A new demo in the work

This one would be less about usage real-world objects, more about phone camera as 3d mouse.

Another obscure part of symbian lore

Renaming/changing application uid is not so trivial. Not only you have to change name and uid, but also the name of aif file(application icon). There could not be two different application with the same icon name.

Augmented reality contact lenses could be little more close

Slashdot report Samsung showed prototype of colored E-paper based on the carbon nanotubes. Carbon nanotubes used as underlying electrodes, they are completly transparent and very thin. So we may be more close to AR holy grail – working active matrix contact lenses(Vinge’s Rainbow Ends anyone ?. Or at least good enough transperent see-through display glasses.

Advertisement

Nokia E50, E51 camera API – strange

For some reason crushing on

SetExposureL
SetDigitalZoomFactorL
SetContrastL

Though they should always by supported…

.

Version 0.02 of the demo is up

At my homepage
New
1. Objects scanning improved
2. The manual improved and extended.
3. Some bug fixes and rebalances.
4. Some game environment changes.

5. support for E51 and hopefully for E50

Nokia N96 cpu-wise is worse than N95

Mystery about Nokia N96 resolved – it’s CPU is not OMAP 2420, like N95 (ARM11 330Mhz, floating point, 3d accelerator). It has 264mhz ARM9 instead, without floating point, without 3d accelerator. So it can’t possibly run version of the demo for 9.2

Hands detection

Being doing some reading on hands/skin detection. Most of methods, even “fast” are computationally heavy and would hardly suit for modern mobiles real-time. Well may be for highest end mobile 500+ Mhz some would work, if cpu doing nothing else. Looks like for 300+ Mhz, with a lot of other stuff going one have to improvise something not quite robust.

RjDj

I’ve been seen some RjDj videos This audio augmented reality is really impressive. I’m wondering what effect would have tying together audio and 3d input into single augmented environment.

Reception – first imppression

First reception of the demo is positive. It seems generally users have no problems with concepts of fiduciary markers and keeping them in the camera frame. Mostly people want more stuff and gameplay, as always, but seems there is no problem with usability. Well, of cause there is problem of Nokia platform fragmentation, but that is not exactly AR related. Well, it could be. Mobile AR applications are CPU hungry, and that force them to squeeze device hardware to its extreme. Which make AR apps especially vulnerable to platform fragmentation. I’ve already mentioned strange problem with N96 in previews post.
The concept of tabletop marker-based game seems viable.
The main question is, how to bring more real-world interaction into game ?
Right now we have phone as 3d mouse and small objects scanned and placed into game.
Make markers movable ? Wouldn’t it hinder gameplay ? Scan surface of the table and use it as game “terrain”, with game physics properties imported from real table surface/objects ?
Questions, questions…

What have I learned ?

I’ve finished porting my AR Game prototype to Symbian 9.* (Nokia N95 and like).
OK, what have I learned ?
1. 300+ Mhz CPU is quite enough for multimarker-tracking based AR.
2. Floating point and fixed point calculations. First version was for Symbian 7.0s, without hardware floating point support, so I tried to make exposure to floating point minimal. As result, on the Nokia N95, which have floating point (VFP) I don’t see any difference in performance between software floating point version and hardware floating point.
3. Symbian platform fragmentation strike again. The game run on the Nokia N95 (s60 FP1, Symbian 9.2), but 9.2 version refuse to run on the Nokia N96 (s60 FP2, Symbian 9.3) Ironically 9.1 version is running on N96.
4. Use optimization ! performance difference between O3 and no optimization is huge.(I use GCCE compiler)