Research On Landmark-based Pedestrian Navigation Methods In Complex City Environment

Pedestrian navigation can be viewed as an approach of guiding pedestrians to their destinations by means of verbal or written directions using digital maps on personal digital assistant devices or mobile phones. In recent years, pedestrian navigation applications have received increasing attention from the mobile phone industry as well as from researchers studying location-based services. In particular, pedestrian navigation has become a critical application for the mobile phone industry. One key requirement of pedestrian navigation services is to guide pedestrians efficiently with a minimal spatial cognitive burden on the users.Many studies have shown that landmark-based pedestrian navigation systems can offer the clearest cues to enhance wayfinding instructions. Landmarks enhance pedestrian navigation systems compared with automated navigation systems in the following ways:(a) landmarks are easier for pedestrians to follow than turning instructions with distance, street name, and direction as given by automated navigation systems,(b) landmark-referencing route instructions can decrease the pedestrian’s cognitive load and reduce the feeling of route confusion, as well as increase confidence in the navigation instructions, and (c) landmarks help to shorten pedestrian navigation time.Although the importance of landmark has been highlighted by many studies, there are still some difficulties exist when including landmark in pedestrian navigation applications:(1) Very few navigation data models in the geographical information science and transportation communities support modeling of landmarks and use of landmark-based route instructions for pedestrian navigation services.(2) Before using landmarks as navigation cues in wayfinding and navigation, people need to recognize them in real scenes. The usability of landmarks can be evidently affected by their recognition efficiency. Few common attributes of landmark (such as name, color, shape, etc.) can reflect the recognition efficiency exactly. The existing landmark salience attribution can partily describe the usability of landmarks, but the description is not directly and exactly.(3) Although the importance of landmark has been highlighted by many studies, relatively few studies have included landmarks in route selection algorithm. In addition, many studies have used landmark as navigation cue to generate route descriptions by the semantic features (eg. name) of landmarks. But it is difficult to use these route descriptions to find landmarks if the semantic features of landmarks are not salient, landmarks lack explicit marks (eg. signate) or people are not familiar with the wayfinding area.Against to the above three issues, several models, algorithms or approaches are proposed in this thesis, including landmark-based pedestrian navigation data model, the analysis of typical landmark features in pedestrian navigation scenarios, and landmark-based route selection and route direction. This study is organized as follows:(1) The research background and significance of this study are introduced, including:the nature of human wayfinding and navigation, the usability of pedestrian navigation system, and the differences between pedestrian navigation and vechile navigation. Then, the related works of landmark, navigation data model, visual searchs, route selection and direction are reviewed. After the discussion of existing problems, the research objectives and contents of this study are proposed.(2) A landmark-based pedestrian navigation data model is proposed. This data model can model landmarks in several pedestrian navigation scenarios (buildings, open spaces, multimodal transportation systems, and urban streets). This study implements the proposed model in the ArcGIS software environment and demonstrates two typical pedestrian navigation scenarios:(a) a multimodal pedestrian navigation environment involving bus lines, parks, and indoor spaces and (b) a subway system in a metropolitan environment. These two scenarios illustrate the feasibility of the proposed data model in real-world environments.(3) An approach is proposed to analyse the usage of landmarks in pedestrian navigation scenorias based on three typical features of landmarks, including:salience, visibility and recognition time. These features can evidently affect the usability of landmarks, and will be included in the proposed data model, route selection and direction approaches in this study.(4) Landmarks provide the most predominant navigational cue for pedestrian navigation. This study proposes a landmark-based route selection algorithm, in which landmarks are included to reduce the complexity of route choice and help people to confirm route choice decisions. Furthermore, a landmark-based route direction approach, so called landmark sequence, is used to describe and represent route direction information, and can provide pedestrian with continuous route directions. Two generation approaches of landmark sequences are proposed later. These approaches can generate optimal landmark sequences based on different optimal objectives.(5) Several experiments are implemented to test and analyse the proposed model and approaches in this study, including the landmark-based pedestrian navigation data model, the analysis of recognition time of landmarks, the generation of optimal landmark sequences based on the proposed multi-objective model, to illustrate the usability of this study.