3D Visualization And Analysis For Motion Trajectory Data

Trajectory data are increasingly becoming available for researchers due to the advances in technologies like Video Surveillance Systems, GPS and wireless communication devices. Most existing work has focused on the analysis of the shape and aggregation of the track, but little on the trajectory attribute. However, these trajectory data have important applications in many fields, such as architectural planning, traffic management, and recommended routes, et a1. Therefore, it is a particularly important thing to find an appropriate visualizing to handle these.In this thesis, we propose a novel hybrid 2D/3D trajectory wall visualizes trajectory attribute data by stacking 3D color-coded bands on a 2D map to explore the trajectory attributes. we can also to detect trajectory anomalies through an interactive way. A 2D map serves as a reference for the spatial context, and the trajectories are visualized as stacked 3D trajectory bands along which attribute values are encoded by color. Time is integrated through an appropriate ordering of bands, ie, trajectory wall is stacked order by the start time of trajectory. With the temporally ordered stack of trajectories, vertical neighborhood of band segments corresponds to temporal neighborhood of the trajectory points. Trajectory speeds are encoded by color, in our solution we provide an interact speed input mechanism, and we can detect the abnormal tracks by entering some boundary value. To solve the overlapping that maybe occurs in the 3D space, our solution allows the user to rotate and to observe the trajectory and scenes in different angles. For the large amount of trajectory data or those multi category data, we first cluster the data into different groups and then analyze each group by the interact mechanism. Due to the limited height of the screen, we provide some more important interact operations to help the analysts to explore some details and potential information of the trajectory data. The research has been experimented with simulation data and real data, and by comparing with other visualization methods we prove the design is feasible and effective.