Quick Project Update

So I've been spending quite a bit of time playing in Simulink trying to track the fish from pre-taken videos. What I have would work from a real time input as well, but we don't have continue access to the fish tank. There are still some random blobs that need to be filtered out, but this is off of a single video stream with some minor signal manipulation. The simulation runs much faster than my initial attempts as well (working on a net book can be a pain), and produces very similar results. At the beginning of the video there are a bunch of mistake blobs this is because of the webcams built in auto focus. I also have not yet figured out how to track the lighter fish. And the dark fish cross paths and become "one" fish for a little while. But we're definitely much further along than we were before. Prof. Lubos says he has some suggestions as well. But we've not yet had a chance to discuss these with him yet.

(The video is pretty highly compressed for the uploading sorry)

There is still a lot of work to be done, but I think what I have thus far is pretty cool, and has been quite fun to work on thus far. My friends have to drag me away from the computer to go do things.