In this part, we create a 3D model of a 2D image by following the paper 'Tour into the Picture' by Horry et. al. I computed the focus as a function of the height and width of the backplane. By varying the constant associated with this function for few times, I could obtain reasonable results. All images (apart from the one provided in the project and another one I got from Flickr) are captured by my mobile phone. In this work, the user specifies the backplane and the vanishing points. From these inputs the dimension of the 3D box is computed and then each plane is rectified using the homography code of part A of project 5 and then texture mapped onto the surfaces of the 3D box. I added few wacky novel views by rotating the box by more than 180 degree. I think they look pretty cool!!!
I used the webpage of Natasha, our TA, to format my submission.