[HELP] NEED Everyone's Help for MOUSE Selection!

allenpan

[HELP] NEED Everyone's Help for MOUSE Selection!

i need a mouse that can still track with highest Z-axis/height!

i am utilizing mouse as my robotic encoder, as result i need for my robot, but i dont want robot to drag is along the floor, but rather suspended! please let me know what your mouse is , type and max height

*Z-Axis is the height that you can held the mouse in mid air and without loosing the tracking.,

Example:
Mouse NAME:
Connection Type: USB/BT/RF
Max Z-Axis:

GHoosdum

What about a gyro mouse?

CB

I've got a Logitech G9x (USB). Seems like its limit is only about a millimeter from the pad.

RyanFodder

Razer Naga
USB
approx 3-4 mm

I could sometimes get it to go as high as ~ 5 mm but it had to be nearly perfectly flat.

allenpan

gyro mouse only working when due to acceleration/tilt, nor matter how I mount in the robot it only can take note of inertia but not much of location; when u move the gyro mouse in mid air in any direction I tells the api to move system cursor in that direction perpendicular to the screen movement.... and my robot is moving in a horizontal plane, which does not work in that manner...

however with a mouse, whatever you move it can be correlated the pixel counts with actual distance, hence actual movement tracking.

car's Odometers is very similar to how a encoder/a scroll wheel works in a mouse, it has slotted wheel what keep tracks of axial rotation, but it does not translate to real life how far it travels, why? like any vehicles the tire might be slipping due to road condition, and this happens even more with tread base vehicles, so if tire slipping it still assuming the distance is traveled due to the rotation.

this problem can be corrected through machine learning and error filter, but that means much more computational power, I don't want to use "any" gpu cycle for something like that, so mouse is that I think the most easiest and simplest way to go for it!

RyanFodder

I made a robot with photo sensors once. They ate unreliable at best, unless you are operating on a giant, uniform surface. You will likely get better quality with accelerometers in a phone and integrating to get position data. Else, you could use your tracks with rotation sensors... more processing to determine turning, etc.

What about laser range finders? if all you want is relative position (x1 - x2) you could so that pretty easily with two range finders with a digital output.

allenpan

it is using kinect, which does have laser range finder, however, it is subject to noise and additional processing power, so far mouse tracking gives me the best result...uneven surface can be caliberated with 3 mice triangulation (1 cosh funtion, very little cycle)

rotation sensor does not correct when the treads or tire slipping on the road and need error correction and machine learning, i really dont want to add another control algorithm

midga

Why not just bust open an old ball mouse and roll the tracker along the floor?

allenpan

same problem, contact floor does not work well, if it slipping i will get error data again, the optic or location tracking is the best

the ball mouse will be exactly the same as if i put tacktometer on the motor shaft.

midga

Right, okay. Well, you might be able to build some sort of isolation thing for the optics. Basically, put it in a tube that points at the floor so it's dark. That way less external light interference would mean a better signal to the receiver. If you used a flexible material kept within a tight enough tolerance, you should even be able to have it reach the ground without interfering with the signal or your operation. That should give you a bit more distance you can raise it up.