Efficient Preprocessing and Interactive Rendering of Scanned Point Clouds
Nowadays, digitizing 3D geometry has actually evolved from a cutting-edge and high-cost technology to a method performable even with low-cost consumer hardware like smartphones. This development comes along with a steady increase of performance, accuracy, and resolution of professional scanning techniques, resulting in a dramatic increase of data-set sizes to several orders of magnitude more points than available screen pixels. From this we see two consequences: it becomes feasible and necessary to use points as rendering primitives, still the amount of rendered points has to be filtered down to a feasible amount for interactive display.
In this talk, an overview over the challenges of handling and interactively displaying real-world massive point clouds will be given. First results, current work, and further challenges of the project "enercloud" will be discussed. This project tries to tackle these challenges on the background of practical needs and develop an intuitive software package for practitioners.
Paul Rosenthal (Universit?t Chemnitz)
Fr. 23.11.2012, 13:00 c.t. , G29-R335