2017-09-20 Meeting Notes Week 38

Date

20 Sep 2017

Time	Item	Who	Notes

Mariusz Eivind Grøtte Investigate references for HSI atmospheric correction (lit review)
Unknown User (joaoff) Get some analysis of HSI data in open ocean
Unknown User (julianve) Get CAD model for camera interior (lenses etc.), must be something we can model with?
Unknown User (orlandic) Software architecture and onboard processing pipeline
Unknown User (rogerbi) Ask Tor-Arne for computers and Terje/Nina for access to the lab
Unknown User (julianve) / Mariusz Eivind Grøtte Ask Fred how easy it is to change the optics and SNR designed for
Mariusz Eivind Grøtte Redefine what is 100 GSD in our documents and for mission reqs.
Unknown User (sivertba) Ask someone who might be interested in thermal modeling
Mariusz Eivind Grøtte Create Gantt Chart

Working on timeline, need input from everyone
Gantt charts and milestones etc.
Looking at the slewing motion and trying to model the target area. - Optimal slewing rate calculation.

Viable results from 400-700 nm range
3000 lumen ref. to what (from a Master's thesis) for signal measurement. FFI tools may support to infer noise --> SNR
Previous calculations on light intensity in space http://nuts.cubesat.no/upload/2015/03/06/prototype_design_for_cubesat_by_thomas_hanssen_nornes.pdf
SNR discussion, multiple sources of noise: water to camera, thermal noise, electronic noise, conversion noise. Not just light intensity.
Seeing the ocean is not the problem, if it is dark it’s dark (no data). If we are looking at something we want to see, like algal blooms, there’s a lot there that will reflect the signal.
Drone data in Joao’s report was from shallow water

There’s a graphic that explains the latest prototype.
Julian will start the pcb layout this week.
The sensor is a NIR enhanced (more sensitive over 700nm) version of the one that is flying on the drones right now.
Got CAD files from Fred but only of the house, not lenses.

Working on getting data from memory to any type of processing system.
Agreed that the first data from the sensor can be 12-bit data packed in 16-bit buffers. If needed we can repack it as 12-bit data for communication.
will need updated code from Idletech with a more realistic data cube.
writing code for different matrix implementation.
working on basic jpeg compression.

No update until we have met with all the platform providers, possibly in October or November.

Next week, Tuesday from noon to 1pm Julian, Milica and Joao will meet with Idletechs.
Follow up question: Is it possible to have computers in the lab soon? Ask Tor-Arne for computers and Terje/Nina for access to the lab.

Need to find anyone who is interested in looking into Thermal expansion simulation.
- Need CAD model to do simulations on thermal impact. Information from Fred on how much displacement of the lens can be tolerated. Then compare with simulations of different materials.
- Sivert might have someone interested in doing that.

100m ground sampling distance and ground sampling resolution is both being used in the specification. Need to decide which one that is going to be used for presentations.

- We might do 50m ground sampling distance. Better processing means lower ground sampling distance and higher resolution.

- If ground sampling distance 100m is the case, that’s achievable and a conservative estimate.
Ask Fred how easy it is to change the optics. Maybe include Fred on one of the weekly engineering meetings.
Is atmospheric correction is necessary?