GEOSCAN Search Results: Fastlink


TitleAssessment of convolution neural networks for surficial geology mapping in the South Rae geological region, Northwest Territories, Canada
AuthorLatifovic, R; Pouliot, D; Campbell, J
SourceRemote Sensing vol. 10, no. 2, 307, 2018 p. 1-19, (Open Access)
Alt SeriesNatural Resources Canada, Contribution Series 20170360
PublisherMDPI AG
Mediaon-line; digital
File formatpdf; html
ProvinceNorthwest Territories
NTS74O/13; 74O/14; 74O/15; 74O/16; 75A/04; 75A/05; 75A/12; 75A/13; 75B; 75C/01; 75C/08; 75C/09; 75C/16; 75G/01; 75G/02; 75G/03; 75G/04
AreaAbitau Lake
Lat/Long WENS-108.4000 -105.5725 61.1833 59.8167
Subjectssurficial geology/geomorphology; Science and Technology; Nature and Environment; remote sensing; satellite imagery; photogrammetric techniques; airphoto interpretation; alluvial fans; organic deposits; bogs; glacial deposits; tills; ice contact deposits; bedrock geology; lithology; metamorphic rocks; models; Archean; Rae Province; Canadian Shield; geological mapping techniques; convolution neural networks; predictive mapping; Landsat; digital elevation data; digital elevation models; till blanket; till veneer; glaciofluvial sediments; glaciofluvial terraced sediments; glaciofluvial hummocky sediments; esker sediments; machine learning; alluvial sediments; alluvial floodplain sediments; alluvial terraced sediments; glaciofluvial blanket; glaciofluvial outwash fan sediments; glaciofluvial outwash plain sediments; glaciofluvial veneer; glaciolacustrine blanket; glaciolacustrine deltaic sediments; glaciolacustrine beach sediments; hummocky tills; Random Forest method; Phanerozoic; Cenozoic; Quaternary; Precambrian
Illustrationslocation maps; geoscientific sketch maps; satellite images; digital elevation models; aerial photographs; tables; flow diagrams; plots
ProgramGEM2: Geo-mapping for Energy and Minerals, TransGEM
Released2018 02 16
AbstractMapping of surficial (earth) materials is an important requirement for broadening the geoscience database of northern Canada. Surficial materials maps are an integral data source for mineral and energy exploration. Moreover, they provide information such as the location of gravels and sands, which are important for infrastructure development. Currently, surficial geology maps are produced through expert interpretation of aerial photography and field data. However, interpretation is known to be subjective, labour-intensive and difficult to repeat. The expert knowledge required for interpretation can be challenging to maintain and transfer. In this research, we seek to assess the potential of deep neural networks to aid surficial materials mapping by providing an objective initial layer that experts can modify to speed map development and improve consistency between mapped areas. Such an approach may also harness expert knowledge in a way that is transferable to unmapped areas. For this purpose, we assess the ability of convolution neural networks (CNN) to predict surficial materials classes under two sampling scenarios. In the first scenario, a CNN uses samples collected over the area to be mapped. In the second, a CNN trained over one area is then applied to locations where the available samples were not used in training the network. The latter case is important, as a collection of in-situ training data can be costly. The evaluation of the CNN was carried out using aerial photos, Landsat reflectance, and high-resolution digital elevation data over five areas within the South Rae geological region of Northwest Territories, Canada. The results are encouraging, with CNN generating average accuracy of 76% when locally trained. For CNN extended beyond training areas (i.e. trained over one area and applied over other), accuracy dropped to 59-70% depending on the classes selected for mapping. In the South Rae region, significant confusion was found between till veneer and till blanket as well as glaciofluvial subclasses (esker, terraced, and hummocky ice-contact). Merging these classes respectively increased accuracy for CNN used in extension to 68% on average. Relative to the more widely used Random Forest machine learning algorithm, this represents an improvement in accuracy of 4%. Furthermore, the CNN produced better results for less frequent classes with distinct spatial structure.
Summary(Plain Language Summary, not published)
In this research we evaluate CNNs as means to improve surficial geology RPM for the case where a model is trained and applied in the same spatial domain and where it is trained from one area and applied in another. This CNN is compared against the more widely used RF algorithm as a benchmark to assess potential improvement.