Dynamic programming algorithms for reconstructing indoor Manhattan scenes

This code was exported from https://manhattan-vision.googlecode.com/svn and checked in here, using git-svn to retain versioning history etc. That SVN repository (and hence the code here) represents all the latest code at the point that I left Oxford. However, there are numerous data files, build artefacts, plots, and feature files that were too large to check in to SVN, so those are not here. They are on my backup drive at Backups/AVL\ Home\ Dir/Code/indoor_context, and on my macbook pro at Code/old/old_indoor_context_repos.