| Abstract |
The activity in human navigation research field during the past years has been increased drastically. There are and will be more and more areas where electronically assisted navigation becomes an important part of a system solution. Moreover, electronic navigation devices have strongly influenced the procedure of human wayfinding, especially with the aid of GPS technology for the outdoor environments. However, examining the launched indoor navigation systems, it is obvious that opportunities for innovative research suggestions are available. Taking advantage of these opportunities, the original motivation is to create a semantics-based contextual guide for indoor environments, designed for pedestrian navigation regarding user’s physical and mental profile. The Contextual Guide is divided into three different research areas. The first area is occupied with the user profile management while the second is focused on spatial representation and path finding techniques. Finally, the third research area, which represents current thesis subject, is occupied with user tracking during navigation process. Specifically, the system is responsible to track user’s movement on the suggested path and to provide useful information about user location and motion state. For example, if user enters an elevator in order to move from the ground floor to the first floor of a building, the information provided by the system will be: “/User is inside elevator and moves upwards/”. Furthermore, the system is responsible to check if user is following the proposed path and detect any deviations. In case of a path deviation, the system is responsible to dynamically re-plan user’s suggested path from his/her current position. Another case, in which the tracking mechanism is obligated to execute the dynamic re-planning process, is after user feedback. For example, if user finds an obstacle blocking the way, he/she sends feedback to the application about the collision point and system proposes a new, collision-free, path to user’s desired destination. The most important system architectural characteristic is the semantics-based modeling of user motion and location, a feature which provides a powerful advantage in the indoor navigation research field, as the potential for describing relations between humans and space elements is almost unlimited.
|
Semantics-based user tracking and dynamic re-planning for a contextual guide
