(3D point to 2D) Point and Inverse (2D point to 3D ray) Camera Projection

The goal of this project is to implement forward (3D point to 2D point) and inverse (2D point to 3D ray) camera projection, and to perform triangulation from two cameras to do 3D reconstruction from pairs of matching 2D image points. This project will involve understanding relationships between 2D image coordinates and 3D world coordinates and the chain of transformations that make up the pinhole camera model that was discussed in class. Your specific tasks will be to project 3D coordinates (sets of 3D joint locations on a human body, measured by motion capture equipment) into image pixel coordinates that you can overlay on top of an image, to then convert those 2D points back into 3D viewing rays, and then triangulate the viewing rays of two camera views to recover the original 3D coordinates you started with (or values close to those coordinates).