Sunday, July 31, 2011

IGARSS 2011 Vancouver - Redux (So that you don't fall asleep)

It took a lot of messing around and changing flights to get there but thanks to the excellent folk at Flight Centre Melbourne Airport I arrived on time , left Melbourne 25th July Morning 10:30am and arrived in Vancouver 25th July 7:30am. Again proving the fact that a teleporter is indeed equivalent to a time machine.

Screen shot 2011-07-31 at 6.03.47 PM

This monday turned out to be full-on 48hour monday, lots of interesting sessions to go to. I attended the plenary by Shimada San introducing ALOS-2 and ALOS-3 with next generations of the great optical and SAR sensors we have come to expect from the Japanese. May be I can sneak into the CALVAL stage and score another trip to Japan.

More flight dramas continued in the background, the presentations blurred into a constant stream of words and diagrams on my sketchbook in the foreground. I learnt a lot about what has been done in the past year and what is coming over the horizon.

I saw science solving political problems with cross-boundary river basin monitoring with Poseidon, Jason-1 and in the future Jason-2. Important basins across unfriendly countries such as Indus, Ganges, Brahmaputra and the Mekong can be monitored by the international science community and appropriate alerts sent and action taken when one of the partner nations or the weather in the river basin begins to misbehave.

I saw the cold-gas (NO2) driven engines that keep TerraSAR-X and Tandem-X locked in their helical dance, not forever, only till the global elevation model project is complete or the gas runs out - whichever comes first. Some extra effort will be required to collect the elevation model of Antarctica, since the satellites will need to be rolled to face left. This is where the polar observation satellites - Cryosat-2 and more, will make a great impact.

I used the conference services a fair bit as well and got labelled "the troublemaker" - thanks people. If you want to get in touch with me just search for the troublemaker at IGARSS.

The real highlights of the week however were the impromptu dinner get-togethers. After the poster session on Monday I drifted off with SAR people from Spain and a few other places. Tuesday night was spent with the OTB team and a fun Columbian guy from ITC. Scientists on one night, engineers on another, we discussed the transition from Science to Engineering to Products, and the effort and mindsets it takes to go the next step, and how the last step provides continuing fuel for the first. The next night it was back to science setting up a collaboration aimed at just such a transition for the OTB ESA Summer of Code in Space (ESA SOCIS 2011). Thursday evening I had great stroll through the Sun Yat Sen garden, dinner at Shukaku and a movie at the Scotiabank Theatre - ultimate self indulgence. Friday, went to very few sessions and spent most of the time in Stanley park sketching totem poles, before catching a flight out. Thank you for a great time Vancouver - I will definitely be back.

Thursday, July 21, 2011

Getting some fish into WorldWind

Last few days I have gone back to biological data, well animal tracks. CSIRO has put together a very nice oceantracks website using the amazing Indie game engine - Unity 3D (amazing in every way except running on Linux). It runs on Android, so with some effort on GL-ES front it can be made to work on Linux.

Talking about multi-platform game engines the other impressive one is Ardor3D,  not that OGRE or Irrilicht are bad. It is just easier to get going without getting platform specific binaries or needing to set up a large tool-chain. It is also easier to maintain the Java based game engine on top of JOGL and LWJGL instead of maintaining JNI bindings to a C++ engine (especially if your bindings are outside the source tree) . The other large draw from Ardor is the JOGL based rendering which lets us share the GLContext between Ardor and WorldWind, as I have done in jaxb-collada (lots of thanks to the original Ardor hack post and improvements from niccastel and brujito). Some pretty famous use-cases have picked Ardor up as their rendering engine including the Mars rover project, I hope to catch up with them in a few weeks. May be the local guys down here in Hobart will adopt it as a mission control system for the SLOCUM Gliders.

There are some altitude setting bugs in the Ardor3D+WorldWind code , and the animations key-frames are not enabled, but the models load nicely and will look great with some more water surface shaders and actual track data.


Friday, July 15, 2011

GSOC 2011 - Sub-surface Features in WorldWind

Honestly I did not sign up to mentor this project, but I kind of drifted into throwing some advice by the way of the student. Doing a prototype of sub-surface visualisation is not at all difficult, but it is difficult to get it right, to convey the information without losing the context provided by the draped textures.

California Seismic or Australia Boreholes data goes down into the earth's crust. The crust is not that thick (only a few 10's of km in the 6000 or so km's of earth radius). Thinks get a bit weird in portraying things accurately while using vertical exaggeration. The surface of the earth becomes a bit of an event horizon. Nevertheless good progress has been made with changing the transparency of the surface textures and deploying secondary meshes below the normal terrain mesh. The aim is to get something close to Rockworks. The implementation so far tweaking secondary meshes in WorldWind can be seen here: