(meteorobs) Determining Altitude & Azimuth From All Sky Cam Images

Chris Peterson clp at alumni.caltech.edu
Mon Dec 7 19:08:43 EST 2009

I find that a zenith pointing fisheye image is the easiest of all systems to 
determine altitude and azimuth. With my cameras, I use a simple 
transformation. I identify the (fractional) pixel that represents the 
zenith, and define polar coordinates around that point. The azimuth is just 
the trivial angle. For altitude, I use a cubic polynomial to relate r, the 
distance from the center, to altitude.

I initially worked out the polynomial coefficients by looking at the 
positions of the centroids of about 50 stars and doing a fit. Since all the 
lenses and cameras are the same, for new setups I really only need to 
identify the center position and reference rotation angle, which is easily 
done with a few images containing some stars, or even the Moon over the 
course of an evening.

I test occasionally against bright stars and planets (the only things that 
show in a single video frame) and the transformed centroids of these objects 
are always within 0.5° of their true position (usually better for the higher 
altitudes), so I know the system is working well.


Chris L Peterson
Cloudbait Observatory

----- Original Message ----- 
From: "Meteorites USA" <eric at meteoritesusa.com>
To: "Global Meteor Observing Forum" <meteorobs at meteorobs.org>
Sent: Monday, December 07, 2009 2:43 PM
Subject: (meteorobs) Determining Altitude & Azimuth From All Sky Cam Images

> Hi list,
> A repost in it's own thread of the question I posed earlier.
> What I would like to learn is how to determine altitude and azimuth of
> the retardation point of a bolide event as recorded from a fish-eye lens
> from an all-sky cam. It's easier to figure from a camera with a 90
> degree FOV pointing horizontally in any given direction, but an all sky
> camera pointing straight up with a 180 degree FOV is much harder to
> calculate. The horizon to horizon view coupled with the distortion of
> the image and aliasing can cause mis calculations.
> Can someone please explain how to determine altitude and azimuth please
> from an an all sky image? Links? Papers? Discussions?

More information about the Meteorobs mailing list