Abstract: This seminar will focus on localization in GPS challenged urban
canyons using
skylines. In our experimental setup, a fisheye camera is
oriented upwards
to capture images of the immediate skyline, which is generally
unique and
serves as a fingerprint for a specific location in a city. We
estimate the global position by matching skylines extracted from
omni-directional
images to skyline segments from coarse 3D city models. Under
day-time and clear
sky conditions, we use a sky-segmentation algorithm using graph
cuts. In
cases where the skyline gets affected by partial fog, night-time
and occlusions
from trees, we use a shortest path algorithm that computes the
location
without prior sky segmentation. I will also briefly highlight
some of our other
results in pose and motion estimation using points, lines,
planes
and light-paths.
- This event has passed.