3DMIT

• 

  Contours of rooms recognized from floorplan

Inspiration

As freshmen new to MIT’s campus, we’ve all gotten lost our fair share of times among the confusing building names and connections between them. The maps currently available to us through the Internet and the MIT mobile app are only useful enough to tell us where each building is, but not how to get to specific rooms within those buildings, which is often why freshmen get lost in the first weeks of the semester. We wanted to build a system that would provide us with directions to get to a specific room on MIT’s campus, between floors, buildings, and even the underground tunnels.

What it does

3D MIT gives indoor directions between rooms within MIT buildings. Ever felt lost, not been able to find a specific room for an event? 3D MIT is designed to solve your directional problems!

How we built it

MIT publishes floorplans of all of its buildings online. We used Python and OpenCV to pull information out of these floorplans to create routes from one room to another, and then displayed this information on a webpage.

Challenges we ran into

As one might expect, the image processing was the most challenging part of the project. We tried extracting room numbers using Google’s Tesseract OCR Engine, but we weren’t able to get it working by the end of the hackathon.

Accomplishments that we're proud of

We’re proud of the fact that we participated, and of the effort we put in. We also learned a lot about how to use a lot of new tools and overcame many of the interesting challenges involved in creating a 3D map of MIT.

What we learned

How to use text recognition and image manipulation (Tesseract and OpenCV) How to use Git, JQuery, and HTML Various shortest path algorithms and how they work (Dijkstra’s algorithm, Floyd-Warshall)

What's next for 3DMIT

Recognize room labels using OCR Improve website design Use Google Maps to help direct users between MIT buildings Options for wheelchair accessibility Search (for the nearest bathroom, for instance) Give directions for shortest walking distance or least time spent outside Providing mobile device support or application for on-the-go use Give directions based on user’s class schedules

Built With

• html
• opencv
• python