Henry:
Camera calibration consists of measuring the field of view upon recovery;
the rest is just monkey motion that provides no useful insight.
Exactly this technique was used—off duty—by staff at JPL to estimate the
height at which an image was taken during the RRS Dart flight of 1996; the
relevant variance analysis had to do with the ability to measure the bulge
in the Earth horizon image, which was accounted for in the analysis as
documented in the article currently available at rrs.org.
If one insists upon correlative data, it can be had from direct imaging of
the ground—as was done w/ the RRS Dart data. Correlating w/ GPS or any
alternative data source is merely redundant to what is already proven.
The rest of what you propose apoears—to me—to be utterly irrelevant: it
is—in my experience—a waste of effort and money to spend time on problems
the answers to which can not effect the certainty of our existing knowledge.
Bill
On Mon, Jul 30, 2018 at 7:17 PM Henry Spencer <hspencer@xxxxxxxxxxxxx>
wrote:
On Mon, 30 Jul 2018, William Claybaugh wrote:
The method for determining altitude from earth limb imagery is detailed
in the 75th anniversary issue of the RRS News... Determining altitude
from a well know surface feature distance is mere trigonometry if we
make the flat earth assumption and only slightly more complex if we
account for the earth’s bulge...
Which is why my suggestion was not for a paper study of what the math
behind it would be, but of an *engineering test* to evaluate *how well* it
works with real hardware in real flight conditions, by carefully comparing
it to GPS data at multiple points on the trajectory. Which tends to focus
attention on gritty real-world issues like calibration rather than
trigonometry.
Camera calibration is admittedly a lot less fun than trigonometry, but
nobody is going to trust altitudes determined this way without at least
one solid experimental test that pays attention to such issues (if only to
establish that they are insignificant, which is *not* self-evident).
Henry
