[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

An amateur fireball detection/tracking system

In these days of high technology, it amazes me that potential meteorite impacts resulting from fireballs are
still localized the very old fashioned way through interviews with (notoriously inaccurate) witnesses. 
Occasionally, security camera video is used when we are lucky enough to get it.

I propose that a computer-based AMATEUR fireball camera system be established.  With the price of computer
power and CCDs rapidly dropping, it seems possible to create an affordable, standardized, stand-alone computer
system dedicated to monitoring the sky for fireballs 24hrs/day which would be affordable by amateurs. 
Individual systems with their CCD cameras perched wherever they had a clear view of the sky and individually 
paid for by the amateurs involved could monitor the sky from many different locations.  Once per day, the 
systems would automatically dial their local Internet access provider of choice (or 1-800 number if no local 
service is available) and upload to a central Internet site images suspected to be those of fireballs as 
determined by image processing software running on the individual machines. Uploaded with the images would be, 
of course, the exact location in latitude and longitude of the camera, the camera involved in the shot (true 
north, south, east, or west facing if a single camera with fisheye lens is not being used), and the exact time
of the shot (with fairy accurate timing maintained by resetting the system clock daily through a time update 
via the Internet during the "reporting in" communication to the central Internet fireball site).  

A less automated, but still completely acceptable, system would dispense with the difficult real time image 
recognition of fireball tracks and automatic reporting to an Internet site.  It would rely on a small array 
of very inexpensive infrared photodiodes pointed skyward, mounted on the camera module, and a small, cheap 
microcontroller (like a PIC) to detect sudden, directional increases in IR radiation from the sky.  This 
would cause video frames from x seconds prior to the IR event to x seconds afterwards to be retained on the 
computer system hard drive (MPEG2 encoded to save LOTS of disk space and, if necessary to further limit 
storage requirements, limited to x frames/second).  Such a system would also be triggered by the flashes 
from lightening, but there are workarounds, if desired.  Automated video recording of lightening would, in 
itself, be interesting and might contribute to the science of meteorology in some way.  However, to give the 
option of rejecting lightening, I considered the idea of using a simple and cheap (4 transistor) VLF radio 
receiver to detect the characteristic radio noise of lightening strikes.  However, since it is suspected that 
fireballs also produce VLF radio energy, it would be a REAL shame to reject the rare fireball event by 
misinterpreting it as a common lightening event.  I still plan to implement a VLF receiver as an option in 
the system, but only because it would be productive to monitor and record VLF frequencies during a fireball 
event. The safest way to reject lightening without incorrectly rejecting a fireball would be to keep track of 
IR event frequency.  Three IR events in, say, 15 minutes would tend to indicate lightening.  If not, 
ITEOTWAWKI ("it's the end of the world as we know it").  After, say, a half hour of no IR events, the system 
would go back to the assumption that IR events were due to fireballs.

A microphone in the camera module of each system could be switched on by the image recognition software or  
the PIC microcontroller when a potential fireball is detected.  Thus, a precise measurement of the time of 
arrival of any sonic booms could be made.  Or, perhaps more simply, the microphone could be monitored 
continuously for any major transients indicating a possible sonic boom, this data only to be stored 
permanently if image recognition routines or the PIC IR detection determine that it is closely associated 
in time with a potential fireball image.

Data from the simplified system would be reviewed manually (i.e. view the video and listen to the audio) and
then sent manually to the central Internet site for analysis if determined to be a valid fireball event.  No 
real hassle at all, unless you're on vacation from home.

The hardware system (using off-the-shelf components) and software of the more sophisticated system could be 
developed by university students and/or amateurs as a cooperative project.  The operating system of a 
dedicated, stand alone system could be Linux to avoid software costs. Manufacturers might be interesting in 
selling higher profit margin hardware components at cost for the positive publicity.  A little kit building 
would not be unacceptable as in the case of the amateur robotic and amateur radio communities.  

BTW, I would welcome any assistance from the readers of this list with programming or hardware design 
experience or from their friends/aquaintances.  I am not an electrical engineer, but am a serious amateur.  
The simplified detection/manual notification system is something that would be "do-able" for me.  Image 
recognition and automated contact with a central web site might be a bit deep for me, but I haven't really 
investigated that yet.

The non-dedicated version of this system would consist of a camera/sensor module and a video capture card
(approx. $200).  The module would be attached to your home system, the CPU of which you would need to leave 
on 24hrs/day.  While you're at work or asleep, the system would be totally dedicated to the fireball 
detection/recording task.  While using the machine, the detection could be turned off or relegated to a low 
priority background task (to avoid slowing your machine excessively).

There would be the natural tendency for larger cities to be filled with an unnecessarily large number of these
systems.  Instead, amateur volunteers could pay for and build their systems and then find, through publicity,
individuals in low population density areas willing to supply a 115V outlet, a tiny amount of floor or desk
space, a spot on their roof for the camera module, a few dollars/month of electricity, and a telephone line
which would be used for an outgoing, no-cost-to-them call at some randomly selected time (to avoid potentially
overloading the main fireball site server) in the middle of their local night.

Amateur astronomers often spend thousands of dollars on their equipment.  Many would love to contribute to the
science of astronomy, but don't either because their equipment is inadequate for the task or because the time
and effort involved in such things as the monitoring of variable stars or the search for comets is more than
they can afford.  The fireball monitoring project would, I believe, be a great opportunity for the amateur to
make a significant contribution with very little investment of his or her time.  I ANY form, such a system 
would be VASTLY better than the crude, horribly inaccurate, "eyeballs-only" fireball tracking system we have 
now.

Bill Blair
Colorado Springs, CO  (with the unusually large number of spectacular but poorly tracked fireballs we've had 
in Colorado in the last few years, you can see why I thought up this idea)

P.S. After thinking up this idea, I searched the web for more info to avoid as much as possible "reinventing 
the wheel."  I found this excellent site:

http://www.imo.net/video/

It's a bit dated as far as video and computer technology/capabilities are concerned (1995), but it still has
a lot of useful information on video recording of meteors and fireballs.

TO UN-SUBSCRIBE: 
----------------
Send an email To: 
meteorite-list-request@meteoritecentral.com 
with UNSUBSCRIBE in the subject field of your 
email. That's all there is to it!
----------------
Thank You
Follow-Ups: