• Telecon number: 866.603.2146 access code 4538594

August 18, 2004

• Installers are mostly done. Files are placed onto the target system "mostly" the way they want. Problem with the difference between the usual toolkit and the special internal toolkit that Leica uses. They might just choose to distribute the internal toolkit instead of fighting install issues with the external kit (with one or two things added to it).
  • Not a problem to distribute the internal toolkit.
  • Just trying to keep the distro simple.
  • IDK would reduce installers by one and make things simpler.
  • Source installer works and builds.
  • Resolved by end of the week (Aug 21).

• Some combination of Janno, Scott, and such will run Dan's test code in the near future.
  • Ensures we understand their code.
  • Updates our timing and estimates of their code performance.

• Leica guys found and will be fixing a VNIR processing bug.

• Leica's curious if all the boresights are figured out yet?
  • Don: it's safe to say that we have boresight estimates for all four cameras.
  • No such thing as a "perfect" boresight.
  • Residual errors are expected, but existing numbers are definitely usable.

August 4, 2004

• Small meeting, lots of people not here (schedule conflicts) on both teams, RIT and Leica.

• Don Light's been in discussion with Paul Kennedy and others about the boresight. Don and Dan Fava are following up further today, with better angles; this is a continuation from the email on the mailing list from Jim Greetham.

• Leica: someone (either Leica or RIT) needs to take into account the camera rotation. Preference is that RIT does it. If Leica does it, software will have to be rebuilt when cameras are changed; presumably, we (RIT) could solve it in a non-recompilable way.

• Software installers from Leica are "on the way"; Jeff's system had a meltdown, so there's a few more days before it's out there. "Couple days" away, probably early next week.
  • Source code installer
  • Support libs installer
  • Unit test installer

• Leica (and Don) were curious if I (Bob) saw the updated documentation.
  • Yes, I have. Complements on the improved docs. More feedback.

• When Leica is on site Tuesday, they'll have a new release and installers. They feel like they're finishing base development, they want to declare a "beta" soon. We should start integrating very soon, in order to get good support from Leica.

• No telecon next week, since Andrew will be on-site.

July 14, 2004

• Software distribution from Leica expected late today or early tomorrow.

• 2pm, Friday, we'll have another telecon to review the distro.

July 7, 2004

• Lost a lot of the conversation between Don and Harvey and Leica
  • POS/AV "OREMA" transformations of Applanix ω φ κ into useful reference frames
  • ...

• BobKrzaczek caught a comment where it sounded like the libraries they send us make a remote call to a separate Imagine process running on the system. This makes him nervous, given that the client of these libraries are CORBA objects.
  • In a week, they'll send me something.
  • Assure me it shouldn't be a problem.
  • Imagine process needs to be running when WASP ADP is running.
  • ScottLawrence asks, how well does this work with so many objects and other processes?

July 1, 2004

• Attendees
  • them: Jerry, Andrew (in Spain), Daniel
  • us: ScottLawrence, HarveyRhody, DonMcKeown, BobKrzaczek

• Leica needs boresight information for some of the modules. They have the test data from RIT. Boresight work is "under way" at Leica (Dante?).

• They've sent some prelim data to Jangho. He's seeing problems with using the Leica data with the Appl utilities (certain EO data can't be read with the utilities). Still working on it.

• Leica continues to evaluate and unit test. Daniel will be doing the testing, so that when he returns to RIT he's familiar with everything.

• EO parameters: ω φ and κ are in degrees. X, Y, and Z are in "map coordinates". WGS UTM '84. DonMcKeown verifies that this makes sense and should be doable.

• Meetings moved. Wednesday 9:30am moved to Wednesday 11am. Monday 1pm moved to Wednesday 2pm.

• End-of-flight mosaicing question from Leica: files or other interface?
  • Due to performance, priority, and other constraints placed on us by Windows, this isn't going to run during dead legs.
  • Starts when plane begins the return to airport, runs to finish (even if after landing).
  • Will be a module that is "pointed" at a set of images, generates the "medium" resolution. Leica thinks we're somewhere "on the order" of an hour or so processing time. Don thinks that's reasonable, we can sit "on the ground" after a flight for half an hour, for example, no problem. Sounds like the fire camp mosaic is reasonable.
  • We supply a "directory" full of GeoTIFF images. They supply an output mosaic. RIT overlays explicit fire information.

June 14, 2004

• Attendees
  • us: ScottLawrence, HarveyRhody, DonMcKeown, BobKrzaczek
  • them: Andrew, Paul, Jerry, Jeff

• The big issue for Leica is that they need "put a price on" the boresighting proposal.

• Leica has added Yuenyin (sp?) to their boresighting team.

• Been working the boresighting issue all of Friday and today (Monday).

• Also working through IRSM equations

• Another programmer joining their team from Huntsville

• Leica will be using Boost libraries for memory management, possibly more.

• To Do List
  • Leica
    • Communicate w/ DonMcKeown re: boresighting
    • Communicate w/ BobKrzaczek re: ADP
  • RIT
    • Define the in-flight "electronic footlocker"

June 9, 2004

• Attendees
  • us: ScottLawrence, HarveyRhody, DonMcKeown, BobKrzaczek
  • them: Andrew, Paul, Jerry, Jeff

• Each thumbnail would be associated with geo info and canvas coordinates
  • each thumbnail is scaled down by some configurable amount
  • transformation on thumbnail: scaling (downsample), rotation (ω, φ, κ) into canvas coord.
  • Is there more?

• Thumbnailing mosaic (note new term!) is "per leg", not per flight. Weird disconnect between what I thought and what everyone else thinks. Need to think on this further. Talk to Jason, or someone who's "in flight" and will actually need and use this information. See what they think.

• Final result: their server will process image-at-a-time. Will return "thumbnail" with meta data mentioned above. We (RIT) will be responsible for managing display list.

June 7, 2004

• Attendees: BobKrzaczek, Leica crew

• Main discussion revolved around the so-called "waterfall" display.
  • plane orientation vs. "up" changes as flight progresses along its legs
  • constant resize or rescale operation as new images are added will likely be expensive
  • Leica wants to share a buffer (the canvas) between UI and their server
  • I don't see how to share buffers between Java GUI and their stuff (not a very distributed system)

• Wednesday, 10:30, Leica will present their ideas on a new waterfall model

• Leica is still learning STL and templates in general.