Due to not having knowledge of image processing, I decide to use Kinect for my final project to reconstruct three dimensional vectors of a physical object. My idea is firstly to detect points with designated colors; secondly, to get screen coordinates of the points; thirdly, to map the coordinates to the depth image, then retrieve the z coordinates and infer other points; lastly, map the inferred points from depth coordinates to screen coordinates. This tedious process eventually enables me to reach my partial goal of my final project.

[youtube_sc url=”″]