Band-to-band registration Cassini

Hello!

I am looking to create 3 band cubes from the 1 band cube Cassini ISS data. Right now, I was thinking about doing this band-to-band registration by using the findfeatures function on ISIS to create a control network for the 3 images, and then ‘populate’ said control network with the 3 cube files corresponding to the bands. Is this a good way of going about it?
If so, is there a way to ‘attach’ 1 band cubes to the control network to create a 3 band cube?

I’m pretty new to ISIS, so any help/advice would be greatly appreciated! Thank you!

generating small control networks and adjusting each frame will likely work okay. You will need to pick one band and hold that image fixed when you run jigsaw.

I would also recommend looking into the coreg app. It does what you’re looking for in a much more straightforward fashion.

1 Like

Thank you so much! I have two subsequent questions: I tried using coreg but it would only allow me to do this with two cube, and would yield a single cube with 1 band (instead of 2), is there a way to use coreg to input two 1 band cubes and obtain one 2 band cube?
This might be a dumb question, but I was just exploring the possibility of using jigsaw, and I am not sure where to find the sensor parameter data for Cassini to create the file to input into SCCONFIG… I’m still exploring though.

Again, thank you so much, and thank you for your patience as I learn how to navigate the software!

Coreg works on single band cubes and produces single band cubes. So, you’ll need to coreg your bands as separate cubes and then stack them back together. You can use explode to split out each band into a separate cube and then cubeit to stack them back together. You can also use some fancy band-selection to avoid needing to separate the input cube.

For example if your image is called test.cub then you can run coreg match=test.cub+1 from=test.cub+2 to=coreg_w.cub to coregister band 2 of test.cub against band 1 of test.cub. See cube attributes for more documentation on this stuff.

Going the jigsaw route is very complex, I would highly recommend working with coreg before going down the jigsaw path.

For jigsaw, I would recommend against using scconfig, it’s designed for when you are bundle adjusting data from multiple different sensors and you want to solve for different things on each data set. For Cassini ISS data, you’ll probably just want to solve for pointing. So, you would use camsolve=angles and spsolve=none. Here’s a very very rough example command line I would use assuming you are registering test_2.cub to test_1.cub and you have a control network file called 2_to_1.net.

cubes.lis

test_1.cub
test_2.cub

held.lis

test_1.cub

command

jigsaw fromlist=cubes.lis heldlist=held.lis cnet=2_to_1.net onet=2_to_1_jig.net camsolve=angles spsolve=none camera_angles_sigma=10

Run that and look at the output files. Particularly check your measure residuals. If you’re happy with the results then add update=yes to update the pointing on test_2.cube. If there are measures with really high residuals then you can open them up in qnet and either manually fix them or delete them.

Also, the camera_angles_sigma value is just a really rough guess.