Efficient detection of motion patterns in spatio-temporal data sets

Abstract

Moving point object data can be analyzed through the discovery of patterns. We consider the computational efficiency of detecting four such spatio-temporal patterns, namely flock, leadership, convergence, and encounter, as defined by Laube et al., 2004. These patterns are large enough subgroups of the moving point objects that exhibit similar movement in the sense of direction, heading for the same location, and/or proximity. By the use of techniques from computational geometry, including approximation algorithms, we improve the running time bounds of existing algorithms to detect these patterns.

How Do People’s Concepts of Place Relate to Physical Locations?

“…Because these shapes contain multiple Dots, knowledge of the associations of individual dots in the group must be extracted. Spatial and temporal data mining techniques, such as motion pattern detection algorithms [7], may be employed to address these problems. Specifically, for Paths, extending existing spatial clustering algorithm to incorporate temporal data processing may be a viable solution.…”

Reporting Leaders and Followers among Trajectories of Moving Point Objects

“…They defined several movement patterns, including flock (co-ordinately moving close together), trend-setter (anticipating a move of others), leadership (spatially leading a move of others), convergence (converging towards a spot) and encounter (meeting at a spot) and gave algorithms to compute them efficiently. Later Gudmundsson et al [21] considered the same problems and extended the algorithmic results by primarily focusing on approximation algorithms -'Any exact values of m and r hardly have a special significance -20 caribou meeting in a circle with radius 50 meters form as interesting a pattern as 19 caribou meeting in a circle with radius 51 meters.' Benkert et al [5] and Gudmundsson and van Kreveld [20] only recently revisited the flock pattern and gave a more generic definition that bases purely on the geometric arrangement of the moving entities and thus excludes the need of an analytical space as with the initial definition of the patterns [36,39].…”

“…Such a pattern is characterised by two values m which is the size of the set of followers, and k which is the length of a pattern. As mentioned in related work [5,21] specifying exactly which of the patterns should be reported is often a subject for discussion. For instance, a leadership pattern of length exactly k + 1 (starting at time-step t x ) implies the existence of two leadership patterns of length exactly k (albeit 'overlapping', one starting at time-step t x and the other starting at time-step t x+1 ).…”

Moving objects: Combining gradual rules and spatio-temporal patterns

“…A moving object cluster can be defined in both spatial and temporal dimensions: (1) a group of moving objects should be geometrically close to each other, (2) they should be together for at least some minimum numbers of certain timestamps. In this context, many recent studies are interested in mining moving object clusters including moving clusters [4], flocks [6,7,13], convoys [1,2,12] and trajectory [5,11,3,14].…”

“…For instance, let take a look at Fig 1, there are 12 objects moving independently in 10 minutes. Individual moving objects clusters techniques including moving cluster [4], flock [6,7], convoy [1,2,12] and trajectory [5,11,3] detecting are not efficient in this situation. In fact, these techniques require objects to be together for at least some minimum numbers of certain timestamps; therefore, they are not adapted to extract some interesting patterns in this context.…”