average recall (AR) were 79% and 73% for the CenterNet model, and 86% and 72% for the MobileNet model with 900 training images. The accuracy of cotton plant detection and counting was higher with testing dataset 1 for both CenterNet and MobileNet models. The results showed that the CenterNet model had a better overall performance for cotton plant detection and counting with 900 training images. The results also indicated that more training images are required when applying object detection models on images with different dimensions from training datasets. The mean absolute percentage error (MAPE), coefficient of determination (R2), and the root mean squared error (RMSE) values of the cotton plant counting were 0.07%, 0.98 and 0.37, respectively, with testing dataset 1 for the CenterNet model with 900 training images. Both MobileNet and CenterNet models have the potential to accurately and timely detect and count cotton plants based on high-resolution UAS images at the seedling stage. This study provides valuable information for selecting the right deep learning tools and the appropriate number of training images for object detection projects in agricultural applications.

scholarly journals Flood vulnerability assessment in the Accra Metropolis, southeastern Ghana

2021 ◽  
Vol 11 (7) ◽  
Author(s):  
Benjamin Wullobayi Dekongmen ◽  
Amos Tiereyangn Kabo-bah ◽  
Martin Kyereh Domfeh ◽  
Emmanuel Daanoba Sunkari ◽  
Yihun Taddele Dile ◽  
...  
Keyword(s):  
Drainage Density ◽  
Geological Structures ◽  
Flood Vulnerability ◽  
Rock Types ◽  
Digital Elevation ◽  
Annual Population ◽  
Density Values ◽  
High Runoff ◽  
Elevation Model ◽  
Very High

AbstractFloods in Ghana have become a perennial challenge in the major cities and communities located in low-lying areas. Therefore, cities and communities located in these areas have been classified as potential or natural flood-prone zones. In this study, the Digital Elevation Model (DEM) of the Accra Metropolis was used to assess the drainage density and elevation patterns of the area. The annual population estimation data and flood damages were assessed to understand the damages and population trend. This research focused primarily on the elevation patterns, slope patterns, and drainage density of the Accra Metropolis. Very high drainage density values, which range between 149 and 1117 m/m2, showed very high runoff converging areas. High drainage density was also found to be in the range of 1117–1702 m/m2, which defined the area as a high runoff converging point. The medium and low converging points of runoff were also found to be ranging between 1702–2563 m/m2 and 2563–4070 m/m2, respectively. About 32% of the study area is covered by natural flood-prone zones, whereas flood-prone zones also covered 33% and frequent flood zones represent 25%. Areas in the Accra Metropolis that fall in the Accraian and Togo series rock types experience high floods. However, the lineament networks (geological structures) that dominate the Dahomeyan series imply that the geological structures in the Dahomeyan series also channel the runoffs into the low-lying areas, thereby contributing to the perennial flooding in the Accra Metropolis.

scholarly journals Modelling Maritime Pine (Pinus pinaster Aiton) Spatial Distribution and Productivity in Portugal: Tools for Forest Management

Forests ◽  
2021 ◽  
Vol 12 (3) ◽  
pp. 368
Author(s):  
Cristina Alegria ◽  
Natália Roque ◽  
Teresa Albuquerque ◽  
Paulo Fernandez ◽  
Maria Margarida Ribeiro
Keyword(s):  
Spatial Distribution ◽  
Forest Management ◽  
Species Distribution ◽  
Pinus Pinaster ◽  
Landscape Planning ◽  
Environmental Data ◽  
Maritime Pine ◽  
Future Climate Change ◽  
Sequential Simulation ◽  
Model Species

Research Highlights: Modelling species’ distribution and productivity is key to support integrated landscape planning, species’ afforestation, and sustainable forest management. Background and Objectives: Maritime pine (Pinus pinaster Aiton) forests in Portugal were lately affected by wildfires and measures to overcome this situation are needed. The aims of this study were: (1) to model species’ spatial distribution and productivity using a machine learning (ML) regression approach to produce current species’ distribution and productivity maps; (2) to model the species’ spatial productivity using a stochastic sequential simulation approach to produce the species’ current productivity map; (3) to produce the species’ potential distribution map, by using a ML classification approach to define species’ ecological envelope thresholds; and (4) to identify present and future key factors for the species’ afforestation and management. Materials and Methods: Spatial land cover/land use data, inventory, and environmental data (climate, topography, and soil) were used in a coupled ML regression and stochastic sequential simulation approaches to model species’ current and potential distributions and productivity. Results: Maritime pine spatial distribution modelling by the ML approach provided 69% fitting efficiency, while species productivity modelling achieved only 43%. The species’ potential area covered 60% of the country’s area, where 78% of the species’ forest inventory plots (1995) were found. The change in the Maritime pine stands’ age structure observed in the last decades is causing the species’ recovery by natural regeneration to be at risk. Conclusions: The maps produced allow for best site identification for species afforestation, wood production regulation support, landscape planning considering species’ diversity, and fire hazard mitigation. These maps were obtained by modelling using environmental covariates, such as climate attributes, so their projection in future climate change scenarios can be performed.

Sign in / Sign up

Export Citation Format

Share Document