Tracking And Detection assignments - Winter Semester 2018
The first exercise deals with multiple images of a teabox. A 3d ply format is provided of the teabox. We have to compute the pose of the camera by manually selecting the corners of teabox in the image. This gives the correspondence from image to 3d polygon. Next, we have to compute camera pose using these correspondences. Once camera pose is known, compute SIFT feature points of each image of the teabox and backproject the SIFT 3d points on to the surface of teabox in 3d. Once we have SIFT features corresponding to location on 3d surface of teabox, we can advance to next exercise.