From: diego@minerva.st.dsi.unimi.it (Diego Montefusco)
Subject: Re: TECH: Kalman Filtering Trackers
Date: Sun, 31 Jan 93 02:47:12 CET



In <9301301635.AA11487@ghost.dsi.unimi.it>, on Jan 30, you wrote:

> From: Roy_W_Latham@cup.portal.com
> Subject: TECH: Kalman Filtering Trackers
> Date: Thu, 28 Jan 93 15:40:09 PST
> 
> I spent a number of years using Kalman filters for aircraft
> positioning using various navigation sensors as inputs. I suspect the
> technique would be well suited to the processing of tracking data,
> although systems like the Polhemus are already performing noise
> filtering of some sort internally.

Since I started this thread, I can say I was originally referring not
to "clean" the signal from noise with a Kalman filter, but actually to
PREDICT the movements of the head of a user, so to eliminate the lag
tipical of head tracked systems. This is exactly what the Kalman
filter does in the MR Toolkit of Alberta University.

Since you worked with this kind of filter, you could answer me: how
computationally intensive is it?

Thank you.

Diego