In the last two days I alone wrote a prototype that can take a whiteboard photo, and automatically turn it into a mindmap-like zoomable chart. Pieces of the chart then can be rearranged and altered individually:

https://prezi.com/igaywhvnam2y/whiteboard-prezi-2015-12-04-152935/

This was part of a company hackathon, and I had some infrastructure to help me regarding the visualization, but with the shape recognition/extraction, it was just me and the nasty python bindings for OpenCV.