 
 
CASSINI ISS Improved (C-Smithed) C-Kernel File
===========================================================================
 
     Created by Josh Riley, CICLOPS, 2005-06-17 11:57:53 MDT
 
 
 
 
Orientation Data in the File
--------------------------------------------------------
 
This file contains c-smithed c-kernel pointing information for
Cassini spacecraft frame, CASSINI_SC_COORD (NAIF ID -82000), relative
to the 'J2000' inertial frame. This pointing was computed by optically
navigating the images taken with the Cassini ISS NAC and WAC cameras.
 
Pointing is defined only for image times where auto-navigation software
was successful.  The c-kernel is of type three.  Each segment in the file
represents the pointing for an image from shutter open to shutter close
plus a four-millisecond buffer on each end of the segment.
 
 
 
 
Status
--------------------------------------------------------
 
This file was created by CICLOPS for archiving with the Planetary Data
System (PDS) from daily images produced by the Cassini ISS and from
kernels produced by the Cassini Project.
 
 
 
 
Pedigree
--------------------------------------------------------
 
The pointing information in this file is derived using an automated program
called AUTONAV (described in full below).  The accuracy of AUTONAV is
dependent on the accuracy of the mission kernels and on the correctness of
the AUTONAV software.  Several validation steps are executed to reduce the
number of poorly-navigated images, but there is a chance that these
validation steps will not catch all errors.  So, users should perform their
own validation steps to ensure the c-smithed c-kernel data looks accurate.
 
Graphic overlay output files are provided for users to visually inspect the
accuracy of AUTONAV.  This file is described further below.
 
 
 
 
Approximate Time Coverage
--------------------------------------------------------
 
The file covers the following interval of the mission:
 
     Coverage Begin UTC    Coverage End UTC
     --------------------- ---------------------
     1999-230T01:40:00.050 1999-230T02:23:50.856
 
 
NOTE: Interpolation between segments is not implied.  Pointing is defined
only for image times where auto-navigation software was successful.  Please
see the Segment Summary section below for exact coverage times.
 
 
 
 
Usage Restrictions
--------------------------------------------------------
 
The use of this file is restricted to Cassini-project internal-use only
until the file is placed in the PDS.  Any proposed or intended scientific
usage of this file or its contents before its placement in the PDS or
publication by the Imaging Team, whichever comes first, must be discussed
and accepted by the Imaging Team leader (as called for by the PSG-developed
Cassini/Huygens Rules of the Road).
 
 
 
 
Usage Notes
--------------------------------------------------------
 
There is a limit of 20,000 segments that can be loaded at one time by SPICE
software.  Software performance can degrade severely if this limit is
exceeded.  ISS C-smithed C-kernels are delivered in one-month blocks of time
which will keep the number of segments in the file to well under this limit.
Most files are on the order of a few thousand segments; none exceed 10,000.
However, the limit can easily be exceeded by loading several of these
C-smith files at once.
 
In order to use this file a Cassini SCLK file containing coefficients
mapping Cassini on-board time to ET must be loaded into a user's program.
 
Loading an Cassini Frames Kernel (FK) containing information that SPICE
software needs to combine camera orientations together will allow a user
to get the combined orientation through the high level SPICE Toolkit
interfaces.
 
 
 
 
Related Kernels
--------------------------------------------------------
 
This file was created using the following LSK, SCLK, and FK files:
 
     naif0007.tls
     cas00089.tsc
     cas_v39.tf
 
 
 
 
Contacts
--------------------------------------------------------
 
If you have any question regarding this data contact the Cassini Imaging
Central Laboratory for Operations (CICLOPS) at the Space Science
Institute (SSI):
 
     CASSINI IMAGING CENTRAL LABORATORY FOR OPERATIONS
     SPACE SCIENCE INSTITUTE
     4750 WALNUT STREET, SUITE 205
     BOULDER, CO 80301
     USA
 
     Carolyn Porco
     720-974-5849
     carolyn@ciclops.org
 
     Josh J. Riley
     720-974-5856
     josh@ciclops.org
 
 
Contact information can also be found on the CICLOPS website at:
 
     http://www.ciclops.org
 
 
 
 
AUTONAV Description
--------------------------------------------------------
 
AUTONAV, a program built on top of MINAS, corrects spacecraft camera
pointing without human interaction.  This allows large batches of images
to be processed into the ISS CICLOPS image archive database.
 
AUTONAV improves the accuracy of the geometry fields which pertain to
foreground objects.  Therefore, foreground features in an image -- limbs,
rings, terminators, etc. -- take precedence over stars when correcting the
pointing, so if the trajectory used in the pointing correction is inaccurate,
then the pointing produced by AUTONAV will not be physical.  In other words,
while the foreground objects should fall where predicted, the stars will not
necessarily line up because the pointing angles determined by AUTONAV will
not correspond to the actual spacecraft orientation due to parallax.  The
goal with AUTONAV is to eliminate false solutions entirely.  Therefore
AUTONAV is tuned conservatively with the philosophy that it is better to
throw out a few good solutions than to accept one ridiculous solution.
Nevertheless, the only way to completely prevent absurd solutions is to
summarily reject all solutions, so in practice ridiculous solutions do occur
on occasion.
 
 
Stages
- - - -
AUTONAV proceeds in stages, starting generally with more global approaches
and progressing toward more local methods.  Each stage attempts to correct
the pointing and evaluate its performance, and its ending condition
(line/sample offset) is used as the starting condition for the next stage.
Stages proceed based on the results of the previous stage, or they may be
forced.  Many stages record their results for further evaluation later in
the procedure.
 
There are two possible cascades.  If useful foreground objects are expected
in the scene, then the foreground cascade is selected, otherwise stellar
navigation is attempted.
 
 
Stellar Navigation
- - - - - - - - - -
Stellar navigation is a low priority in AUTONAV, so its implementation is
not sophisticated.  Two stellar navigation stages are executed:
 
 Global Star Fit
  An algorithm is used to detect all sources in the image that look like
  PSFs.  That distribution is compared to the distribution of known stars
  in the vicinity to produce an association between known stars and image
  point sources.  The line/sample offset solution is computed using a least-
  square fit between the computed and detected star positions.  The line/
  sample uncertainty is computed from the covariance matrix of the fit.  The
  next stage proceeds only if this stage fails or if forced.
 
 Local Star Fit
  A search is performed for PSFs within a small region surrounding the
  computed location of each known star.  Once the association between
  known stars and image point sources is made, the procedure is identical to
  the global star fit.
 
 
Foreground Cascade
- - - - - - - - - -
Some stages use a global grid search, which proceeds as follows: a criterion
is evaluated over a coarse, uniform grid (typically 4x4 locations) in the
image.  The grid is reduced in size and re-centered at the best point from
the previous search until the grid spacing becomes subpixel.  A bias in the
form of a Gaussian weighting function is usually applied to give preference
to solutions that lie close to the initial condition, to allow for the fact
that the Cassini predict pointing is usually fairly accurate.  The line/sample
uncertainty in the result of the grid search is computed by looking at the
width of a Gaussian model fit to the correlation peak in the vicinity of the
solution.
 
Some stages make use of predicted sharp edges, which consist of limbs, ring
edges, and shadows.  Terminators are also included as sharp edges if the
phase angle is favorable.
 
The following stages are executed in the foreground cascade.
 
 Edge centroid comparison
  An edge detection algorithm is used to find the sharp edges in the image.
  AUTONAV computes the expected locations of edges in the image and compares
  the centroids of those two distributions to produce a line/sample offset.
  This algorithm is extremely imprecise, but nearly always improves the
  pointing when the foreground objects being compared fall entirely within
  the image and do not subtend too large an area.  If those conditions are
  not met, then this stage is not executed.  Otherwise the next stage is
  simply allowed to proceed under the assumption that its input condition
  is valid.
 
 Enclosed Flux comparison
  If only one object is being compared and it is expected to fall entirely
  within the image, then this algorithm is applicable.  The outline of the
  object is computed and a grid search is performed throughout the image for
  the line/sample offset that maximizes the difference between the flux
  contained within the outline and that exterior to it.  This algorithm is
  more precise than the centroid comparison, and nearly always succeeds to
  within a few pixels.  The next stage proceeds under the assumption that
  this stage produced as valid result and the result of this stage is
  recorded.
 
 Sharp Edge Comparison
  A grid search is performed throughout the image to find the line/sample
  offset that gives the best correlation between the sharp edges detected in
  the image and those that are expected.  This algorithm is generally more
  precise than the flux comparison, but less reliable as the features of
  interest (edges instead of an extended body) tend to be more sparse in the
  image.  The correlation coefficient is used to evaluate the performance of
  this stage.  If unsuccessful, the procedure is repeated using a low-pass
  filter on the edges to reduce the noise signal.  The result of this stage
  is recorded and the next stage proceeds only if this stage was unsuccessful,
  or if forced.
 
 Sharp Edge Comparison, no terminator
  For appropriate phase angles, the previous stage will have included the
  terminator as an edge.  This stage performs the same procedure except
  that the terminator is excluded.  The result is recorded and the next
  stage proceeds only if this stage was unsuccessful or if forced.
 
 Comparison of preceding results
  Although the preceding stages may evaluate their own performance, the
  reliability of the evaluation criteria is questionable.  Therefore, those
  stages are tuned to act conservatively, rejecting some results that may be
  acceptable rather than accepting results that may be unacceptable.  This
  stage compares all previous results that have been recorded and if two
  or more agree, then the result of this stage is taken as the average of
  those and their uncertainties are combined in the Pythagorean sense.
  The result of this stage is recorded and the next stage proceeds only if
  this stage is unsuccessful or if forced.
 
 Least-square Limb Fit
  This stage is forced to proceed regardless of the results of previous
  stages.  Using the current line/sample offset, computed limbs are scanned
  in search of image limb points.  A least-square fit is performed to find
  the line/sample offset that gives the best chi-square between the two sets
  of points and the uncertainties are computed using the resulting covariance
  matrix.  This procedure requires the initial condition to be close to the
  actual solution.  The performance of this procedure is evaluated using the
  chi-square value and if unacceptable, the procedure is repeated using all
  previously recorded solutions as the starting condition, or until an
  acceptable solution is found.
 
 Enclosed Flux Validation
  This stages is forced to proceed regardless of the status of previous
  stages.  If only one object is being considered and it is expected to fall
  entirely within the image, then the current solution is evaluated by
  comparing the flux enclosed by the outline of the object to that
  exterior to the outline.
 
If there is an acceptable solution after all stages have executed, then a C
kernel is written and a graphic is generated to illustrate the result.
AUTONAV attempts to produce pointing with a precision of at worst 5 pixels.
Therefore, AUTONAV will never output an uncertainty in line or sample that
is greater than 5 pixels.  Regardless of the actual computed uncertainty,
if AUTONAV accepts the pointing, then that is because it thinks it is good to
at least 5 pixels.
 
 
 
 
Validation
--------------------------------------------------------
 
Graphic Output
- - - - - - - -
 
If AUTONAV determines that the pointing has been successfully corrected, then
it produces a graphic showing features overlain on the image so as to allow
the result to be validated by eye.  The graphic includes all edges used in the
procedure (though various edges may or may not be used in various stages), as
well as all stars and the centers of all bodies labeled with their abbreviated
names.  An overlay graphic is produced for each image observation included in the
c-kernel.  All graphics are delivered with the c-kernel as well and are available
in the corresponding CK directory.
 
 
Angular Difference Output
- - - - - - - - - - - - -
 
Also delivered with the c-smith package is an ASCII file containing the
angular rotation deltas between the reconstructed ACS c-kernel and the
Autonav c-kernel for the observation times covered in the c-smith
c-kernel sampled at some small time interval.  The columns are:
 
1.  SCET Time
2.  Total angular rotation delta in mrads
3.  The rotation delta about the X-Axis in mrads
4.  The rotation delta about the Y-Axis in mrads
5.  The rotation delta about the Z-Axis in mrads
6.  The X-Axis S/C body rate in mrad/sec
7.  The Y-Axis S/C body rate in mrad/sec
8.  The Z-Axis S/C body rate in mrad/sec
 
See file: 99213_99243cb_ISS_bc_err.txt
 
 
Target Center Line/Sample
- - - - - - - - - - - - -
 
The target name and target center line and sample calculated by Autonav
are provided in the Segment Summary section below for each image.
The target name comes from the planned target body in the observation's
pointing design.  Line numbering starts at one with first line in the
image file.  Sample numbering starts at one and increases with
increasing sample in the image file.  The line and sample are for the
actual vicar image (not the 512 x 512 Autonav overlay graphic) which can
vary in size up to 1024 x 1024.
 
 
 
 
 
                     Segment Summary: ID -82000
-----------------------------------------------------------------------------------------------------
Image Name      Start (UTC)             Stop (UTC)              Target Name        Line    Sample
-----------------------------------------------------------------------------------------------------
W1313632011_1   1999-230T01:40:00.050   1999-230T01:40:03.858   SKY                N/A     N/A
W1313632115_1   1999-230T01:40:91.849   1999-230T01:41:47.857   SKY                N/A     N/A
W1313632133_1   1999-230T01:42:05.289   1999-230T01:42:05.857   SKY                N/A     N/A
W1313632165_1   1999-230T01:42:25.849   1999-230T01:42:37.857   SKY                N/A     N/A
W1313632182_1   1999-230T01:42:54.737   1999-230T01:42:54.865   SKY                N/A     N/A
W1313632206_1   1999-230T01:43:16.261   1999-230T01:43:18.869   SKY                N/A     N/A
W1313632224_1   1999-230T01:43:35.664   1999-230T01:43:36.872   SKY                N/A     N/A
W1313632280_1   1999-230T01:44:06.868   1999-230T01:44:32.876   SKY                N/A     N/A
W1313632298_1   1999-230T01:44:50.556   1999-230T01:44:50.884   SKY                N/A     N/A
W1313632347_1   1999-230T01:45:39.783   1999-230T01:45:39.891   SKY                N/A     N/A
W1313632382_1   1999-230T01:46:13.687   1999-230T01:46:14.895   SKY                N/A     N/A
W1313632522_1   1999-230T01:48:32.905   1999-230T01:48:34.913   SKY                N/A     N/A
W1313632540_1   1999-230T01:48:51.713   1999-230T01:48:52.921   SKY                N/A     N/A
W1313632614_1   1999-230T01:50:06.596   1999-230T01:50:06.924   SKY                N/A     N/A
W1313632663_1   1999-230T01:50:55.824   1999-230T01:50:55.932   SKY                N/A     N/A
W1313632680_1   1999-230T01:51:10.924   1999-230T01:51:12.932   SKY                N/A     N/A
W1313632698_1   1999-230T01:51:29.732   1999-230T01:51:30.940   SKY                N/A     N/A
W1313632754_1   1999-230T01:52:00.935   1999-230T01:52:26.943   SKY                N/A     N/A
W1313632772_1   1999-230T01:52:44.619   1999-230T01:52:44.947   SKY                N/A     N/A
W1313632821_1   1999-230T01:53:33.850   1999-230T01:53:33.958   SKY                N/A     N/A
W1313632838_1   1999-230T01:53:48.950   1999-230T01:53:50.958   SKY                N/A     N/A
W1313632856_1   1999-230T01:54:07.754   1999-230T01:54:08.962   SKY                N/A     N/A
W1313632912_1   1999-230T01:54:78.957   1999-230T01:55:04.965   SKY                N/A     N/A
W1313632930_1   1999-230T01:55:22.641   1999-230T01:55:22.969   SKY                N/A     N/A
W1313632979_1   1999-230T01:56:11.873   1999-230T01:56:11.981   SKY                N/A     N/A
N1313633721_1   1999-230T02:08:33.027   1999-230T02:08:33.855   MOON               281.2   821.4
N1313633738_1   1999-230T02:08:50.470   1999-230T02:08:50.858   MOON               280.1   759.4
N1313633756_1   1999-230T02:09:08.814   1999-230T02:09:08.862   MOON               278.3   693.6
N1313633773_1   1999-230T02:09:25.814   1999-230T02:09:25.862   MOON               553.9   1265.5
N1313633806_1   1999-230T02:09:58.778   1999-230T02:09:58.866   MOON               551.7   1034.9
N1313633839_1   1999-230T02:10:31.798   1999-230T02:10:31.866   MOON               554.1   808.9
W1313634054_1   1999-230T02:14:06.822   1999-230T02:14:06.845   MOON               526.9   393.8
W1313634071_1   1999-230T02:14:23.822   1999-230T02:14:23.845   MOON               526.5   383.1
W1313634088_1   1999-230T02:14:40.821   1999-230T02:14:40.844   MOON               526.1   372.7
W1313634105_1   1999-230T02:14:57.821   1999-230T02:14:57.844   MOON               525.8   362.3
W1313634122_1   1999-230T02:15:14.821   1999-230T02:15:14.844   MOON               525.7   352.1
W1313634139_1   1999-230T02:15:31.821   1999-230T02:15:31.844   MOON               525.6   342.0
W1313634156_1   1999-230T02:15:48.821   1999-230T02:15:48.844   MOON               525.5   332.1
W1313634196_1   1999-230T02:16:28.754   1999-230T02:16:28.852   MOON               526.8   309.4
W1313634213_1   1999-230T02:16:44.343   1999-230T02:16:45.851   MOON               526.9   299.7
W1313634247_1   1999-230T02:17:18.343   1999-230T02:17:19.851   MOON               528.2   281.5
W1313634264_1   1999-230T02:17:36.383   1999-230T02:17:36.851   MOON               528.3   272.0
W1313634281_1   1999-230T02:17:53.793   1999-230T02:17:53.851   MOON               529.3   262.7
W1313634298_1   1999-230T02:18:10.723   1999-230T02:18:10.851   MOON               529.4   253.6
W1313634315_1   1999-230T02:18:27.818   1999-230T02:18:27.851   MOON               528.8   240.7
W1313634332_1   1999-230T02:18:44.838   1999-230T02:18:44.851   MOON               531.1   234.9
W1313634349_1   1999-230T02:19:01.783   1999-230T02:19:01.851   MOON               528.9   222.6
W1313634366_1   1999-230T02:19:18.722   1999-230T02:19:18.850   MOON               529.0   217.2
W1313634383_1   1999-230T02:19:35.162   1999-230T02:19:35.850   MOON               527.9   208.6
W1313634400_1   1999-230T02:19:52.837   1999-230T02:19:52.850   MOON               527.0   197.0
W1313634417_1   1999-230T02:20:09.827   1999-230T02:20:09.850   MOON               525.1   190.0
W1313634434_1   1999-230T02:20:26.827   1999-230T02:20:26.850   MOON               524.8   181.3
W1313634451_1   1999-230T02:20:43.692   1999-230T02:20:43.850   MOON               524.5   170.4
W1313634468_1   1999-230T02:21:00.802   1999-230T02:21:00.850   MOON               525.0   156.8
W1313634485_1   1999-230T02:21:17.692   1999-230T02:21:17.850   MOON               525.7   142.6
W1313634502_1   1999-230T02:21:33.342   1999-230T02:21:34.850   MOON               525.7   128.2
W1313634536_1   1999-230T02:22:07.349   1999-230T02:22:08.857   MOON               526.6   100.9
W1313634553_1   1999-230T02:22:24.849   1999-230T02:22:25.857   MOON               526.4   87.2
W1313634570_1   1999-230T02:22:42.749   1999-230T02:22:42.857   MOON               527.2   73.3
W1313634587_1   1999-230T02:22:59.629   1999-230T02:22:59.857   MOON               528.3   61.5
W1313634604_1   1999-230T02:23:16.799   1999-230T02:23:16.857   MOON               529.0   43.6
W1313634638_1   1999-230T02:23:50.698   1999-230T02:23:50.856   MOON               527.4   19.8
 
