Home // IMMM 2012, The Second International Conference on Advances in Information Mining and Management // View article


Efficient Extraction of Motion Flow Data From a Repository of Three-Dimensional Trajectories Using Bi-Dimensional Indexes

Authors:
Antonio d'Acierno
Marco Leone
Alessia Saggese
Mario Vento

Keywords: information retrieval; spatio-temporal queries; indexing; segmentation

Abstract:
Motivated by the growing presence of acquisition peripherals throughout the world, we propose a novel method for storing and querying moving objects’ trajectories extracted from surveillance cameras. Once moving objects have been detected and tracked, we suggest to store and index the related spatio-temporal data by using an innovative scheme based on widely available bidimensional indexes; moreover, a segmentation stage is performed to increase the overall efficiency. Thus, starting from the limitations of most of the clustering and similarity-based approaches, which restrict the choice of the query parameters, we present a trajectory storing system which efficiently supports Dynamic Spatio-Temporal (DST) queries, which are unrestricted time interval queries over moving objects. For the statistical description of the motion flow in the scene, we use a novel query typology, namely the Flow-DST, that is formulated as a sequence of DST. The experimental results, conducted over real and synthetic data, show the efficiency of the approach.

Pages: 79 to 84

Copyright: Copyright (c) IARIA, 2012

Publication date: October 21, 2012

Published in: conference

ISSN: 2326-9332

ISBN: 978-1-61208-227-1

Location: Venice, Italy

Dates: from October 21, 2012 to October 26, 2012