|
Abstract: Suppose an asteroid is so far away that only a single
pixel of light can be measured at any time. The measurement is
assumed proportional to the area of the profile of the asteroid in the
viewing direction - that is, the area of its orthogonal projection
onto a plane orthogonal to this direction. Once several such
measurements have been taken, they must be processed somehow to
produce an image of the asteroid. How can this be done?
The talk will describe joint work with an electrical engineer, Peyman
Milanfar of UC Santa Cruz. We have designed algorithms for this purpose,
and with the help of a Western Washington University undergraduate, Chris
Street, these have been implemented. The talk will include a computer
demonstration of one of the algorithms.
|