Discipline: Computer Sciences and Information Management
Subcategory: Computer Science & Information Systems
James Stigall - Bowie State University
Co-Author(s): Sharad Sharma, Bowie State University, Bowie, MD
In emergency situations such as building fires and collapses, lives are lost partially because evacuees have not been properly trained on how to evacuate a specific building. Evacuation drills illustrate to participants the proper route to take and the most efficient evacuation patterns needed. Augmented reality is a technology that blends the real world with computer-generated elements, allowing the user to still interact with the physical environment while being assisted by virtual objects. Motivated by the popularity of mobile technology and increasing interest in mobile emergency evacuation applications, a mobile augmented reality application (or MARA) was developed to help users evacuate the Computer Science Building located at Bowie State University. The goal of this project is to create an emergency MARA system to help people evacuate a building. The system will give a visual representation of a building in 3D space, allowing people to see where exits are in the building through the use of a smartphone or tablets. MARA also shows animation of people evacuation the building as well as smoke and fire. Our hypothesis is that the MARA will lead to better and safer decision making during building evacuation. The MARA was built using the Unity3D game engine and features 3D models built using SketchUp. The Android Software Development Kit was used for the development of the MARA application. The application incorporates the use of markers and camera in the smart phone and tablets. When the user holds his or her mobile device to a marker, the application augments the appropriate floor plan in 3D space, containing avatars navigating towards the exit. Marker detection capabilities were implemented using ARToolKit. Our computer vision method gives good results when the markers are closer to the camera, but accuracy decreases when the markers are further away from the camera.
Funder Acknowledgement(s): The authors would like to thank the National Science Foundation for supporting the project. This work is funded by the Grant Award number HRD-1238784.
Faculty Advisor: Sharad Sharma, ssharma@bowiestate.edu
Role: Developed each 3D building model featured in the application, placed avatars within the models, and matched each building model with a marker.