Indoor Navigation Map Auto-generation System Based On Smartphone

Indoor navigation is a service which can provide users with accurate location information and guide them to their destinations in hot spots like shopping malls,museums,airports,etc.The indoor navigation system requires corresponding indoor map data.However,the majority of existing indoor map data are collected by field sketch with professional staff and equipment.The high cost of map data gathering hinders the development of indoor navigation.However,on the other side,the ubiquitous smartphones have integrated a lot of sensors,which can be used to collect real-time data about the motion of the users.According to the idea,this disser-tation presents and implements a novel low cost indoor map automatic generation system using sensory data.The main works of this paper are as follows:(1)Based on the research of the existing indoor navigation technology,we proposed a scheme for automatic generation of indoor maps using sensory data from smart phones.The scheme is made up of two stage:pedestrian trajectory acquisition and map gener-ation.Firstly the smartphone uses its built-in sensors to collect data and extracts pe-destrian trajectory from these data,then it uploads the trajectory to the server.Secondly the server gathers the pedestrian trajectory from different users and automatically gen-erates the corresponding indoor map.Compared to other traditional schemes,the data needed by our scheme is less and the complexity of operation is lower.(2)Based on the research of the existing dead reckoning technology,we proposed a tra-jectory estimation mechanism.The mechanism uses the acceleration and gravity sen-sor data to recognize the steps,and then uses the data from direction sensor to separate the trajectory into segments according to the direction of steps.Finally each segment is fitted into a straight line or a Bezier curve.The algorithm can reduce the amount of data in the trajectory information,meanwhile improve the data transmission efficiency and processing speed.In order to improve the accuracy of the trajectory,this paper designs an algorithm to dynamically adjust the threshold parameters according to the training data.(3)The trajectory data collected by the system only represent the relative displacement,which is no help to determine the actual location of the trajectory,so we proposed a trajectory aggregation scheme.The system records the Wi-Fi and iBeacon signal strength along the roads while collecting the trajectory data.When aggregating the trajectories,system compares these wireless signal patterns to determine the spatial relationship between the trajectories and therefore integrate them into a whole map.In order to reduce the cumulative error caused by the dead reckoning,the algorithm inte-grates the trajectory segment by segment to restrict the cumulative error in a single segment.(4)According to the above schemes,this paper implemented a prototype system and tested our scheme in a shopping mall.The system collected the trajectory from the APP on the smartphone and uploaded it to the server.Then the server integrated the trajectory information to automatically generate the indoor map.The experiments showed that the system can automatically provide the information about road structure and room location in the shopping mall without the help of professional staff or tools,which is necessary for indoor navigation service.