| Summary: | 碩士 === 國立臺南大學 === 電機工程學系碩博士班 === 107 === In recent years, navigation for pedestrian has been widely used on mobile devices. By marking the pedestrian’s location and destination in the map, a recommended route can be planned through the navigation system. According to the researches, if the pedestrian is in an unfamiliar environment or does not know where he is, it is easy to cause anxiety. We propose to calculate the most familiar and safest route based on the pedestrian’s familiarity and safety. To alleviate this problem, the historical travel data are recorded in the system, and then the road usage frequency and time are referred to conduct each road’s familiarity weight. We design the road cost function to calculate the cost of each road with the road familiarity weight when planning route. Open Source software’s are used to build our system, such as OpenStreetMap and Graphhopper, which are exploited to implement our proposed algorithms in this paper. The system provides navigation service on the form of Web Service, which can be accessed by not only mobile phones but various wearable devices, such as watches, glasses, etc. This makes us easy to integrate various devices with our navigation system in the near future. Finally, the algorithms proposed in this paper is verified by real experiments and field trial. The accuracy and effectiveness of the system are displayed in the paper. Traditional navigation algorithms are also compared to the proposed method. It turns out that our method does provide the most familiar route for pedestrians.
|
|---|
