The overall aim of this project is to expose citizens to the consequences of climate change outdoors and in real-time, using Augmented Reality (AR) glasses and event cameras. With this, we create experiences that visualize multiple aspects of climate change, e.g., increase of sea level and temperatures, at familiar areas, such as a park they typically enjoy, the path they take everyday to commute, or near their homes. Thereby, users will have a personalized connection to the known areas, and better understand the consequences of climate change, which might eventually lead to adapted behaviors. The cornerstone ideas of this project are: (1) to robustly detect environmental features, such as buildings, fields, parks, under unknown and varying lighting and weather conditions (enabled by event cameras). Based on such detections, we will (2) overlay visualizations presented in AR glasses that will take precise locations in consideration to create a realistic user experience. We will use publicly available environmental data and predictions for weather changes (e.g. from ClimateCentral) to visualize as real-time overlay, what and how the environment will change in the user’s everyday environment.

In this thesis project (a sub-part of the overall project above), we develop a wearable outdoor AR glasses prototype by augmenting available indoor AR glasses (here: XREAL Air) with miniatured embedded event cameras.
Following, we develop, implement and evaluate algorithms for robust outdoor scene detection using the event cameras. Those algorithms shall detect basic environmental features (e.g. edges or the base-line of houses and streets), which allow realistig superposition of climate aspects in a separate part of the project.

Prerequisites
Programming in C (high-speed), some experience with computer vision.
 
Supervisor: Jörg Conradt, conr@kth.se