Inspired by a series of competitive matches on the foosball table in the informatics leisure room on campus, a fellow PhD student and I decided to build a “Foosball Game State Tracking” system as our computer vision semester project. Besides from having an excuse to play even more foosball and test my new camera, we were mostly interested in implementing a real-time vision system and working with the almighty Kalman filter.
In a nutshell, we filmed ourselves playing foosball and then used MATLAB to build a system that takes in the video stream and is able to locate all players, estimate their rotational angles, track the ball position and do all of this in real-time. Our test results can be seen in this YouTube video. If you are interested in a more in-depth explanation of what we did, we have a project report as well.
During spring break 2012, my girlfriend, her father and a friend of his went down into the heart of Kentucky to go hiking. We chose the five mile Buffalo Canyon Trail, which – both due to the rough terrain and me still thinking in terms of kilometers – quite honestly felt more like ten. Anyway, it was a lot of fun and finally a chance for me to play with my new camera. Here are some results:
This month, I finished up my bachelor thesis at Cologne University of Applied Sciences. I fully designed and implemented a webcam-based system that can optically capture the three-dimensional pose of a puppet and map it onto a corresponding skeleton. The skeleton can viewed in 3d and can even be saved as a COLLADA file (and thus could be imported into professional 3d programs like Maya®).
A quick demonstration of what my project can do can be seen in this nifty YouTube video. Also, if you are really interested or just really bored, you are more than welcome to take a look at the whole thesis.
Los Angeles/Hollywood, Universal Studios:
Malibu, Santa Barbara, Santa Maria, Pismo Beach:
Monterey Bay, Aquarium:
