Modeling and querying vague spatial objects using shapelets

TL;DR: A new technique for modeling and querying vague spatial objects using shapelets, an image decomposition technique developed in astronomy, as base data type, and an implementation of this data model as an extension to PostgreSQL, including an indexing technique for shapelet objects.

Abstract: Research in modeling and querying spatial data has primarily focused on traditional "crisp" spatial objects with exact location and spatial extent. More recent work, however, has begun to address the need for spatial data types describing spatial phenomena that cannot be modeled by objects having sharp boundaries. Other work has focused on point objects whose location is not precisely known and is typically described using a probability distribution. In this paper, we present a new technique for modeling and querying vague spatial objects. Using shapelets, an image decomposition technique developed in astronomy, as base data type, we introduce a comprehensive set of low-level operations that provide building blocks for versatile high-level operations on vague spatial objects. In addition, we describe an implementation of this data model as an extension to PostgreSQL, including an indexing technique for shapelet objects. Unlike existing techniques for modeling and querying vague or fuzzy data, our approach is optimized for localized, smoothly varying spatial objects, and as such is more suitable for many real-world datasets.