3D Tracking
Cameras are placed orthogonal to each other. Each web cam provides only 2D information, x and y. The system gets the "depth" information, z, by the 2nd camera.
The tracking program finds the 2 brightest spots on each camera, applying a smoothing algorithm to detect sharp changes in light position.
Every frame the 3D position of each LED is written in the shared memory reserved for communication with the design program and read by the rendering program.


Hardware/Software configuration
The system uses 2 web cams, 2 LED’s, 1 laptop, OpenGL Performer for 3D graphics and DirectX 9.0 to manage the two web cams, tracking two lights on both web cams continuously. Each light detected on the first web cam is matched to itself on the second web cam. The light’s position information is written on shared memory by the tracking device and read by the design module generating the surfaces. The software exports the shapes to VRML format to be read in 3D Studio and Auto Cad.
The system requires minimum calibration from the user. The surface rendering application reads the LED's positions from the tracker. These positions are them united in triangles forming a triangle strip.
The shape is generated as the result of an imaginary line uniting the two lights as they move through space.
Unlike other 3D sketching systems, this method allows for a varied range of surface types to be rendered because of the varying distance between the two lights. Their relative position can be constantly changing with different widths along its length.

Open 3D tracking diagram >>>>

...loading