At 08:25 AM 6/24/99 -0500, Matt Bennett wrote:
[snip]
>... You also run into problems
>that moving targets appear in the wrong place in the image because of
>doppler shifts, and you need to sample at a higher rate with a lot of
>precision, because you end up needing to do a backprojection type of image
>where precise phase data of the incoming signal is vital to the quality of
>the image.

   Matt, this is not of much help and you obviously have a thorough
knowledge of RADAR and SONAR. I was once involved with a TI airborne
multi-mode RADAR with an MTI function to detect trucks on a `trail'...
The RADAR used the KU band and was tunable.

   We took advantage of the fact that moving objects added a Doppler
`phase history' on the received pulse. We filtered this info to blank
ground clutter and only show moving targets. It seems to me that you could
use a similar approach to ignore moving objects on a pulse by pulse basis.
Obviously I haven't thought this through and you are using ultrasonic.
Anyway, good luck. This is a very interesting project.

   - Tom