Clustering Multidimensional Trajectories based on Shape and Velocity

Recently, the analysis of moving objects has become one of the most important technologies to be used in various applications such as GIS, navigation systems, and locationbased information systems, Existing geographic analysis approaches are based on points where each object is located at a certain time. These techniques can extract interesting motion patterns from each moving object, but they can not extract relative motion patterns from many moving objects. Therefore, to retrieve moving objects with a similar trajectory shape to another given moving object, we propose queries based on the similarity between the shapes of moving object trajectories. Our proposed technique can find trajectories whose shape is similar to a certain given trajectory. We define the shape-based similarity query trajectories as an extension of similarity queries for time series data, and then we propose a new clustering technique based on similarity by combining both velocities of moving objects and shapes of objects. Moreover, we show the effectiveness of our proposed clustering method through a performance study using moving rickshaw data.