Geodynamic pattern recognition by satellite images of some areas in Asian island arcs for mineral exploration

Methods of pattern recognition by digital processing are applied to analyze edges, lines, density of lines, and the standard deviations of pixel brightness of some satellite images in Asian island arcs. In South Kyushu, where activity of younger volcanics occurs from Tertiary to present age, Mt. Takakuma is composed of Miocene granite surrounded by Cretaceous to Quaternary sediments and is about 20 km/spl times/20 km in area. Peripheral areas of the granite include some gold-bearing tungsten deposits that have been already mined out. Edges and lines extracted in images mostly coincide with the geodynamic fracture system including faults, joints, cracks, and sedimentary stratification. The directional frequencies of the lines are well concordant with the observed major faults. Composite images coincide with the pattern of formations. The heavily vegetated areas can be classified geologically using the authors' method.<
>