Font Size:  Small  Medium  Large

Identifying and modeling moving regions: A case study in deep-sea corals monitoring

Marwa Massaâbi, Jalel Akaichi


The increasing availability and use of GPS equipped mobile systems facilitates the collection of data generated by tracking the geographical positions of moving objects in all their forms (points, lines or regions). The generated streams of spatiotemporal data constitute the trajectories of the moving objects. From these trajectories, we can understand the behavior of the movement and interesting patterns can emerge. The challenge in this work is how to identify moving regions from raw data then model their movements in a way that respects their features. Examples of moving regions are for instance, sea surface temperature, animal herds, massive precipitations.
Therefore, in this paper, a new approach for modeling moving regions semantic trajectories is introduced. First, a novel algorithm, called region identi cation based on density-based spatial clustering of applications with noise (RI-DBSCAN), is used to extract moving regions from raw data. Second, the algorithm annotates the moving regions with their movement behavior. Third, outliers are also considered in the algorithm and processed using fuzzy logic. Additionally, we adjusted the moving region data model to establish a more realistic representation. Thus, a moving trajectory data model is proposed. This model will serve as basis for knowledge discovery and prediction of future location and behavior of moving regions. Finally, we used a real dataset to empirically verify the effciency and effectiveness of our solution for identifying moving regions.

Full Text: PDF

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.