search

DOI-USGS / ISIS3

Integrated Software for Imagers and Spectrometers v3. ISIS3 is a digital image processing software package to manipulate imagery collected by current and past NASA and International planetary missions.
https://isis.astrogeology.usgs.gov
Other
198 stars 168 forks source link

Jigsaw problem solving for Position #1171

Closed ascbot closed 5 years ago

ascbot commented 5 years ago

Author Name: Tammy Becker (Tammy Becker)

Original Assignee: Debbie Cook

What Jigsaw considers to be the original position is based on a polynomial fit to the data ...this can be significantly different from the true original data.

This is evident with many Mini-RF scenarios; for instance with two images, one has 32000+ lines and a second image has 101,000+ lines...the resulting jigsaw pulls the 'apart'

Jigsaw needs to solve for hermite splines

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2011-12-29T20:59:38Z

One idea for a short term work around:

Crop the long orbit image data into shorter length pieces. Test1: Crop a single image from 101,0000 lines to 33,000 lines. Spiceinit-ed shorter segment, updated the line measure coordinates in the control network Jigsaw results were the same as the original probably due to the AlphaCube group

Test2: Deleted the 'AlphaCube' group from the labels of the cropped image in Test1, Spiceinit-ed the shorter segment Jigsaw failed, Apriori spice was way off and no longer could be mapped into the segmented image.

Test3: Try deleting the 'AlphaCube' group from the original cropped image in Test1, this time don't re-spiceinit (i.e., keep the tables from the original uncropped image)...we'll see what happens....stay tuned...

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2011-12-29T21:15:54Z

Test3: results in the same wacked spice as in Test2. Even though a tabledump of the instrument position has no difference from the original spice.

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2011-12-29T21:19:35Z

Idea ala Randy Kirk:

Is it a reasonable/quick-fix to improve the way the polynomial is set up for a segment of an image (i.e., "quickly" solve the problem with the "alpha cube" for a workaround).
"We don't want the program telling us our points are outside the image, but on the other hand we don't want it setting up a polynomial over the time range of the full image if we're working with a subimage."

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2011-12-29T21:21:10Z

One more bit of information from Randy regarding Cassini Mini-RF:

"...I did a quick calculation of how much a quadratic function (parabola) departs from a circular orbit. If you are willing to tolerate up to 10 m of error then you can have an orbit arc of 3.7 degrees, which is 112 km or about 15,000 lines for Mini-RF zoom mode. Doubling the length of the image (roughly) increases the error to 100 m because the error goes as the image length cubed."

ascbot commented 5 years ago

Original Redmine Comment Author Name: Debbie Cook (Debbie Cook) Original Date: 2012-01-25T03:53:29Z

Ken Edmundson and I are working on adding 2 new interpolation options to jigsaw for spacecraft position. The first option is to expand the current 2nd order polynomial function to allow higher order polynomials. The second option is to fit a polynomial function over (added to) the Hermite cubic spline.

I created a directory under /work/projects/isis/latest/m00646

for this task.

ascbot commented 5 years ago

Original Redmine Comment Author Name: Debbie Cook (Debbie Cook) Original Date: 2012-03-01T01:47:01Z

The completion of this issue is pending user testing. Some of the original test data (polar) is not usable because of outstanding issues with dem software (Mantis ticket 694). Randy suggested testing with only nonpolar data for now so that the updated software can be made available.

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2012-03-01T18:29:35Z

changed Type to Bug from 'Block'

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2012-07-21T00:46:06Z

This post is linked to #783. Any further testing for MiniRF-RADAR will be performed under #783. Refer to the #783 notes for additional information.

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2012-07-21T00:48:16Z

On 20-July-2012 Ken Edmundson compared #646 and #783 results for MiniRF RADAR and confirmed an exact match...this helped for future testing of RADAR under #783 (more recent version of jigsaw that contains 646 functionality).

ascbot commented 5 years ago

Original Redmine Comment Author Name: Debbie Cook (Debbie Cook) Original Date: 2012-10-20T01:13:03Z

Ken Edmundson worked on this task as well. The 80 hours is a guess at both of our time and includes testing.

ascbot commented 5 years ago

Original Redmine Comment Author Name: Tammy Becker (Tammy Becker) Original Date: 2012-10-22T18:21:48Z

Post 783 has been closed..this version of jigsaw is now checked-in. This ticket should be closed now also.