Right now, I am involved in a neat university research project that uses the camera on a camera phone to enable motion sensing. So instead of using dedicated hardware such as accelerometers (which all of the current motion sensing phones use), it's all software-based. YouTube Video: (dated, but shows some of the stuff that has been done) While the main purpose of this is for productivity, the gamedev side of me can't resist the prospects of developing motion-sensing-based games. The tricky part, however, is realizing that unlike a Wii, which separates the controller from the screen, a lot of ideas that work using that paradigm fails, because moving the phone moves the screen too. What game ideas do you have for a setup like this? A WarioWare Twisted sort of compilation comes to mind as does a Monkey Ball type of game, but I'm sure that there are a lot of potentially great ideas out there. - Jon

Very impressive :) Simple but effective. It’s one of those ideas that are right in front of you but you never bother to look :P

But as you stated before. The main problem is that the screen moves with the camera so the player wouldn’t be able to see what’s going on when the camera moves fast :/ obviously you have to make games that uses simple motions like tilting.

Your project can recognize up, down, left, and right motions. How about zooming?? Can it identify if you are moving forward or backwards??

If it can, you can make a FPS (with 2D graphics if it’s not possible for a cell) where you can rotate the camera around you and look up and down and see the game room rotate with you. And if you move forward or backwards you can move in and out of the game room. If you can’t than a simple directional button will still work fine.

Again, beautiful work :) this is one of those idea that can change the game industry :P

Quote:

Original post by Farraj Your project can recognize up, down, left, and right motions. How about zooming?? Can it identify if you are moving forward or backwards??

It can recognize tilting (pitch, yaw and roll if you call it that).

3D graphics support depends on the phone model. Some have OpenGL ES, but many still do not unfortunately.

- Jon

It's a cool idea, but isn't the technology one $2 gyro chip away from obsolescence?

There is a game on the market, Tilt-a-World, that uses this same basic premise to roll a ball trough the world.

I certainly like the idea and the originality behind it but there is a lot of machine vision problems involved here and the controls became very difficult when you used very plain backgrounds the camera was pointed at.

Regardless, keep up the good work! I love seeing new people coming to the mobile industry bringing unique and new ideas, its what mobile games need more now than ever.

Quote:

It's a cool idea, but isn't the technology one $2 gyro chip away from obsolescence?

This technology is coming, and its going to be cool! Have a look here for an older phone that has it. Also, of note, is the new Java Platform 8 coming from Sony Ericsson which has support for JSR-256 (J2ME accelerometer support).

Quote:

Original post by Sneftel It's a cool idea, but isn't the technology one $2 gyro chip away from obsolescence?

If the technology matures, then cell phone manufacturers will do whatever it takes to trim costs, even if it's a couple bucks for an accelerometer or gyroscope.

Price aside, the camera approach has its pros and cons.

Pros
- Completely software based.
- Works when you are in a vehicle. (I remember WarioWare being impossible to play in the car due to this!)
- Detects motion in all 6 axes. Some hardware-based approaches don't detect all directions.

Cons
- Does not work in pitch darkness, but still works in relative darkness. Then again, I don't know who would play in such conditions without gouging their eyes out. :P
- Consumes more battery life.

So there are advantages and disadvantages to this approach, but it's a viable alternative to the accelerometer based phones on the market today.

What's more important though is developing smart applications for them, something that isn't available at all right now. The phones that do have motion sensing, sadly don't take advantage of it at all yet, and a big component of this project is developing applications, menu systems and other components that work better.

A few interesting applications come in the area of text input and gesturing, and we're working hard at developing new menu systems.

As we have found out through usability studies, navigating a linear menu with this sucks. :)

Even if our approach ends up taking the backseat to hardware-based approaches, all of our applications will still be relevant, and that's where I see a lot of promise.

Once we publish a paper comparing those menu systems, I'll be sure to share the results with you. I'm very curious myself to see which of our approaches pan out.

- Jon

Theres already a game on the market that uses this it lets you shoot flies off ppls faces.