Before we got too deep into the SDK portion we decided to try out a Media Center application that I read about online a few weeks ago on Engaget (http://www.engadget.com/2012/02/15/kinect-for-media-center/). Initially we had some trouble getting the Kinect to recognize us but eventually we managed to get it working. We quickly noticed there seemed to be some latency in the interaction, and it really did not feel "natural" to wave through the interface. We were unable to determine how much of this was from our position and movements, the lighting in the room, or just poor coding. It worked, however it felt far from perfect and we were often times pushing the interface in one direction only or we were unable to select our intended target. For a cost of $6.99 we didn't find the app to be anything revolutionary, but it was still helpful in giving us an easy way to experience the Kinect and it fueled discussion for things we should consider for our project.
We decided to set aside the Windows Media Center application, while Karen dug into the Microsoft SDK several of us started sketching ideas and acting out scenarios of how a user may want to interact with the energy controls at the CDR. Later we regrouped and tested some of the code that Karen started implementing. The unit has a motorized tilt function that is controlled through an interface in the SDK, we also tested the range values for optimal distance from the sensor. Through looking at the range precision we can tell the hardware is capable of reading precise distances, however we also foresee some challenges with the limitations of range and lack of peripheral vision on the Kinect sensor.
Slideshow from our meeting on March 4th at the CDR
No comments:
Post a Comment