An integrated approach for extraction of lithology information using the SPOT 6 imagery in a heavily Quaternary-covered regionNorth Baoji District of China

摘要

We studied the applicability of SPOT (Systeme Probatoire d'Observation de la Terre) 6 satellite imagery for red sandstone interpretation in the heavily loess-covered northern Baoji District, China, which is always a crucial but challenging issue in remote sensing geological mapping. First, the non- or sparsely vegetated subareas were outlined in order to narrow the interpretation target area; second, band-math and the fractal DN-N (digital number-frequency) algorithm was conducted to extract the sandstone-associated anomalies; third, a series of spatial analysis, for example, overlaying and clumping, was employed for elimination of the false anomalies; finally, we took advantage of matched filtering to discern the non- (red) sandstone outcrops in local areas. Generally, the interpretation accuracy based on this novel technical proposal is acceptable74% or slightly less (taking the geological sketch as reference criteria)and can be well proven by field investigation. This study may have contributed a useful case study for remote-sensing lithological mapping in the Quaternary-covered areas.