search

DOI-USGS / ale

Abstraction Layer for Ephemerides (ALE)
Other
13 stars 33 forks source link

Issues with the spiceinit'd cube driver #104

Closed jessemapel closed 5 years ago

jessemapel commented 5 years ago

  1. The sensor position and velocity are in the J2000 reference frame. They need to be rotated into the body fixed reference frame.

  2. The sensor orientation is the rotation from the J2000 reference frame to the sensor reference frame. This needs to be fixed so that it is the rotation from the body fixed reference frame to the sensor reference frame

  3. The sensor orientation is missing the constant rotation. The binary table data defines the time dependent rotation which needs to be followed by the constant rotation from the table label.

  4. The body orientation needs to read the rotation coefficients off the table label instead of trying to read the binary table contents.

jessemapel commented 5 years ago

A bit more detailed info:

The body orientation defines the rotation from J2000 to the body fixed reference frame. The rotation coefficients are stored in the PoleRa, PoleDec, and PrimeMeridian keywords of the label. How to use the coefficients is described in the NAIF docs.

The constant rotation for the sensor orientation is stored as a rotation matrix in the ConstantRotation keyword of the label.

jessemapel commented 5 years ago

One more problem:

  1. The camera model expects all of the ephemerides to be regularly spaced and ISIS doesn't always do that. So, we may need to re-interpolate some values. IE. the camera model would expect the first ephemeride to be -6, the second, -3, the third 0, and the fourth 3. ISIS sometimes does things like the first is -6, the second is -5, the third is -2, and the fourth is 3.
jessemapel commented 5 years ago

Done! See #122