An Algorithm for Bus Trajectory Extraction Based on Incomplete Data Source

As the transportation system usage grows, there remains of some challenges. On the one hand, the emergence of intelligent transportation system becomes increasingly attractive; on the other hand, researchers are hardly to access real-time and dynamic information about traffics. In this paper, a new solved approach is designed. At first, real-time data on incomplete public traffics are collected via Internet. Then to achieve reliable extracting bus trajectory, the algorithms are proposed that consists of two aspects: trajectory fragments are generated based strong correlation of original data and trajectory fragments are connected with minimum connection distance. To the best of our knowledge, there are the first results for solving these problems, which will lay on data basis for subsequent traffic forecast. At the same time, the proposed algorithm on eliminating abnormal data and tackle with mass data effectively will provide new knowledge and experience for similar research areas.