MountainWatch – Discovering mountain peaks with mobile augmented reality.

The MountainWatch project is a CHEST Call 3 winner from Roman Fedorov and the Department of Electronic and Information Bioengineering at the Polytechnic University of Milan. The project vision is to boost social engagement for environmental monitoring by engaging the user in a societal challenge of producing high quality images, by identifying peaks in real time and overlaying their name onto the user generated photo. MountainWatch will thus present itself to the user as a “cool” augmented reality camera for mountain peak detection and mountain photo processing. It will also as an active crowdsourcing interface for massive environmental data collection.

The project team is working to create an augmented reality mobile application that provides an entertaining experience of mountain discovering and sharing. MountainWatch will be a powerful tool to have while trekking or climbing a mountain, allowing users to discover mountain peaks with high precision and test their knowledge. It will also offer an unusual way to document achievements and share photographs on social networks.

Since their successful application to CHEST, the project team have carried out much development work. This includes devising computer vision algorithms for precise mountain peak identification and producing real-time augmented reality architecture for outdoor natural objects identification. The project team have also started the testing phase, by producing a testing framework allowing the evaluation of exactly the error that would be perceived by the user using the application (avoiding subjective time vs error trade-offs), collecting a dataset of example sequences of videos, simulating a user using the application, to be used in the testing framework, and by performing the first field tests. The image below (which can be clicked for an enlarged version) is an example of algorithm execution from the MountainWatch prototype.

MountainWatch - Example Window

The project is currently in the prototyping phase and we look forward to bringing you further updates as the project progresses.