Monday, November 21, 2011

While the last experiment had a somewhat constant thickness for the line drawing, in this experiment the line is less constant in its thickness. I've changed the settings for the response time of the line, while 0.1 leaves the line to react in a later time, this experiment is set to 1 and reacts right away to any blob it detects and sometimes create large triangular shapes. At the start, after the waiting, I do a simple tutting position in which it creates a large triangle between the two arms.

The graphic style as a vector is different for both the thin constant and large random thickness lines. While the former graphical style seems much sophisticated and the user may see the direct corrolation with his/her arm movement to the line. The latter creates a much abstract rendering of our movement. Either line is much more interesting when the user is able to see it as a 3D object. In which it becomes an enlightment stage of the object being directly related spatially to their body movements.


Thursday, November 17, 2011

Trying out a vector graphic style drawn in 3D space. By erasing one of the vertex points created for the 3d line, it creates a vector line object. The disadvantage to this is that, the drawing looks seemingly flat without the 3d rotation tool to work. But when implementing the 3d rotation by detecting 2 blobs, most users got confused or had difficulty controlling it. There seems to be a different space to realize drawing versus controlling the object with 2 hand blobs, but most users do not realize it for they only watch the line drawing.

The other alternative is to increase the minimum detecting size of the blob so a full human body maybe detected. So one person in front of the system draws lines, while participation of two people controls the rotating of the object in 3D space. In this video the camera detects smaller blobs, so the object rotates even with one person.



Using: Memo Atkens kinect3Ddraw demo, open Frameworks, kinect