TangoAalto is a mobile indoor map app for Aalto University in Finland. The project was supervised under professor Juho Kannala at Computer Vision Lab in Aalto. The involved researchers in the project are
- Professor Juho Kannala
- Arka Sadhu(responsible for point cloud correction)
- Jonathan Granskog(responsible for point-cloud-to-geoJSON conversion)
- Me(responsible for the app)
The objective of the project is to present a complete pipeline of capturing and visualizing indoor environments. Indeed our team captured a university building with Google Project Tango, processed the point cloud, converted the point cloud into a geoJSON and built an iOS app. At the moment the project is semi-open source. It will be fully open sourced once we get extra funds. If you need some help or base code, feel free to contact me.
Demo video – a video is worth a zillion words.
1. Map data comparison
We have both auto- and manually-generated map data. By manual, we mean hand drawn data. We used QGIS for drawing geoJSON from a floor plan. If you are interested in drawing one, check my tutorial. The auto-generated data was created completely automatically with Tango device. We also georeferenced the point cloud automatically by laying the data over the Earth and project it on the surface of the Earth.
Unfortunately the “auto” data has only corridor since we were not given an access to every office rooms.
FloorView is like Google Maps StreetView. You click a path and you see the image that is taken at that location. It would have been much more useful if we had a 360-degree camera.
The path is an actual camera path where a videographer walked.
You can search a building by its name.
There were many features which I didn’t have time to implement such as resource(rooms/hardwares) search, extensive map data annotation, better UI/UX and etc. As well, we need much more data. Let’s hope we get funds to continue this project!