LibFloorplan Topological parsing and classifying of large floorplan datasets INSTALLATION You will need Boost Graph Library > 1.42 or equivalent WORKS: - Loading and saving graphs from XML, boost::serialization - Area and degree distributions can be generated & saved to file - Filtering based on size, frequency, connectivity etc. - Graph statistics: * Clustering coefficient * Degree distribution * Characteristic path length * Pairwise counts * Computing vertex-wise attributes TODO: Regarding geometric properties, the polygon layout of a room: - Shape (simply square or elongated) - Perimeter length - Number of vertices of the room's layout (i.e. how "zigzaggy" the perimeter is) - Convexity measure of the room's layout (ratio of polygon's area to it's convex hull) - Average vertex angle of the room's layout - Door width Regarding other structure: - Number of windows (only available in the KTH dataset)