FreeTrack Forum

I'm sure you've already took big steps in this guys but I thought you may want to check it out if you've not already done so:

http://wiiyourself.gl.tter.org/

it's a wii remote library for C++, just like Brian Peek's one (that being on .NET).

Thanks

I would also point that due to high frame rate and resolution of wiiremote's IR cam, I believe practical uses of Free-Track  will be more than head tracking.

Do you guys plan to implement data capture from more than 1 wiiremote for highly accurate 3d positioning.
That way point model can be used as a real time  3d pointing device.

Thanks

Hi!

Apart from Wii Remote there are coming new devices that might be interested for an integration:

http://=http://www.engadget.com/2008/05/30/asus-eee-stick-motion-controller-brings-wii-to-the-eee/

Bye!
S.

Motion sensors tend to loose their calibration, that's why Nintendo came up with IR solution, which is great. I can't see a similar thing on Asus' product.

What I mean by 3d pointing device is something (in this case 3 ir leds and Free-Track API) to map real life 3d movement to 3d world behind your screen through your screen without loosing calibration.

I'm not talking about a solution for existing games and software, maybe there are some titles that can take advantage of it, but this technology can be used for new ideas in interaction and entertainment software.

I don't think that a Wii remote would offer much of an improvement. (At least not for that cost!) I'm doing some research on this subject, and it seems that a regular webcam should be accurate enough to a distance of about 4m at a resolution of 320x240 (and accuracy scales linearly with resolution), and 30 fps with interpolation is already quite smooth. This way or another, the main issue with using FreeTrack (or anything else) for anything other than headtracking is that when you're doing headtracking, you may take the liberty to make some assumptions when matching projected points with real points, based on the fact that the user is facing the screen. Such assumptions are difficult to make in other cases (tracking a sword with IR lights, for example) without constraining the possible motions. Support for several webcams could be useful, though, for improving the accuracy of headtracking and solving the narrow field of view problem. (If the cameras are strategically positioned, that is. Putting two webcams some 20cm apart horizontally, as you'd do to take stereoscopic images, doesn't improve accuracy much and doesn't solve the narrow field of view problem at all.) A funny idea that crossed my mind is to dump lightwave-based tracking altogether and try tracking using sound waves. You could have several microphones placed around the room and several sound sources with different frequencies representing the points to be tracked. You could then use some FFT to break the recorded samples into the basic frequencies constructing them and to analyze the delays and differences in intensities from the different microphons to approximate the locations of the points. That would eliminate the narrow field of view and the point matching problems, allowing for tracking of the most arbitrary things in the most arbitrary positions.

I thought sound waves thing as well, but major problem with it is the fact sound waves will bounce within room walls and returning to receiver, decreasing accuracy a lot imho. If it was possible some company would have implement it already. IR light is the best way to go so far.

As for working with a limited FOV, that's true, yet within that area we would have a quite good implementation.

Well, you see, the reverberations obviously WOULD decrease accuracy, but the question is by how much. You'll never know until you check. And "another company would have already implemented it" is not an excuse. In fact, I'm pretty sure somebody has already implemented it, for different purposes. (Perhaps in the context of robots and AI?)

Another thing to note is that you generally don't need accuracy.
You just need your sword or whatever to swing in roughly the right direction.  
Either way, this sounds too appealing to skip without trying.

[EDIT]

My main problem is that I don't have a lab here, with microphones, webcams, Wii-motes and IR light sources, so I can't check the stuff I come up with.