Motion Capture Project
The Motion Capture Project is about making a computer able too ascertain the position and orientation of a object using a single video camera. Thus this deals with computer vision. By processing the input image from the camera, the program tracks anchor points on the object. Some fast image processing is necessary for this to keep up with the frame rate. Using the anchor points and doing a self invented (don't know if it existed but might since it is not too difficult) reverse projection technology, it calculates 3D coordinates of the anchor points thus defining its position.
You will find the project report (with source code) here
If you are interested in seeing it at work more than just the few samples provided with the project report, download the workspace and samples and have a go at it. You will find instruction in this page which is also included in the zip file above.
|