A modified Hausdorff distance based algorithm for 2-dimensional spatial trajectory matching

With the development of mobile technology, Internet and GIS, LBS plays an important role in various applications. As the basis of LBS, the research on acquisition, processing, storage, query of mobile object trajectories has been quite mature, but the analysis and the application of movement patterns contained in mobile object trajectories is relatively lagging behind. The calculation of trajectory similarity is the key area for mobile object trajectories data mining, and the similarity algorithm based on Hausdorff distance is one of the major algorithms. In this paper, we presents a modified Hausdorff distance (MHD) based algorithm for 2-dimensional trajectory matching in light of the problem of correlation algorithms which are based on Hausdorff distance. This algorithm uses interpolation algorithm to shield the impact to the distance between trajectories due to different position updating polices, sampling granularity, initial position and so on in Mobile Object Database (MOD). The experimental results show that the algorithm can better reflect the trajectory similarity between 2-dimensional mobile objects.