Compression of GPS Trajectories using Optimized Approximation Minjie Chen 1 , Mantao Xu 2 , Pasi Fränti 1 1 University of Eastern Finland, Finland; 2 Shanghai Dianji University, China E-mail: {mchen,franti}@cs.joensuu.fi, xumt@sdju.edu.cn Abstract A large number of GPS trajectories, which include users' spatial and temporal information, are collected by geo-positioning mobile phones in recent years. The massive volumes of trajectory data bring about heavy burdens for both network transmission and data storage. To overcome these difficulties, GPS trajectory compression algorithm (GTC) was proposed recently that optimizes both the data reduction by trajectory simplification and the coding procedure using the quantized data. In this paper, instead of using greedy solution in GTC algorithm, the approximation process is optimized jointly with the encoding step via dynamic programming. In addition, Bayes' theorem is applied to improve the robustness of probability estimation for encoded values. The proposed solution has the same time complexity with GTC algorithm in the decoding procedure and experimental results show that its bit- rate is around 80% comparing with GTC algorithm. 1. Introduction Location-acquisition technologies, such as geo- positioning mobile devices, enable users to obtain their locations and record travel experiences by a number of time-stamped trajectories. In the location- based web services, users can record, then upload, visualize and share those trajectories [1]. However, these trajectories often incur a large amount of redundant storage to the end-users as well as the mobile service providers. For example, if data is collected at 10 second intervals, a calculation in [2] shows that without any compression, 100 Mb of storage capacity is required to store the GPS trajectories of 400 users for a single day in server side. To overcome these difficulties, a number of compression algorithms have been presented not only considering the data reduction for visualization purpose but also investigating the encoding process for the storage use. Due to the inherent characteristics in GPS trajectories, conventional error measure, e.g. the perpendicular Euclidean distance is not suitable for GPS trajectories as both spatial and temporal information should be considered. Therefore, the so- called top-down time-ratio (TD-TR) algorithm [3] was developed, where synchronous Euclidean distance was used instead of the perpendicular distance in the Douglas-Peucker algorithm [2]. Threshold-guided algorithm was also proposed via estimating the safe area of the next point using the position, speed and orientation information [4]. In [5], a multi-resolution simplification algorithm has also been designed with linear time complexity. Two error measures, called local integral square synchronous Euclidean distance (LSSD) and integral square synchronous Euclidean distance (ISSD) are used jointly, which can be calculated in O(1) time. Semantic meanings of the GPS trajectories are also considered during the compression process in urban area in [6] whereas trajectory compression algorithm with network constraint has been developed in [7]. Performance evaluations are also made for several traditional trajectory simplification algorithms [8]. It should be mentioned that there is not one algorithm that always outperforms other compression approaches in all situations. However, these methods lack a rigorous analytical approach on the encoding procedures of the reduced trajectories. Namely, fixed bits are allocated after data reduction to store latitude, longitude and timestamp information. On the other hand, when encoding techniques are used, a better compression ratio is achieved for the spatial trajectory data, which is appropriate for data storage. For example, quantization-based approach has been analytically investigated in the so-called vector map compression problem [9, 10]. In these algorithms, differential coordinates of adjacent data points are 21st International Conference on Pattern Recognition (ICPR 2012) November 11-15, 2012. Tsukuba, Japan 978-4-9906441-1-6 ©2012 IAPR 3180