Project nearing completion

The program now recognizes different strokes, instead of just one long one. I also attempted to smooth out the data by using bezier curves, but there are just too many points being sampled to smooth it noticeably.
I am now using a utility called WiimoteConnect that I found on http://www.wiimoteproject.com . This drastically decreases the amount of time it took me to connect the wiimotes to my laptop.

Screenshot time

As you can see below, I have drawn my name in the program, but the amazing thing is, I drew it in the Z direction! I then rotated the drawing so it could be read.

Drawing is a success

I was able to replace the movable sphere with an ArrayList of vertices, then draw the vertices as a line list. I initially had some problems, the program would run out of memory, and there would always be stray lines at the beginning and end of the drawing. It turned out that I was setting the vertices whether the wiimotes were receiving a signal or not, causing the ArrayList to bloat up with unused points, and waste memory. This was also the cause of the stray lines.
I also made a modification to the IR pen. It appears that if the LED is pointed directly at the wiimote camera, then the light is refracted causing some erroneous data. Also, the side of the LED was not bright enough, so the solution was to wrap the LED in a couple layers of scotch tape to diffuse the light. The IR pen looks even uglier than before, but it works a little better now.
More pictures and screenshots to follow.

Screenshot

Here is the screenshot of what I have in the program so far. The IR pen controls the sphere moving around the screen.

Great progress!

I made an IR light pen according to the schematics I posted earlier. It is not pretty, but it works.
I also have DirectX working together with the Wiimotes in a test application where I can move a sphere around in 3D space. Screenshots to come.

Updates

I found a schematic for an IR pen from Johnny Lee's website http://www.cs.cmu.edu/~johnny/projects/wii/

I plan to go to the hardware store this week and get some ideas for an adjustable apparatus to hold the wiimotes.

Current Status

Apparatus to hold wiimotes - pending. (Needed to wait for paycheck)

Decide on a 3d environment to use - I have done some research, and I believe the 3d portion of my project will use Microsoft XNA, as it is what most people appear to use for 3d environments with C#.NET

Here are the screenshots and pictures that I promised.

The first picture is of my IR light source. Right now I am just using the remote control that came with my laptop, it works adequately, but since it pulses the light, and is not very powerful, I will need to create an IR wand soon.

Here is the Wiimote setup. Without a solid apparatus to hold them, they can swivel,and rock back and forth, making data collection almost meaningless.  But it works fine as a proof of concept.

Here is the screenshot of my program in development. Please click on it to view full sized, as some detail is lost in the compression.

Some ideas and some progress

I was able to create a simple program that used the IR cameras from two wiimotes to determine the relative distance of a light source. This program also combines the two images to agree on a single x and y value for the light source. Screenshots will be added later.

I attempted to make an IR "light-wand" out of an LED flashlight and an IR LED. Attempted. It turns out that the LED that I chose required more power than the one that was originally in the flashlight. So I will either have to find a lower power LED or find a different board to solder this one onto.

My goals for this next week are to make a working IR wand and to compare ΔZ_real with ΔZ_relative with the wiimotes at a fixed distance to be determined later.

Goals For Sept. 16

In yesterday's meeting, two goals were set for next week.

1. Work out the geometic equations for stereoscopic vision

2. Write a small program to test these equations with real input.

Already I can tell that the math is going to be harder than I thought.

Update: The math is not hard. It's the opposite of hard. Easy.  That's a relief.

Introduction

This blog will be used to catalog the goals and milestones of my independent study to create a 3D tracking system using Nintendo Wiimotes