On Wed, May 7, 2008 at 9:53 AM, Doug LaRue <[EMAIL PROTECTED]> wrote: > ** Reply to message from Karl Cunningham <[EMAIL PROTECTED]> on Wed, 07 May > 2008 > 08:09:33 -0700 > > > > Smoke must be > > distinguished against a variety of backgrounds -- sky, clouds, rocky > > hills, dry meadows, etc. Software to distinguish smoke from clouds > > sounds tough. Humans have a hard time with this one, and it's especially > > needed during thunderstorms. > > probably true but a combined system would beat the pants of the 100% > human-at-the-site system. Having a camera system up there looking for > smoke and then alarming back to a single manned post who is "watching" > all of southern CA shouldn't be out of the question. And in another 10 years > we could probably have dozens of mini drones which each could be > airborne in short order and do the IR detection mentioned. > > > > > Given today's technology I suspect this is better done by humans. > > Fine, put cameras out there so one or two humans are remotely monitoring > dozens of sites. Lightning detection should be a piece of cake and after that > the humans keep and eye out for either smoke or fire at the strike location. > > We do have alot of tech which gets us close to having manned observation > platforms but without the required manpower and salaries. > > Doug
LIDARs on the ground with Bayesian signal detection trained to look for smoke. See the Kooler for a work thread on this. As Doug says there is a _lot_ of technoloogy waiting to be properly developed and deployed. For a tiny fraction of what we spend on the military we could probably do one hell of a good job of early on fire detection ... BobLQ BobLQ -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
