Week 3 began with an unboxing. I had just got my hands on the Azure Kinect DK, which I could use as a body tracking camera, as well as harness its inbuilt IMU (inertial measurement unit). The Kinect system has four different cameras to choose from. Three in 2D; infrared, colour video and depth. As well as the 3D camera.
The system runs on command prompt. Something I had never used. This would be a challenge. The first probelems arose when the guide provided by a certain tech giant assumes you are the epitome of an expert in this field...I am not. Setting the camera up took 3 days of trial and error and watching countless videos on youtube (other streaming platforms are available).
Finally, the camera was up and running. However, not all was smooth. Literally.
In order to make use of the camera, i needed to record more than just video footage. I needed the position data of the individuals the camera is picking up. This is easily done with Microsoft's Body Tracking. However, for those unfortunate enouh to remember skype, you will remember the shocking framerate issues. This was the equivalent of skyping the skeleton of yourself whilst looking through a kaleidoscope as the camera output when recorded was approximately 7fps.
My hardware wasn't powerful enough to do the task at hand. A problem when sampling data, but given the scope and duration of my research, it is not essential to have smooth running, it is more important that it works.
Week 3 has been a challenge. I would be lying if i said it wasn't. However, it has taught me valuable lessons in problem solving and troubleshooting, with the use of resources around me proving to be invaluable.
I sign off for a few weeks now, with weeks 4-6 coming during August. Hopefully week 4 is less gremlin-ridden than week 3!
Arron J Thompson