Proceedings Article | 12 September 2021
KEYWORDS: Remote sensing, Radar, Optical sensors, Detection and tracking algorithms, Data modeling, Algorithm development, Visualization, Visual analytics, Spatial resolution, Landsat
Lineament mapping is a genuine issue for deciphering the tectonic setting, geological history, stability evaluation, engineering constructions localizations, prospecting orogenic gold deposits, prospecting polymetallic vein-type mineralization,….etc. Consequently, the substantial objective of this study is to compare various remote sensing data sets concerning with their potency in lineaments elicitation to recommend the superior for future usage. Toward this aim, approximately most of the common remote sensing datasets, are tackled in this study, including optical sensors (Landsat 8 OLI, ASTER, Advanced Land Imager, Sentinel 2A), active radar data (Sentinel 1), digital elevation models (ALOS PALSAR, SRTM, NASA, ASTER V3). In this scope, an entirely automatic lineament derivation environment is created through integration of edge detection algorithm, line-linking algorithm, as well as digitizing the pre-existing lineaments from geological maps. Subsequently, thematic comparison, visual interpretation, geostatistical analysis and accuracy assessment are performed. Results revealed that the used optical sensors are less efficient than DEMs having the same spatial resolution. Also, sentinel 1 radar data (C-band, f = 5.4 GHz) is more competent than optical data sources. ALOS PALSAR DEM is more eligible than any other utilized data type. Wholly, DEMs built from L-band (f =1.27 GHz) radar data (PALSAR DEM) proved their leverage in automatic lineament extraction to a limit that can deviate from the well-known relationship between number of extracted lineaments and pixel size.
