| Summary: | 碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 96 === The study of the trajectory identification has been long studied in tradition spatio-temporal databases. In wireless sensor networks, several challenges arise for acquiring the object trajectory including imprecise localized locations and the limited sensor storage and bandwidth. In the current localization technology, the localized locations are inaccurate and imprecise, and the results of such techniques consist of multiple possible object locations. To address these challenges, we propose Possible Presence Zone Trajectory Identification Method (PPZTIM) for extracting the object trajectory from the imprecise localized locations. PPZTIM constructs trajectory based on multiple possible object locations observed/measured by the sensors. The constructed trajectory contains the most probable traveling path of object and all possible object locations observed by the sensors. On the other hand, the constructed trajectory of PPZTIM can represent the object locations with respectable accuracy and contains low data size. We have conducts experiments on a broad variety of object trajectories, including synthetic trajectories and real-world trajectories. PPZTIM significantly reduces the data size of the trajectory by fusing the localized locations, while achieving the user specified error tolerance on the location represented the trajectory.
|
|---|
