Developments in photorealistic mapping, modeling and analyses of terrestrial scanning of geological outcrops

Contributions are made to improve the data acquisition, surface model construction and subsequent geologic analyses of 3D photorealistic models from terrestrial laser scanning (TLS or ground LiDAR). A system (DOAS-Digital Outcrop Annotation and analysis System) is created for annotating photos used for model construction while observing outcrops in the field or virtual 3D models with a computer using an applet on a tablet such as an iPad. Geologic tracings can be converted into 3D digitized points (from the pixels) with a Matlab program so that geologic information such as strike/dip can then be determined and quantitatively analyzed. The geological features can then be visualized in 3D or 3D stereo which is especially useful in education and also in research. Three sets of road-cut outcrops including the Cambro-Ordovician carbonates of the Arbuckle anticline, Oklahoma, the Paleozoic fluvial deltaic sedimentary rocks of the Breathitt Group, Pikeville, Kentucky, and the Cretaceous Eagle Ford Formation, Del Rio, Texas, have been captured, modeled and preliminarily analyzed. The quality of the image control (tie-points) is a major factor in building photorealistic surface models and also effects the time spent in the field and in post-processing. A coaxial camera mounted on the scanner in the field automatically and simultaneously determines the transformation coefficients. This capability is utilized to improve photorealistic surface mapping and modeling resulting in a faster procedure generating higher resolution and more accurate models. A 3D analysis approach using the multipatch ESRI (Environmental Systems Research Institute) format of 3D models and a suite of ArcScene based modules was designed for geological analyses. Semi-auto and auto filter programs are shown to diminish the errors in surface modeling tested in Pikeville due to the effects of vegetation.