Haptically Guided Filtering
Reverse engineering of mechanical systems often begins with large
datasets produced from laser scanning of physical
artifacts. Commonly it is necessary to remove noise and filter
them; however, selecting noisy regions and preserving sharp edges
on desired features is difficult using standard GUI interfaces. We
demonstrate a haptic interface for marking and preserving features
in noisy data and for performing local smoothing operations. The
force-feedback provides a natural interface for these
operations.
Heightfield Haptics
We present a system for haptically rendering large height field
datasets. Inasmuch as, height fields are naturally mapped to
piecewise bilinear patches. We develop algorithms for
intersection, penetration depth, and closest point tracking using
bilinear patches. In contrast to many common haptic rendering
schemes for polygonal models, this approach does not require
preprocessing or additional storage. Thus, it is particularly
suitable for the large scale datasets found in geographic and
reverse engineering applications.
