More Kinect SDK v2 from the field...
- Posted: Dec 26, 2013 at 6:00AM
Looks like it's going to be a Bruno Capuano week. He's recently received a Kinect for Windows v2 Developer Preview Kit and has continued (Here's the first set of posts El Bruno Kinect for Windows v2's) to share his experience.
What, you want one? Well you better hurry, [Special Edition] There are 500 more slots now available in the Kinect for Windows v2 Developer Preview Program
So here this
- It uses the Kinect Sensor of XBoxOne
- Better quality and more horizontal and vertical amplitude for the camera and a depth sensor
- Full HD Color
- Infrared light (30 fps)-independent
- Improvement in 3 X depth sensor
- Improvements in the microphone (zero balanced, that I don’t have very clear it is)
- The Skeleton can detect up to 25 joints with 6 people at the same time (25 joints in the skeleton are detected for each person)
- Detection of thumbs, contour of the hand, and open and closed hand gestures
- Improvements in the detection range (from 0.5 meters to 4.5 meters)
- Several applications can use the sensor at the same time
In the last post of the series with basic stuff we can do with KinectSdk V2 and the new KinectOne today I’ll write about how we can “draw a body” or what we use to call a skeleton.
As in the previous examples, we must create a Reader of the type BODYREADER, initialize it and subscribe to the FrameArrived event. In it then
- a couple of null validations
- We get the frame of the Body feed
- We update the array of bodies (remember that with the KinectOne, we can follow up to six bodies simultaneously!)
- for every body access to its collection of Joints and paint them