Similarity Query Of Indoor Moving-Object Trajectories

Similar trajectory query is a key problem in moving-object trajectory management.The existing research focuses on the similarity measure and moving-object trajectories query,however,the spaces are outdoor or road network.With the variety of indoor location services,the trajectory similarity query of indoor space is urgent for the great demands.The current measurements of trajectory similarity mainly investigates the comparison of character equivalence,which leads to two kinds of results: completely similar and completely dissimilar.So,the imprecise measurement is another challenge for the similarity measurement.In order to improve the query efficiency,the corresponding index structure should suite the query of indoor similar trajectories.Moreover,due to the collection of location data with deviation or uncertainty,the similar trajectory query must consider the effectiveness of uncertainty in trajectory query.In view of the above problems,this thesis studies the characteristics of indoor space and the similarity query of moving-object trajectory.The main works are as follows:(1)Combining with the characteristics of indoor space,propose a method of ITSM(Indoor-space moving-object Trajectory Similarity Measurement)to measure the similarity of moving object trajectories in indoor space.In the ITSM,the space distance algorithm of the indoor trajectory based on the trajectory projection can measure the distance of trajectory efficiently.Meanwhile,the fusion trajectory of location semantics context is designed for indoor location semantic analysis.LSR_Tree transfer the text conversion isotropic comparison to the relationship degree calculation.After that,we put forward a location semantic distance extraction algorithm,which effectively reduce the error of the semantic sequence of the locus position as the direct comparison of the text sequence.The experiments show the correctness and effectiveness of the ITSM method.(2)we present a Top-K query algorithm ITSQ(Indoor-space moving-object Top-K Similarity trajectory Query)for the similar trajectory of indoor moving-object.In the algorithm,We propose a moving-object trajectory index HGB-tree that supports bottom-up query on the similar trajectory query.The HGB-tree adopts a hierarchical structure.Firstly,the similar trajectories are grouped through the moving-object trajectory group division,and the Linked list structure is used to record the trajectory data.Then,the record table implements the query with the bottom-up manner to reduce the computation of the trajectory similarity value and improve the query efficiency.The experimental results show that the ITSQ algorithm based on HGB-tree index has better query efficiency and accuracy.(3)Based on the ITSQ algorithm,we propose an indoor uncertain moving-object Top-K similarity trajectory query algorithm U-ITSQ(Uncertain Indoor-space moving-object Top-K Similarity trajectory Query).In the algorithm U-ITSQ,We establish an indoor uncertain moving-object trajectory model,reduce the computation error caused by the positioning system and the fixed uncertainty threshold through introducing the metrology parameter estimation model,and optimize the trajectory uncertainty caused by the indoor location data sparsity issue through introducing the maximum likelihood principle.The experiment results show the effectiveness of the uncertain moving-object trajectory model and the efficiency of the U-ITSQ algorithm.