March 24, 2011 | Team 10 | Progress Update #3
Full integration of our newest pattern matching algorithm proves to be very successful. The above video demonstrates Kylee signing "I want (a) Map". We see her signing, the program tracking and the eventual recognition displaying "I (I am) want map.
March 6, 2011 | Team 10 | Progress Update #2
Our previous progress update showed the successes we had with the recognition of the 'wave' movement. Unfortunately only Kylee's movement was able to trigger a successfully recoginized movement as it was her movement we recorded for the library. To rectify this we developed a normalization function which would allow people of all shapes and sizes (to a degree) to use our program. The normalization function will take a users movement data and normalize their dimensions in order to accurately compare their movement with our library.
February 18, 2011 | Team 10 | Progress Update #1
Today marked the first time we were able to successfully recognize prerecorded movement. Movement is detected using our software with the Kinect to record a segment of data. This Data is then compared against a library of movements we have prerecorded and upon recognition within a certain degree of accuracy we can return the type of sign found. The video found above shows Stephen Lau, a Team 10 member conducting a single 'arms down' movement. We then are able to see the response via the console window as the program is able to calibrate, monitor, and successfully detect the movement.
The OpenNI organization is an industry-led, not-for-profit organization formed to certify and promote the compatibility and interoperability of Natural Interaction (NI) devices, applications and middleware.
PrimeSense's NITE Middleware focuses on enabling Natural Interaction™ in the living room using two key applications: Control by Gesture and Games for All.